/diy/ - Do It Yourself

This thread's digits brought to you by the Realtek ALC892 audio codec *angry crab noises*

Man, audio frequency envelope followers are hard to design.
This is like my 5th attempt.
It works ok at like 1kHz or more and can follow the envelope at about 150Hz without too much ripple, but at lower frequencies the ripple becomes pretty huge.

>>1685892
Kek

>>1686002
Man, you should use a constant current source to charge and discharge that cap man. It ramps up to the peak voltage quick but then lazily drops off according to the RC time constant of that circuit.

>>1686370
Can you explain that in more detail?
What would the advantage of that be over a RC filter?
Like, I can change the RC constant for a quicker drop off, but that also increases ripple.
How would a constant current source change that?

I found this poor cutie stuck on an old pcb

*noms you* :3

These dimmer modules use a MOC304 optotriac. this chip contains a zero-crossing circuit so it will only switch on when the voltage is near zero. Which means it could only ever be on or off during a whole half-cycle.
Doesn't this completely defeat the purpose of using the thing as a phase-angle control dimmer?
I don't really want to buy one to see.

just wanna say THANK YOU to the guy who helped me through the transistor problem, seems like TR711 was the faulty one holding up the line. The play solenoids work as they should now.
Now on to the next issue. the RWD/FFD solenoid is not engaging.
see image for bad transistor values. they read fine in play/on mode but not surprisingly bad in RR/FF mode.
I assume its just time to replace the transistors on both?

>>1686709
The zero crossing detect isolator there is for AC side -> logic side, you're expected to use the zero cross signal and your own timing to decide when to send the signal to fire the output side.

>>1686717
forgot to clarify, the values are E B C expected and E B C measured (going horizontally across the row)

>>1686723
>>1686717
Bingo! Replaced those two and now the Solenoids for rr and ff work! This is great, going to put it back together a bit so I can try to test a tape

I have a receiver with a 230 watt output, with a recommended impedance of 8 to 16 ohms for each speaker. I have two speakers that each have an 8 ohm impedance, but those speakers a recommended amplifier strength of 10 to 60 watts. Doesn't the 230 watts get divided by the 8 ohms, meaning that the speakers would receive 28.75 watts? I haven't had any electrical experience since I failed the course in high school a few decades ago.

>>1686736
I forgot to add to this - if those speakers and receiver are compatible, what gauge wire should I use to connect them? Should it be shielded and stranded? These are genuine questions and not some form of /diy/ troll.

>>1686736
>Doesn't the 230 watts get divided by the 8 ohms, meaning that the speakers would receive 28.75 watts?
That's not how power works. If it instead was 230 volts, then yes 230V would mean 28.75A of current, but power is power and isn't scaled down as long as you have the right impedance speakers for the job. Since that amplifier is built for 8-16Ω speakers, it should be able to drive the full 230W to your speakers. Though you didn't specify whether that's 230W per channel, or 230W total. In either case, don't drive your amplifier too loud or you'll blow the fuck out of those speakers.

>>1686737
Normal speaker cable that I've seen is stranded, but I don't think shielding is terribly necessary. Go for it if it's cheap enough I guess. Consider that your speakers are 8Ω and there's up to 60W going into them, which corresponds to a current of I = √(P/R) = 2.7Arms. For reference, that's also at 22Vrms. So you'll want a cable that can handle 3A without a sweat, which isn't a terribly high demand. This page here:
https://en.wikipedia.org/wiki/American_wire_gauge
is my go-to guide for looking what size cable I need. According to it's chart, 22 gauge wire would heat up by 40°C if you pumped 3A through it, so jump to maybe 16 gauge at a minimum to somewhere around 12-10 gauge as a maximum. There'll be a common speaker cable diameter somewhere near that 14-10 range, which is probably what I'd go for, but I've never worked speakers in my life. There are probably better online cable selection guides than asking here, so see what you can find.

>>1685971

>>1686744
I appreciate the detailed information. I have 16 gauge wire so I should be able to test everything before heading to bed. Thanks for the help.

>>1686717
cheers m8, congrats. if you get signal and main screen turn on, put on an ancient football game and pour one out for dear old dad

>>1686613
oh I thought you meant the source voltage for the VG, never mind
>inferior to a PCB with a ground plane
that depends strongly on layout and lead dress, of course, and on the encapsulating compound if you decide to get cute
>coupling
not necessarily. the dielectric constant of FR-4 is ~4.5, of PTFE substrate is ~2.1, of vacuum is defined as 1. did you test inside the filing cabinet yet?
>thin coax
yes do this. RG-316 is thin, somewhat inexpensive, yet eminently usable at high frequencies. for each 30m, 3.1dB loss at 1MHz, 60dB at 3GHz (obvs you'll want to keep your line short at that freq)

>>1685892
Off topic, but does anyone know the answer to my question here?
>>>/sci/10993016

>>1686239
Seeing if anyone can still help me with this. Since it seems there isn't a universal standard when it comes to this.

>>1686752
are you able to provide a photo?

>>1686239
it's really a style matter. do whatever you think is most explanatory to the poor sucker who has to build/fix the thing. I'd do a detailed schematic of the system-level wiring on its own sheet and ensure each net has a single, same name in each module it touches and on the harness drawing

>>1686750
Looks highly unrelated to /diy/ohm and already solved.

>>1686752

assuming you're talking about industrial terminal blocks

>would you guys recommend labeling Terminal block names instead of just symbols?

yes, and you should dedicate a sheet to each block/few small blocks per sheet with detailed listings of each terminal and corresponding connection, with sheet reference.

Convention I've always used is

Layout sheet -> TB002 - Sheet xxx
Sheet xxx - > TB002 - TB002:1, TB002:2, etc.
also on Sheet xxx -> Table with columns (something like this - not exact)

TERMINAL | FROM | REF | TO | REF |
TB001:1 | TB033:3 | Sheet yy| PLC01:A2 | Sheet zz |


The way you're describing

>It has "CR-xxx" on the terminal block

is unfortunately common for external (to-customer) schematics but if your internal schematics look like that, you need a new CAD manager.

I'm looking for an open source thermostat breakout that I can control and give temperature data from an arduino mega or a water temperature sensor.

Goal is to chill or heat liquid nutrient with a 12v chiller and a 12v heating element controlled over relays by the arduino. I have alot of other code that needs to run on the arduino so I'd like to outsource PID to another piece of hardware. Any suggestions? Duckduckgo aint what it used to be.

>>1686752

Rule of thumb

>>1686763
>>1686777

One poor bastard tasked with repairing your shitbox should be able to stand outside with a binder in hand and literally read connections off to another poor bastard under/in your shitbox holding a cont. tester.

I bought this wireless mouse a while ago, didn't know it has color changing LEDs. They drain my batteries really fast. The other mouse that doesn't have LEDs can run on worse batteries for months, and this piece of shit runs through better ones in a couple of days, even though I turn it off when I'm not using it.

I want to disable LEDs. I tried short-circuiting them with solder and it worked, but after using the mouse for some time, it just stops working. I have to turn it off and then on again for it to work, but it inevitably stops again.
What can I do here? I'm guessing it stops working because of too much power. Is that possible? I tried using weaker batteries but it's the same thing. No matter whether I disable only 1 or all 4.

>>1686787
>short-circuiting them with solder
remove the connection instead.

>>1685892
Is it safe to use this relay to switch wall outlet power on/off

new to electronics. some things I've found during these past few days:
>I am blind without a multi-meter
>LED magnifying glasses (with different lenses) are extremely useful
>wire/pcb holders are indispensable
>anything you can do with your fingers (or teeth) a tool can probably do better
>eat before soldering
>solder fumes will always move in the direction of your face
>double-check that you've put on heat shrink
>test soldered connections before heating the heat shrink
>put heat shrink on any clamps to avoid injuring wires
>I should have bought more heat shrink

>>1686804
is it real or is it greentext

>>1686797
Thanks Anon, I'll try. It will be hard without that sucky sucky thing.
There's a good chance I'll take it out and not be able to put it back in if it doesn't work. At least then I'll have an excuse to buy a new one.

>>1686801
Is the relay rated for 240/120VAC?

>>1686797
>>1686815
I took them out. So far so good. Will report back if it dies.

>>1686810
I don't find it controversial.

>>1686821

wasnt necessary to remove the LEDs. you can just snip one leg off, if there's room to fit your snippers, or cut the trace on the PCB with a knife.
both options take just 30 seconds to undo (should you have a gay friend over for the evening.)

>>1686720
no, that's incorrect. Read, damn you.
the module provides that optoisolator so the MCU can detect the zero cross and know when to fire the triac. yes.
But the MOC304 optotriac, used to drive the load through a bigger triac, contains a zero cross circuit which does not allow it to switch on until the voltage across it crosses zero. it's not meant for use as a dimmer, it's for switching static loads.
so how and why does this work?

>>1686861
not him but
>someone tries to help and tells you something that's (partly?) incorrect
>you say "Read, damn you" and names the file "fuku"
get the fuck out of here until you're not on your period anymore

>>1686749
>>1686733
well fuck. somehow while assembling some of the pieces I blew BOTH fuses on the power board. Ordered some new ones that wont here until monday so the project is on hold but I hope to god that's the only problem. I'm so fucking close to getting this thing working again if I blew the mains I'd never forgive myself.

>>1686749
>>1686733
well fuck. somehow while assembling some of the pieces I blew BOTH fuses on the power board. Ordered some new ones that wont here until monday so the project is on hold but I hope to god that's the only problem. I'm so fucking close to getting this thing working again if I blew the mains I'd never forgive myself.

>>1686864
>you must accept without question anything that passes for help even when it's completely off-base due to complete lack of attention

>>1686884
Your greentext is a straw-man. It's an incorrect representation of what I wrote.
That's not something I have claimed or implied at all.
If person A wants to help you and unfortunately is wrong about something, that doesn't warrant insults.
That DOES NOT mean the nonsense you just wrote.
Again: wait a few days until you're somewhere else on the cycle.

>>1686886
it's not being 'unfortunately wrong', it's picking two or three words out of a post and then writing a completely off-base reply based on that.
Its worse on reddit, but you people do it too and it gets old.
on the schematic or even on the datasheet you see right away they're using the MOC304 on the load side and not to provide that logic level zero-cross signal.
Kicker is I'm not really looking for help or quick answer, I'm not using one of these I just looked into it because I wanted to see how they designed it and why. I specifically chose optotriacs without the zero-crossing circuit because I wanted to use them for dimming. and here these robotdyn people are using what I'm almost completely sure are the wrong parts, why?
you know if somebody asked me something like that and I replied with such an uninformed and low-effort answer, they'd tear me to shreds. yall people just somehow know when to put that double standard crap on me.

>>1686895
>yall
stopped reading write there

>>1686900
>write
stopped reading right there
ok now I'm done. just shitting up the thread at this point.

>>1686836
Yeah the wires were really short and I wanted to be able to reuse LEDs. Still works.

>>1686804
>LED magnifying glasses
yes. personally I repurposed a 1970s-era sewing lamp and put 6W worth of LED beans in it. when you start to move away from the large-print THT components for the elderly, a microscope is essential, be it a shitty USB video microscope or a large-working-distance double-boom binocular microscope
>wire/pcb holders
correct. pace Murphy, you will eventually need one more hand than you have. hemostats > alligator clips
>solder fumes will always blow in the direction of your face
take a page from snipers, exhale slowly toward the work as you melt. it also steadies the hands
>double-check heat shrink
yes, ditto connector shells
>heat shrink on clamps to avoid injuring wires
sometimes necessary, but connectors are also very helpful. so is polyimide tape. use only solid metal clamps for bending leads, ideally with rounded jaws, to absorb all shock and stress to the component body seal rather than just reducing it
you have enough basic sense to go far in the hobby. Asian father tentatively approves

>>1686886
it's a correct interpretation of the context, autie
>doesn't warrant insults
yes it does. this is a low-bullshit general and we prefer to keep it that way. bleach-ammonia crystals go to >>>/b/

>>1686895
zero-voltage switching and dimming are incompatible unless you really like flicker. that's good enough to "dim" most heater applications, not good enough to dim a lamp
look for documentation. those modules usually have PDF schematics available for only the cost of a little web research

>>1686921
>reuse LEDs
y tho, they're a penny a piece in hundreds. let your rainbow flag fly, my dude

My latest attempt at an envelope follower.
This works pretty well at all frequencies, but there's about a 1.4v error since I didn't bother replacing all the diodes with super diodes.
Input frequency range is +/- 1 octave from the sampling frequency.

I don't know why I'm even still trying to implement this with analog electronics.
It's basically just DSP now and a Attiny85 would be much cheaper than this while performing much better.
y[n] = MAX( x[n], x[n-1], x[n-2], x[n-3] ).

>>1687030
Shit, forget the image.

>>1686786
>>1686777
>>1686763

Appreciate it. To cover what it is, I just picked up lead position in a panel wiring department and I have gone over the drawings, harnesses and components and it's just a fucking disaster. To me everything should be labeled that any John Contractor can walk in and fix it. At the same time allow my team to actually do their damn job, which with all the guess work these drawings leave isn't happening. Just wanted to bounce it off some others to see if I may have been overstepping with my views.

>>1687030
Mix the wave up by a few hundred kHz and then put a simpler envelope-follower on it, you'll be able to set the corner frequency pretty precisely without worrying about the frequency of the waveform changing drastically.

>>1687129
Oh, that's a good idea.
What's the cheapest way to do the mixing though?
The way I'm familiar with is using a Gilbert cell or a ring modulator.
Gilbert cells are a little too expensive for my application, and I don't really want to use a load of transformers either.

>>1687172
Dual gate MOSFET or JFET cascode mixer seems like it'd be a good fit for this application. Can be implemented without transformers.

>>1687179
Is that a 4 quadrant mixer, or will I have to bias my signal?

>>1687172
Transconductance amplifier is my bet.

>>1687183
It's a single balanced mixer and you need to apply some bias to your LO.

>>1687086

was reading about the Deepwater Horizon disaster. 2 of the 6 batteries in the Dead Man's switch were dead coz of miswiring. a solenoid valve was also miswired in the same unit.
so what's the connection with you? if you make drawings so explicit that even a moron can read them, then morons will be hired and kept around until they cause deaths and millions in damage.

and your cohorts will undoubtedly label you '"Capt. Autism" for wanting the change The Way Things Have Always Been Done™.

So if I understand correctly, a superheterodyne receiver mixes the radio signal by some hundreds of kilohertz to get to an intermediate frequency, which is amplified and then used with an envelope detector. Ignoring any filters. If you tried to mix the signal with a local oscillator that was instead set to the transmitting frequency, even being a few Hz out could cause the signal to sound terrible, which I assume is why they used an intermediate frequency. But these days, couldn’t you put a PLL on the incoming RF signal (after the tuning circuit) to make a sine exactly equal in frequency to the transmission but without a modulated amplitude to feed into the mixer? Perhaps with a phase shift in there. The advantage being, no diode detector would be required, nor any RF amplifier. The PLL may require some frequency dividers (on both sides so they cancel out) in order to average the frequency better, but I don’t think it’s needed. The tuning circuit at the start would ensure the PLL locks onto the right signal.

>>1687200
Ah, thanks.

>>1687197
Right, I always forget about those. Don't get enough excuses to use them.
I'll probably go with an OTA mixer since that only requires a single OTA and some resistors.
I need like 12 envelope followers, so cost is kind of a driving factor for me.

Since I've been kinda vague: I'm working on an analog 12 band vocoder.

>>1687223
>then morons will be hired and kept around until they cause deaths and millions in damage.

This sort of assumes that they wouldn't be hired and be kept around in the first place. That's a very, very bad assumption to make in Oil&Gas.

Always chart for the Lowest Common Denom.

>your cohorts will undoubtedly label you '"Capt. Autism" for wanting the change The Way Things Have Always Been Done™.

Not wrong, but (old) cohorts label any young engineer/tech/manager something. I'd rather be Capt. Autism then Capt. Jackoff.

>>1687223
So they had no quality control prior to the product going in? Because your point is stupid. As it stands the errors with these products are not found by the people wiring them as they have little to no understanding anyways of how electronics work. They following a drawing and wire it up. You aren't going to get someone who understands how electronics and components work for 15$ an hour like the vast majority of these people.

And to be labeled captain autism for making sure the product is readable and correct isn't a negative. It falls on the dipshit who hired the dumb ass in the first place and thought it was a good idea to do no QC, at any stage and not even do a burn-in test on any of the equipment before hand.

>>1687230
Oh looks like it's not a new circuit, even with the PLL. The method of directly mixing the transmitted frequency with the radio signal is called a "Product Detector". It's also apparently required to listen to SSB, which is something I did plan on doing.

>>1687281
>You aren't going to get someone who understands how electronics and components work for 15$ an hour like the vast majority of these people.
There's hope for me yet!

>>1687284
From what I've come to understand, if you have little to no knowledge of reading a drawing, but understand how to crimp and solder you can land a 17-20$ an hour job in most places. In fact that longer I work in this field the more I realize the vast majority, including the engineers, have no fucking clue what they are doing.

I've been looking for recommendations for tooling my local hackerspace, recently.
Someone suggested the Hantek 2C42, and it looks cool as fuck.
However, I don't know what the stats on o-scopes allow you to actually do. What kind of analysis is usually done with an o-scope, and what does 40MHz of bandwidth with 250MSamples/s allow?

Furthermore, does anyone have experience with the model itself? Is it a good tool?

>>1687288
What kind of shop would I be applying to in order to do so, though?

>>1686521 those can have a nasty bite

idiot here.

if I take two 5 volt portable batteries used to charge cell phones, strip some usb cords and tie the batteries in series, can I use the combined batteries as a big 10 volt one?

>>1687432
You should be more confident in this answer. What happens when you run two batteries in series? Look up the results.

I'm wanting to control 2x 12V DC, 0.3A fans. I was planning on splicing a USB cable and using a phone charger to power this. Would this work or will it blow my face off? The description just mentions the 1.8V, 3V, 5V, 6V, 12V Voltage range and max 2A current. Also should the wiring to the fans be parallel?

you get their combined voltages, what I was wondering is if this would work with two daisy chained cell phone batteries

>>1687439
With a boost converter it should work fine. Boost the charger voltage from 5V to 12V, then feed that to the two fans in parallel. Perhaps with a speed controller in between. Note that a 5V-12V fixed boost converter will be smaller and more efficient than a variable converter.
And if you don't post a datasheet or purchasing link it's hard to say anything about what you're referring to in that picture, though it looks like a speed controller.

>>1687490
Thinking about it it'll probably just be easier to buy some 12V power supply, they're dirt cheap.
As for the listing:
>https://www.ebay.co.uk/itm/DC-Motor-1-8V-3V-5V-6V-12V-PWM-Speed-Controller-2A-Potentiometer-Knob-Switch-SF/362720085707?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649
The listing doesn't say much beyond
>Input Voltage: DC1.8-15V
>Output Voltage Type: Load Linear
>Output Current: 0-2A
>Continuous Current: 2A
I just bought it, £2.69 isn't going to bother me.

>>1687494
Yeah that thing will almost certainly work. Going for a 12V PSU in the first place is probably the right choice.

>>1687497
I almost forgot: As far as I understand if you turn down a potentiometer the excess will be released as heat. I've also bought a small 65mmx45mmx25mm plastic box for this controller, it won't get too hot will it?

>>1687499
The potentiometer is not being used as a variable power resistor, but rather to set a threshold that the onboard controller reads to tell it what duty cycle to PWM the motor at. PWM doesn't produce much waste heat at all compared to using a resistive speed limiter. Google PWM if you don't understand.

>>1687501
Ah that makes sense. Thanks for babysitting me, I feel like I've covered all bases now.

>>1687432
Yes but there are details, e.g. you have to take it apart to recharge them

>>1685892

>>1687386
cute, but I would suggest something that won't so easily walk off
spring for the Rigol and a proper bench meter or two

>>1687523
next time post pin 1 in the lower left corner so that the marking reads correctly

>>1686801
Did you died?
>>1686818
He died.

Is the CD4046 PLL unique in that it has an internal VCO that you can set the range of solely with an external capacitor? Because all the higher frequency ones I’m looking at require external VCOs, all the circuits for which look like a bit of a nightmare.

>>1687716
what do you need a PLL for

>>1687730

>>1687730
>>1687732
weapons grade lighting

>>1687730
I bought the biggest module digikey had at the time, was round about 4cm diameter. Man that shit was bright.

>>1687736
filename

>>1687741
probably going to run these at something like 30% power to light my living room. putting together a driver for these rn

>>1687723
To try my hand at demodulating AM or SSB SW radio, perhaps even demodulating FM broadcasting. The 4046 only goes up to I think 12MHz, which isn't enough to cover all of SW (which goes up to 30MHz or so), and is certainly insufficient for ~100MHz FM demodulation.

That circuit you posted should work for the SSB radio work, though I'd somewhat prefer to be pushing a sine into the mixer instead of a square wave, the required filtration would be less. But the VCO is required for negative feedback of the control voltage for FM demodulation.

>>1687751
>anon brings a girl home
>"what's that noise, anon?"
>oh it's just the coolant pump station for my lamp
>turn on the lamp
>"ANON WHAT'S HAPPENING? IT'S SO BRIGHT, I CAN'T SEE!"
>s-sorry

>>1687763
y tho, the harmonics will be so far out of band you can suppress them with a fairly low-Q filter. besides, you'll lose 1-(pi/4) = 21% of your signal strength with a sine multiplier. otoh you'll let fewer harmonics thru, but if you filtered them with a gross tuning tank you shouldn't see much of them

>>1687792

>>1687800
My plan was to use a tuning tank after the antenna, and a simple RC low-pass filter between that and the audio amplifier (which itself should block high-frequencies). No need for a tank circuit after the mixer, is there?
Sure I'll lose some signal strength by using a sine wave instead, but will my S:N get any worse?

>>1687807
oh, not for a no-IF receiver. a lowpass on the output would be fine
SNR would be slightly better with the sine, 1-2dB on average, but depends on whether there are signals at odd harmonic frequencies e.g. the 2m ham band's third harmonic is in the 70cm ham band, transmissions on the upper band could come through on the 3rd harmonic of a square LO. the tuner needs to have enough Q to attenuate that. and your post-filter should have a sharp enough cutoff to prevent adjacent channels from interfering. both required values are in the upper single digits or so

New to electronics but a few questions
Trying to make a wireless dog shock collar function similar to a video game controller rumble pack, the two things I need to do are send the signal from the game (which might not have output) to a transmitter and then from the receiver to the shock collar.
Atm im leaning towards using a Saelae logic to pick up the cue (somehow) and send it onwards to an arudino mini hooked up to a cheap dog shock collar 433MHZ receiver.

Are the there any third party programs that monitor cues from games and transmit them to output information I can send as a signal? And would it be better to get a Bluetooth transmitter and do it via Bluetooth instead?

>>1686717
>>1686749
>>1686733
>>1686873

Ok back in action, fuses replaced, all systems are go. popped in a tape and.. BAM. First time anyone has seen this in probably 30 years! holy shit

Okay almost there, but might be scope time. The picture is very rainbow-y and fuzzy, audio comes in out and and sounds faint. COULD be the tape, could be the adjustments on the boards.

I have the following NON-scope adjustments:
>color lock pot
>peaking pot
>skew and tracking front knobs
adjusting them helps at random but the picture isnt there yet.


Still to do:
>FFD does not trip the auto-stop (good) however the motor/belts do not seem strong enough to move the tape
>RWD trips the auto-stop (bad) meaning something is still wrong in the circuit. could be expecting a tape to move one of the capstans (only tested rwd with no tape)

pic related is what a color image looks like and a black and white (there's a swtich to select on the front). really hope its not a scope tune-up..

>>1687820
>your post-filter should have a sharp enough cutoff to prevent adjacent channels from interfering
Say I'm tuned into 500kHz AM, and hence feeding a 500kHz square into the mixer along with it. That square will have harmonics at 1.5, 2.5, 3.5, etc. MHz. The only component anywhere near the audio range will be the 500-500 one, the others will be in the 1MHz, 2MHz, 3MHz, etc. range. It's barely worse than it would be just with a sine, which would also have a lowest harmonic at 1MHz. So all I need is to get a filter that has enough of a cutoff from ~10kHz to ~1MHz. Worst case scenario it's between 20kHz to 500kHz, a factor of 25. A ~3rd order cascaded passive low-pass will sort that out, right?

How would /ohm/ improve upon this idea?

>>1687843
shielding

>>1687845
>shielding
You're reading my mind. But how... maybe wrap it with something but with what...

>>1687846
fill the inside with hot glue and wrap it with aluminium foil, then cover with heat-shrink. Have fun getting the heat-shrink over there in the first place. I'm assuming those are unbalanced lines?

>>1687848
they're RCA connectors. I don't know how you'd balance something like this.
yes, heat shrink is out of the question.

>>1687849
Fusing silicone tape? Just a bunch of vinyl electrical tape?

>>1687850
yeah, but shielding is still an issue.

here's a similar idea. this guy used 90deg connectors that have a female end. I'm guessing he just inserted a piece of copper wire in the middle and used the hose clamps as the outer connector.
I prefer the soldered version myself but this one is better shielded.

>>1687851
I mean still have the insulating filler hot glue/silicone/silastic/etc. with alfoil wrapped around it, and put the tape atop that. I'd fold the alfoil up to make it more solid and less crinkly.

>>1687852
My best idea so far is to use a silicone rubber on the inside and then cut a piece of copper tubing down the middle and just sandwich it.

>>1687854
That could work reasonably well if it was the right diameter, but wouldn't you want it to be connected to the outer of the two connectors?

>>1687857
yeah, I'd sweat it with solder like you'd sweat a water pipe.

>>1687858
Around the threads of the grey part, or around the inner gold-plated part?

>>1687859
this is what I'm thinking.
solder it around the threads part. that's connected to ground anyway.

>>1687860
So long as you sand the coating off and its solderable, I guess that's fine. But if it's aluminium you'll have a bad time.

>>1687862
yeah... you have to sand the copper pipe when fitting plumbing too. a bunch of flux should make it adhere well.

I'm just thinking that maybe I'm missing something even more obvious...

>>1687863
I meant sanding those threads on the existing part.

>>1687866
right.
there's also this idea but this is too complicated. or maybe it isn't...

>>1687868
That one has inferior shielding, not that it probably matters. The copper pipe method is likely to work if that metal is solderable, so before anything else just try scratching a bit of the coating off, covering it in flux, and see if any 60/40 wets to it. But I'd use an iron, not a torch, as there might be plastics inside the right-angle piece that wouldn't like the higher temperatures.

>>1687869
good point. that dark metal could be zinc alloy.
yep, that PCB has no shielding.
also, I found this attempt... lol

>>1687869
>inferior shielding
if it's a 4-layer board, it's bretty gud. you'd think the audiophiliacs would put a bit of effort into their scams

>>1687825
bitchin. did the broadcast signal alignments go well? did they stick when changing from video source to video source?
mag tape does degrade over decades. peek under the head plug and make sure the tape looks ok
if they're front panel controls, play with them. old VCRs required such adjustments, sometimes per-tape. wacky/missing colors were one of the symptoms. adjust them first to get the best picture, then look at the internal adjustments
a weak motor would be at the root of all sorts of trouble. are all the right solenoids moving to engage the mechanism for the ff mode? that right-hand one looks like skew and tracking should clean it up
scopes do take a lot of guesswork out of the job, to be sure

>>1687829
adjacent channels, meaning e.g. 510kHz. suggest you make the LPF adjustable, if you can manage to do it without a $10 triple-ganged pot. the correct AF filter cutoff frequency depends on the channel spacing in the band of choice

>>1687876
duellingbanjos.flac

>>1687883
>adjacent channels, meaning e.g. 510kHz
I was hoping the tuning tank circuit would be high-Q enough to dodge those, but it may not be if they line up just the wrong way. My research tells me that the frequencies are spaced about 10kHz apart, with a bandwidth of 20kHz each, how is this possible?
I can always dampen the square wave's harmonics with a filter after the zero-crossing detector, perhaps a variable one ganged to the same capacitor that tunes the tank.
I have a multi-gang tuning cap I got hold of for pretty cheap lying about, I think it would work for this.

I can always get some RF amplifier or buffer ICs with which to improve the Q factor and add active filters. But still, the VCO method would be preferable to avoid those harmonics.
Also trying to figure out if the 90° phase shift is actually necessary with a PLL, because some rudimentary graphing tells me that it would just leave the signal inverted without it.

>>1687883
>put a bit of effort into their scams
it's not a scam. the whole point of this is to avoid long cables and make a compact connection between a DAC and amp so you can push it against the wall or something. and having a mechanical link between them means they won't slip when you push buttons or turn the knob.

How do I calculate the output voltage for a pressure sensor before the signal is amplified by an instrumentation amplifier? It has a sensitivity of 0.75 mV/V/psi with a maximum pressure of 30 psi. I'm supplying the sensor with 5V. Can I just multiply these three numbers to get Vout?

>>1687888
>tuning tank circuit Q of 50
seems iffy, especially if tunable over a 20% range
>what did they mean by this
broadcast stations are interleaved to prevent interference from anomalous propagation and to protect stations near one another from stomping on each other's sidebands and causing that groovy frequency-inverted sound, which is important in a crowded radio market like NYC or LA
>90° phase shift
assume the VCO is a square wave oscillator and that the PLL input signal has been squared off. now NAND them and filter to get a control voltage, which is at a steady 3/4 of max when locked, and rises and falls proportionally to the actual offset. if the VCO drifts a little slow relative to signal, the control voltage will rise slightly and the VCO will speed up slightly, thus taking up the slack and restoring the phase relationship, and vice versa. thus the PLL follows the phase of the input signal and can self-correct for (theoretically) up to ±90° of input shift. in practice the phase difference between any two cycles is much less than 1° even for broadcast FM. to maintain 0° or 180° offsets, you need to know and care which rising edge came first to know which way to pull the VCO, implyign 2 control loops and crossover distortion

>>1687902
you sure can

>>1687906
No I know why a PLL's operation means it has a 90° phase shift, that's just how its XOR mixer works. I was more wondering why an analog mixer's two signals need to be in phase with one another. Like with an SA612 or diode ring.

>>1687883
i dont have a scope yet thats the thing. and i dont want to spend $$$ for a one time project unless i can find one cheap in chicago.
i also need a twin lead->75oihm adaptor for this unit since it only accepts 300ohm picture IN (but can do rca and 75ohm out..) so i cant even test to see if broadcast/ntsc picture looks ok passing through the unit. ebay time.
the tape itself could be fucked up for all i know, i only have one to test right now, will get the rest out of storage later this week. skew/tracking did help but ther'es still a constant fuzz along the top edge of the picture.
audio noise was coming from a dirty monitor/rec slider, used some contact cleaner on it.
as for the FF/RWD a new observation, neither of them work when a tape in inserted, meaning either the belts are slipping, or the motor isnt strong. (ffd works with no tape, rwd doesnt work either way). I will have to check the lines, and transistors again tonight to make sure all the rr/ff components are working.
There are two motors, a large drive motor in the back and a 7v take up motor in the front. idk how hard it would be to replace if it needed it.
next steps might also be to look for new belts, just gotta figure out how to measure them.

>>1687971
one more sample image, showing the rainbow banding on color mode when it should be just regular color picture. this tape looks like it was taped in 1984/85

is there an electronics challenge list equivalent to this?

>>1687973
yes, old, often criticized, never updated.

>>1687978
oh well, I'll take something over nothing

>>1687988
Voilà.

>>1687997
wow these are all over the place

What's a fast programmable microcontroller in 2019? I'm working with something that provides a brief signal every 6 or more milliseconds. When the signal is revcieved on one pin, I need it to read the voltage of two other pins and provide an output signal for 1-3 milliseconds based on these voltages. Is it even possible to read and execute everthing in such a brief window?

>>1688054
for MCU's you're probably better off asking the arduino general, but there's a lot of crossposting

>such a brief window
several ms is plenty. you can make a basic circuit that'll do what you want

>the more you learn about analog electronics and electronics in general the more you learn about how finicky and limited they are and how much of a pain in the ass they are to design, and that no matter what digital is the only way to go for literally anything more than a toy
wtf not based or cool pilled

>>1688102
>everything you grew up with is normal
>everything that's not normal is wrong

>>1687971
did you look inside the tape to see if the edges were starting to ruffle? have you cleaned the head with 99% IPA and a Kimwipe? (don't use cotton swabs directly on the head, the fibers will fuck your shit up)
get a little mirror on a stick so you can see what's going on down there. Harbor Freight has Chicagoland locations right?
>measure belts
place belt on flat surface, hold two pencils upright, separate until they pull the belt into a narrow oblong with as little stretching of the belt as possible, mark, measure
if you can't successfully google up a "suitable" (forget about exact) replacement belt, another trick you could try is heading to your local auto parts store and picking up some P.B. Blaster, which is sometimes used to rejuvenate automotive belts
mechanical alignments should be in service manual territory

>>1688054
>milliseconds
jeez grandpa
the STM32F030 would be much more than plenty. it's got a core that can do big math at a suitable speed (no hw multiply, sorreh), a 12-bit ADC that can be automated (some STM32F have two or more), and most importantly its timers are fire
you may be able to do this with a 555 and a few more components. need more constraints. need more input

>>1688102
>t.hipster
many magnificent things have been done even within those limitations. open-loop gain and negative feedback absolve many sins. however, it did tend to cost quite a bit and cooperated heavily with the macro-scale physical domain, using odd materials and components that are hard to get and use today thanks to Western deindustrialization among other reasons, and isn't *as* practical anymore for that reason
physics also don't help. the bulk of any interesting analog or digital circuits that you can get to work on a plugboard can be done in software with much less fuss and alignment. analog is really only interesting at speeds where traces need to be treated as transmission lines

>>1687999
show tits and coat hanger walker

>>1688102
>digital is the only way to go
those digital circuits are created from analog components
you haven't even started learning

I have a problem with this speaker, it's never been used, 11 yrs old and doesn't work. No impedance, no continuity. Could it get bad in this time span? I got it out of a car with no radio unit but it had the whole speakers and cables setup for a proper one.

>>1688146
>Could it get bad in this time span?
it was bad already - the difference is that it's now also not functioning.
also yes.

>"cardboard" speaker cones
>poly frames
you should have tossed it years ago.
don't repair it. get a proper speaker driver.

>>1688102

>He hasn't gotten into transmission line dynamics

The digital world is an illusion created by error correction and D̶a̶r̶k̶ ̶B̶l̶o̶o̶d̶ ̶R̶i̶t̶u̶a̶l̶s̶ creative engineering

>>1688144
the bulk of the analog behavior of digital componentry can be ignored tho

>>1688152
damn right you keep quiet about the dark blood rituals that don't exist

>>1688151
The frame's metallic but that don't matter, how would you repair this? I'm curious if I could get it working just for fun.

>>1688155
>I have junk, wat do
did you even try google

Hi, I have a problem with my bass amplifier. Might be the wrong general for this, idk. I haven't used 4chin in years. If there's a better general or if I should just make my own thread lmk. Google doesn't know wtf I'm trying to ask here.

So a while back after leaving my bass amp (similar to the one in the picture) at my parents' house for several months (untouched), I came back to find it no longer working in what I think is kind of a weird way. As I turn it on, for a brief moment there's sound and then it quickly fades away (in less than half a second). Same when I turn it off. So like I can pluck the string, then turn it on, and only for that moment as it turns on do I hear anything. Then same as it's going off, I can pluck the string while the amp is fully turned on, hear nothing, flip the switch and for a brief moment as the LED 'On' indicator is fading out the sound will pop on and fade out just the same. It's almost like it's only working on low power or something? Idk I don't into electronics that's why I'm here.

At the time I did try opening it up and it had some weird case setup and I literally couldn't get it open. And I'm not a moron I swear, I've opened other amps before. It's not supposed to be rocket science. If necessary I can try again now that I'm older and know to look up the manual online and stuff but I want to know what I'm looking for first.

If anyone's got any leads on what it might be lmk. I can upload video if you think you know what's going on and want to see exactly what's happening instead of my autistic description of it.

Thanks guys.

>>1688163
btw the amp in the picture may or may not be the same exact model as mine i just pulled it off google cause it's same brand and looks similar. in case that's relevant at all.

>>1688160
I saw those speaker repair videos, it's a verry shitty job in the way of its outcome.

>>1688134
I learned about stuff traveling through the 'ether' but I learned waves, not whirls.
¯\_(ツ)_/¯

t. seventy-six year old

>>1688163
Wtf why is it starting to work when you power it off? I'm thinking the transistors are gone or something. They allow current through, regulated by a voltage so as a capacitor is charging (when you power it on) it tells the transistor to switch on but in this case after a certain threshold the transistor stops responding? When you power it off the capacitor is discharging and it reaches that sweet spot the transistor is able to let current through. Is this how bad transistors work?

>>1688180
>learned waves not whirls
Dang Confucius

>>1688186
sadly, they couldn't explain what the ether was - just that it was what carried the electro-magnetic waves

>>1688154
Maybe at like 20 MHz you can ignore analog behavior, but as you go up in frequency you quickly have to start considering analog effects on the signal.

>>1688102

today's troubleshooting tip: place a 0.1-0.2mm sheet of polypro (e.g. a food storage bag) atop a circuit to find hot spots without an IR camera

>>1688163
you haven't told us fuck-all about your amp, but that sounds like some stage is way the hell off its operating point, and the most likely cause is mechanical (loose connector, cracked board, loose component, etc.). are there tubes? if so try pulling them and reinserting them into their sockets. find a service manual for your exact model (you cagey fucker) that shows you how to pull the chassis so you can have a look around inside. or even a youtube or some random amp modding site
but if you don't into electronics, that's probably about all the user-serviceability you can expect

>>1688180
I, too, read Fear and Loathing in Las Vegas
t.45yo

>>1688185
shut the fuck up, Donny

>>1688246
>I, too, read Fear and Loathing in Las Vegas
>t.45yo
I've heard of it and know a little about it but haven't read it.

>>1688163
post pic of circuit board

When I use multiple dc power supplies off a common ac outlet, do the dc ground wires need to be jumpered together?

My microcontroller runs off 5v but it will control mosfets and relays for the 12v and 24v PSUs

>>1688264
I forgot to connect dots for Live wires in this diagram btw.

>>1688264
Yes the 0V pins should probably be commoned, unless you need to make a split rail with them. But if they have a connection from the negative rails directly to the ground input then you perhaps don't need to worry. I'd still be a little wary and want to connect them all the same, provided you avoid any ground loop situations.

Then again, if it's measuring current with a shunt between the output and ground it could even be detrimental to connect the outputs. I doubt they'd design isolated PSUs in this manner however.

It would be out of spec to have the USB shield at a different DC potential than the USB 0V, right?

took apart a vacuum cleaner to fix a part of the engine. had to cut the two wires supplying a somewhat high current to it.
re-soldered them and covered the internal cable with heat-shrink when I was done with the engine.
the thing runs now, but after a few minutes of testing the vacuum itself stutters (electrically) and even the LED power button on the extension cord outlet (connected to the vacuum) flickers on and off when I turn it on.
did I fuck up the solder?

>>1688272
Maybe it would help to see more of the circuit.

The arduino analog, digital, and serial devices will all receive their power from the 5v sensor terminal block.

The i2c devices will as well, however the 8 channel relays and the tb6600 motor controllers will also use the 12V 600w supply for their loads.

My understanding is that my LED PWM controller needs to jump to the arduino ground and the 24v 600 ground. I was just thinking about running a jumper to all negative outputs on each power supply and the arduino gnd pin.

Not shown here are mosfets, the power wiring for sensors, resistors, and kickback/rectifier diodes on pumps/motors/solenoids.

Is there like a wiki or catalog of some kind that documents old fluke test tools?

something like http://w140.com/tekwiki/wiki/ that i use to learn about older tektronix equipment.

im trying to buy test equipment and having a hard time finding anything on fluke without knowing exact model numbers first.

>>1688288
you need a crimp connector.

>>1688251
for shame

>>1688283
nah, happens all the time. that said, it's customary to dc ground USB shields only at the upstream. use a cap to the shield if you want ac shielding in your device

>>1688305
diamonds. you in a legal state?
to answer your q, no, but you absolutely can, but note well they might already be coupled (or not) thru the power supplies to mains earth, directly or thru a blocking cap

>>1688308 here, a couple more questions for those more experienced than i.

so i have been looking at multimeters, oscilloscopes, and power supplies on ebay, and it seems pretty much anything in my price range is gonna need recalibrated. ok. so i have expanded my search to include voltage references, voltage calibrators, references meters, null meters, handheld field calibrators. theres a lot of options out there. anyone have any advice on what i should probably be looking for here?

also, will a calibration shop have non corporate prices for the guy who is buying equipment from the 50s-80s on ebay and using them at home for ham/learning?

>>1688322
Fluke's own calibration service, for example, offers four levels of service, from a basic performance test with no paperwork and not even a sticker, to sworn on the blood of Christ Himself and a Mount Sinai of paperwork to prove it. cost doesn't appear to be listed on their home website, maybe send a quick note to a branch near you or look into other shops that do post their prices openly

>>1688322

>multimeters, oscilloscopes, and power supplies
>pretty much anything in my price range is gonna need re-calibrated.

Meh. I don't know what you're going project-wise for but even most uni labs will let things drift a little, unless you're talking about some OG 70's stuff. Anything digitizing/digital will still be well within meaningful tolerances IMO.

If you really want to - rent a calibrated function generator and a voltage source from a local equipment dealer. I honestly have never known a company, small lab, or person to own any of that equipment unless they were actually in the business of repair and recert. Someone nearby you has something, almost certainly.

Resistors are different - decent reference resistors aren't arm-and-leg expensive. you're not looking for fractional PPM accuracy, so you can grab some .01% thru-hole for 5 bucks, chassis mount (which are actually really nice to have) for about 20 bucks. If you rent some "starting" equipment - Func. and Voltage - and own the resistors, you can get everything nice, and use the resistors to circle-jerk your equipment to keep it in spec for years.

>>1688305

>When the EE student wants to grow some weed in his rental

>>1688305
holy shit thanks for posting this. i had a rough schematic like this in my head and you laid it out so nicely.

>>1688305
Link to different LED colours?

>>1688288
>did I fuck up the solder?

it sounds like you have an intermittent connection. it could definitely be a bad solder joint, but it could also be that you tugged on the wires and some connector (to the motor or the power switch) became loose.

the trick to good solid solder joints is to make them mechanically strong enough to resist being pulled apart even BEFORE you apply solder. if you just solder the wires side-by-side, that's asking for trouble.

>>1688305
i would tie the grounds directly between the supplies to cover your ass. as others have said the supplies could already be DC coupled to earth but most supplies aren't in my experience. be sure not to make a usb connection from a PC to this full circuit unless you're sure it's not tied to something funny on the mains in its finished state.

also you can trim this down to one mains PSU with a cheap 24->12v switcher and a linear regulator for 12->5v. adjust power ratings accordingly. find a zener+fuse board so a chinesium power setup doesn't fry your grow op. idk how flammable the overall plant is, but reading up on wiring standards for class 2-3 hazardous locations could save your life

also also you'd know better than me but city water supplies are always pretty cool so you might be able to skip the water chiller.

>>1688288
>did I fuck up the solder?

it sounds like you have an intermittent connection. it could definitely be a bad solder joint, but it could also be that you tugged on the wires and some connector (to the motor or the power switch) became loose.

the trick to good solid solder joints is to make them mechanically strong enough to resist being pulled apart even BEFORE you apply solder. if you just solder the wires side-by-side, that's asking for trouble.

and if your iron cant handle big wires, you can always use a lighter or a candle to melt solder you wrap around the joint.

>>1688288
>did I fuck up the solder?

it sounds like you have an intermittent connection. it could definitely be a bad solder joint, but it could also be that you tugged on the wires and some connector (to the motor or the power switch) became loose.

the trick to good solid solder joints is to make them mechanically strong enough to resist being pulled apart even BEFORE you apply solder. if you just solder the wires side-by-side, that's asking for trouble.

and if your iron cant handle big wires, you can always use a lighter or a candle to melt solder you wrap around the joint. and if you're not using flux (or solder with flux inside it), that's a recipe for failure.

>>1688246
>you haven't told us fuck-all about your amp
idk what info is relevant and what's not

Anyway now that I've looked there unfortunately doesn't seem to be a manual for this type of amp (or any of their bass amps) online.

I can't seem to get the electronic part out because it's connected to the speaker through a hole in the wood (and the whole wood frame is glued together). Is there a usual way to disconnect that that won't involve my soldering it back together after? Pic related.

Also now that it's out of my closet and I'm fucking with it again, I realized it actually only makes sound while being powered off. Not while coming on. Cause it doesn't take a moment to "fade in" to being fully powered like it fades out when it goes off ig. If it matters.

>>1688305
Yeah I live in Colorado but this system isnt going to be growing weed for many years until my lettucecraft is perfected.

>>1688337
If you wanted to grow weed in a rental youd save 4 grand by doing it without these expensive ass electronics.

I wish I did EE, I'm a computer engineering associates. I can code this thing and I know enough with EE to be dangerous. Like this should probably be 2 or 3 phase AC.

The goal of this machine is to accelerate plant growth down to objective programs that schedule ideal horticultural conditions. I want to build a machine that makes 2d dirt farming obsolete because we are going to have alot more people and we're already seeing signs.

My other goal of this project is to shit on ivy League kids building less capable food computers at MIT. Im also reasearching into methane enrichment for processing Titan's atomospheres into breathing air and water via plant cycles.

>>1688347
Its open source, but 4 years past due. I call it æden but it started as project EDEN in 2015 hackaday prize, made semifinals but I had to move states.

Eventually going to be posting everything but I need about 2g in parts to machine before its done. Making it modular if you dont need the fancy stuff. The LED array alone is better than most of whats out there.

>>1688385
>way to disconnect that

are you kidding? those are spade terminals on the speaker. you pull them off with pliers, and push them back on by hand. if you dont even something that basic, how are you gonna fix an amp?

>>1688384
third time's the charm?

>>1688392

how do you know that? is there a plugin or do you have a time machine?

>>1688396
4chan x shows deleted posts
also I saw them before you deleted them

>>1688396
time machines, like free energy devices, are actually pretty easy to make if you're an experienced EE. we just don't tell the normies.

>>1688391
i just don't want to break anything. i thought that might be the case but i wasn't just gonna start pulling shit apart and risking it lol

>if you dont even something that basic, how are you gonna fix an amp?

with /diy/'s guidance i hope! and i'm hoping it's something simple like a wire came undone or idk something is just fucking dusty? like I said it was literally sitting at my parents' house, no one moved it or touched it or anything and i came back to this problem so it's not like something catastrophically broke apart, i have no idea what to expect.

which part do i unclamp? the part around the rubber of the wire or the part lower?

>>1688403
google what a spade terminal is

>>1688352

PAR LED Array

LEDs
-3W Far Red LED (740-745nm) w/ PCB x50 (30) https://www.ebay.com/itm/222373169319
-3W Deep Red LED (650-660nm) w/ PCB x100 (60) https://www.ebay.com/itm/321788397858
-3W Red LED (610-630nm) w/ PCB x100 (60) https://www.ebay.com/itm/321356962292
-1W Blue LED (460-465nm) w/ PCB x100 (60) https://www.ebay.com/itm/221806724965
-1W Royal Blue LED (445-455nm) w/ PCB x100 (60) https://www.ebay.com/itm/221806724965
-1W Violet LED (400-405nm) w/ PCB x50 (30) https://www.ebay.com/itm/252459680066
-3W Full Spectrum LED (400-840nm / White) w/ PCB x50 (30) https://www.ebay.com/itm/253156296962
-3W Full Spectrum LED (PAR) w/PCB x100 (60) https://www.ebay.com/itm/321834649301
-LED 45 degree collimator lens / holder x360 https://www.ebay.com/itm/221642202319
-150W Boost CC PSU 10-32V to 12V-60V x10 (8) https://www.ebay.com/itm/162434501864

-1W 1 Ohm Resistor x50 https://www.ebay.com/itm/302233718061
-5W 15 Ohm Resistor x10 https://www.ebay.com/itm/173362608941
-5W 5.6 Ohm Resistor x10 https://www.ebay.com/itm/371842259017

>>1688409
-TLC5947 PWM LED Driver
5v @35mA
+IRF244N N-Channel MOSFET
-(8) 150W Constant Current Boost Converter
Input: 10-32V @16A Max Output: 12-60V 8A Max

-(30) 1W Violet LED (400-405 nm, 3.2-3.8V 350mA)
42.78W 24V @1.78A (78%, 3.2V Ea, 5 strands of 6 LEDs)
+(5) 15Ω Resistor (5W)
-(60) 1W Royal Blue LED (445-455nm, 3.0-3.2V 350mA)
85.57W 24V @3.56A (79%, 3.2V Ea, 10 strands of 6 LEDs)
+(10) 15Ω Resistor (5W)
-(60) 1W Blue LED (460-465nm, 3.0-3.2V 350mA)
85.57W 24V @3.56A (79%, 3.2V Ea, 10 strands of 6 LEDs)
+(10) 15Ω Resistor (5W)
-(30) 3W Full Spectrum Horticultural LED(380-840nm, 3.2-3.4V 700mA)
85.12W 24V @3.5A (83%, 3.4V Ea, 5 strands of 6 LEDs)
+(5) 5.6Ω Resistor (5W)
-(30) 3W Full Spectrum White LED (400nm-840, 3.2-3.4V 700mA)
85.12W 24V @3.5A (83%, 3.4V Ea, 5 strands of 6 LEDs)
+(5) 5.6Ω Resistor (5W)
-(60) 3W Red (620-630nm, 2.0-2.2V 700mA)
86.45W 24V @3.5A (97%, 2V Ea, 5 strands of 12 LEDs)
+(5) 1Ω Resistor (1W)
-(60) 3W Deep Red (640-660nm, 2.0-2.2V 700mA)
86.45W 24V @3.5A (97%, 2V Ea, 5 strands of 12 LEDs)
+(5) 1Ω Resistor (1W)
-(30) 3W Far Red (740-745nm, 2.0-2.6V 700mA)
51.87W 24V @2.1A (97%, 2.4V Ea, 3 strands of 10 LEDs)
+(3) 1Ω Resistor (1W)

>>1688409
that's a whole lot of nanometers

>>1688406
I did, there are multiple different designs and I didn't see one that looks just like this.

I know it's a stupid question I'm a stupid boy.

>>1688403
>which part do i unclamp?

grab the point shown by arrow with small pliers and pull up. wiggling helps.

>>1688411
Plants absorb mostly red and blue light. I used to work as the head electrician for a vertical hydroponic farm. What I would observe in lettuce that was grown under red or blue light is that strains of lettuce with traditional red coloration would come out green.

This inspired me to add violet as UV can cause coloration to improve but I didnt want to cause as much damage.

In flowering deeper reds like Far Red on the far visible spectrum can help.

I added full spectrum white for coverage of green spectrums without wasting power. I also added PAR full spectrum (pink led) for low power operation.

>>1688413
Thank you! I got it.

Here's the internals people asked to see. Lmk if you need to see anything closer up or want me to test something.

>>1688428

everything looks in tip-top shape. would need schematic and instruments to diagnose further. the two tubes on the RHS are probably fuses covered in plastic. if you can tell the diff between a good fuse and a blown one, you might wanna take an x-acto knife and remove the plastic cover to see the state of the 2 fuses.

>>1688435
>if you can tell the diff between a good fuse and a blown one

Could that cause the problem where it only makes sound as it's turning off? And if I post a picture after removing the rubber could people here see whether it's blown?

Also here's a picture where I marked where everything is in case it helps anyone.

>>1688441
>Could that cause the problem

yeah. the thing runs from a split supply: a positive voltage, a negative voltage, and a ground which is halfway between them. if one fuse is dead, then it becomes a single supply and weird unpredictable things will happen.

>>1688446
If I post a pic without the burka will you/anyone be able to tell me from looking if it's blown?

>>1688454

chances are excellent that you wont be able to get a clear shot from the side. using a mirror, look for signs such as a blackened area, or a discontinuity (break) in the wire. it can be very subtle, so you need good eyes and good light.

>>1688455
I started opening it and there wasn't glass like in that image it was just exposed wire and that didn't seem right so I stopped. Luckily (?) the amp is exactly how it was before. Idk if that means they're not fuses or if they're just a different kind of fuse.

Also I found another weird behavior - if I plug in headphones first of all they don't work but secondly sometimes after unplugging them, the amp will stop producing sound entirely. Then if I touch just the tip into the jack it'll go back to ""normal"" (ie; only produce sound as it's turning off).

>>1688380
Thats enough confirmation for me. I'll ground everything out.

Connecting to the system from a computer wont be an issue. The Pi runs a webserver and I've configured it to host its own hotspot that I can connect and access it from the web or a terminal.

I used to have it on one PSU however the issue with that design was that 1000W+ PSUs typically want 220V past that. I have found used server supplies (god bless hp) that get the job done but if I decide to manufacture this I'd like to use newly sourced supplies.

But at 1300W on a single AC outlet I'll admit I worry about my apartment wiring. Once I fire this thing full blast I want to know its on its own breaker and all by its self. I may need to research 2/3 phase supplies better.

The benefit of two PSUs is that I can make the LED array a standalone piece of hardware from the irrigation system.

>>1688461
They're probably chokes then, not fuses. No need to have 3 fuses anyhow. Pic related. Look for fuses elsewhere, though it is possible that the only fuse is on the primary side of the transformer. I'd specifically be looking for dodgy parts around where the power comes in, since the "working fine for a split second when starting up" does sound like a split-rail or input capacitor/rectifier fault.
If I know what I'm looking at, that's a centre-tap transformer, each tap going through one of three chokes, which then are rectified to both a positive and a negative rail with a pair of 2-diode centre-tap full-wave rectifiers. That yellow box between the diodes will likely be an X suppression capacitor on the AC side as opposed to a DC filtering capacitor of some kind, but in either way it's a bit odd since I'd expect two of them for the centre-tap / dual rail power.
Check continuity of the diodes, also on the chokes if you want to but they're probably fine, check capacitance of the three caps and ensure they have a DC resistance that approaches infinity, and check the resistance through both transformer taps to each other and to the centre itself. Can't think of anything else on the power side, aside from measuring the supply voltages when it's turned on.

I'd also consider checking out where mechanical stresses might be happening, especially the jacks, switch, and power socket. Perhaps also where the PCB itself can flex and see if any solder joints are cracked.

>>1688471
>I'd specifically be looking for dodgy parts around where the power comes in, since the "working fine for a split second when starting up" does sound like a split-rail or input capacitor/rectifier fault.

If you didn't catch it I corrected myself and it's actually only when it's powering down. To me it seems like it's only working when the power is low. Like as a capacitor is draining. As an analogy if I look at the LED "On" light it kind of shows that, as soon as I flip the switch the light comes on but when you flip it off it fades out (as opposed to going right off). Only during that fade out does it produce sound.

Video:

streamable dot com slash 1h1gt

Link broken up cause idk if 4chan spams streamable links.

>>1688477
Same still applies, whether powering on or powering off. Could also be some sort of (incorrect) biasing of the internal transistors that's altered during powerdown.
Do any parts feel particularly hot after leaving it on for a few dozen seconds?

>>1688352
not him but cheap af on ali

>>1688380
I'm much happier to see grows overbuilt than burnt down. pharma merchants don't need any pretexts
>chiller
in aquaponics the nutrient bath temperature needs to be held in a narrow window to avoid root rot on the one hand and root shock on the other. usually, wort chillers meant for brewing are often used

>>1688389
nut
>shit on ivy League kids building less capable food computers at MIT
based af anon. come on up to MI sometime for a bowl and a look at the cannabis abacus :^)
>all those LEDs
that's one way to cover some bandwidth I suppose... but the new wave of microgrowers claim good-CRI 4000-5000K white LEDs are most efficient LEDs in PAR per $TCO, with some red-end augmentation optional during flower (warm white?). even beats blurples. there's a band around 500nm or so that you might need to fill in if it's photosynthetically active in Cannabis, which I do not know

>>1688419
>UV
blue and violet LEDs are special eye hazards due to the eye's weaker pupillary response to them. be careful
it would make for an interesting study to see what the dose-response function of the UV looks like, whether it's a curve, line, or a step
>cooling fans on your LEDs
lel, see >>1687751 (60 out of 240W installed)
>newly sourced
that's what EEs are for. Pic vaguely related, hacked up bench copy of my 100W CC LED driver, being used to qualify components and ply truth from datasheets (see bodge wire right of Kapton tape). having been steeped in the old ways for decades, I'm a bit allergic to duino and don't trust ESP chips for life safety, so homebrew it is. since I2C isn't a great long-haul bus, I'm running CAN bus (cannabus?) and +48V on four-wire cable, connecting small semi-autonomous nodes at point of use (I2C/switches/relays/etc). a Linux box connects to a CAN interface on the bus, logging measured values directly from the bus into a time series database and sending out commands/config as needed

>>1688481
Sort of like pic related. Im making sliders that tune this graph to the power given to the LEDs.

If you google 'par led chart' it should show you more finely plotted stats for chlorophyll consumption.

My eyes where fine in the grow but the previous electrician wore UV glasses. After working in pink light all day everything looks green for the first 30 seconds or so.

I've seen those big boi leds but honestly with 600W of light in less than a square meter I've already maxed out the ppfd absorption rate several times over. It would be interesting to try them though.

Oof, these RF ICs are more expensive than I expected. But their specs are quite something, like the very utilitarian (and very high frequency) AD8302 phase detector, or the good and simple MC1648D VCO. I think I'll hold off buying any of these for the time being.

>>1688489
thanks for the handy search term. the trough I had in mind was at 550-600nm. lots of carotenoids to worry after too
>big boi
um, they're 19mm on a side. 60 dice inside each little square, in 5 series strings of 12 dice each. think of the wiring you save
one point of potential interest for water-cooling your LEDs is that you can reduce air exchanges otherwise needed just to remove excess heat, to keep your CO2 around for longer. I've seen people go to the extreme of using car radiators/fans and an external chiller to control the environment inside a partially sealed area for just that reason
continued from my last post, I hear raspis are hard to source in quantity. for mass production, look into the orange pi series which is a rough functional equivalent. there are many variants which mix and match the peripherals offered by the cheap set-top-box chips. all boards have Ethernet and/or wifi, a few have gigabit Ethernet. some have MIPI DSI, some have MIPI CSI. some have HDMI, others have composite video output. all have 26 or 40 pins of GPIO with some serial buses muxed on them as well, generally following the raspi pinout standard. tech support for the better OS distro (Armbian) is cranky at times but it shouldn't be a complete brain-buster for someone who's seen a couple of years of comp eng
further, you may be surprised how little just a few dozen custom pcb assemblies costs these days, thanks to chineese automation. an engineer can throw a bunch of cookbook circuits onto a board, string them together, and call it a product
>600W/m^2
noice
>UV goggles
I mean, people are employed at glow in the dark mini-golf places. here's Cree's take on it fwiw, including an item-by-item listing of hazard classes https://www.cree.com/led-components/media/documents/XLamp_EyeSafety.pdf
another tip: check out the BME280 air temp/pres/humi sensor, it's super convenient

>>1687741
holy shit I'm running one of these at 125mA * 32V = 4W and maybe that's enough for one room

>>1688492
there's not a whole lot of room in the consumer world for all those discrete-ish functions. integration is the order of the day there. those sorts of chips usually wind up in test equipment and maybe the odd repair job that absolutely must work according to original spec

>>1688517
Doesn't stop there from being a dozen sellers on ali. And there have to be a few hundred varactor sellers.

>>1688520
I'm sure. surplus happens. seconds happen. "deep recycling" happens
>varactors
sucker, use a 1N5822 or a 1N4007, or red LEDs http://www.hanssummers.com/varicap.html

>>1688532
>red LEDs
I think I'd avoid those just because they'd photovoltaically pick up optical interference. Though if they're in an enclosure I don't see a problem.

>>1688492
>MC1648D VCO
If you can read past the headline you will notice that the 1648 itself is not a VCO.

>>1688542
It's a sinusoidal oscillator driver that requires only an external inductor and capacitor (and/or varactor), it's close enough to what I want. Most enclosed VCO ICs I've seen have square wave outputs and only deal with particularly high frequencies, so one that I can easily tune the frequency across several orders of magnitude just with a couple of passives is preferred. Far better than fucking about making my own Colpitts oscillator, though doing so isn't out of the question.

>>1688322
a ham shack doesn't need calibration in the metrology sense; you don't have a process to control, there aren't a hundred assembly stations all working in parallel to produce identical parts. you only need your instruments to more or less agree with each other and your power supplies to output roughly what they're indicating.
with instruments like o'scopes, universal bridges, tube testers, curve tracers, or spectrum analysers used in a troubleshooting role, the result is more of a fortune cookie than a hard number.

>>1688141
The tape itself looks fine, but a bit shiny on one half. From what I can tell, the audio plays just fine but the video is just static at worst or fuzzy at best. I'm thinking its probably a bad tape. The head can actually be removed, so im getting some gloves and carefully cleaning it. see: https://www.labguysworld.com/Quasar_VR-1000.htm
There's 4 main belts, 2 flat, 2 square, The drive motor ones seem okay, not gooey or stiff, but again it seems the tension and weight of the tape itself is too much for the machine to ffd/rwd. it will play at normal speed fine.
Last night i actually rewound the spools by hand, took a while but now the tape is back to the beginning.
If i could find a "20mhz, 2 channel, ac/dc coupling" oscilloscope in chicago I might be able to get somewhere.
pic related is the tape on the machine with the top off. film is ran CCW to the top spool

>>1688305
jesus fuck m8 I grew some tomatoes in a box with a weed pink light and just did a circuit that turned shit off if it got too hot so my rented student room wouldn't burn down. People these days just want to put PWM in every single application even tho you don't need it kek

Really stupid question here.
Let's say I have a BLDC motor 3 phase and I want to make a driver. Motor doesn't have hall sensors.
Can I detect zero-crossing in BEMF, and "shift" to next stage.
For ex.
My current phase output is "0 + -", zero crossing happens, and it switches to "- + 0".
Next zero crossing and "- 0 +". Etc.
For this I can use simple counter and simple logic. I want to avoid microcontrollers at all cost, because everyone can make a motor spin with Arduino... Wanna try without this.

I will make a schematic of my madness a lil bit later

>>1688602
one more image showing the tape visually, a bit of sheen on it the upper half

>>1688640
Also, I have another stupid question.
I'm too retarded to understand how proper bootstrap works. Can I use pic related as FET driver?
Will that SMPS even work?

>>1688577
what oscilloscope function is that called?

>>1688556
>I can easily tune the frequency across several orders of magnitude
But not continuously. Expect two different inductors per decade. Can you imagine why that is?

>>1688661
https://www.jammarcade.net/simple-component-tester-a-k-a-octopus-curve-tracer/

For my job I'm going to have to program microcontrollers. AVRs to be exact.
https://www.digikey.com/catalog/en/partgroup/atmega1284p-evaluation-board/33326

Would this be a good kit to start out? I already got a USB programmer and some jumper wires.
Or should I go the arduino way and program that with an AVR IDE? If that's even possible that is.

>>1688533
put blackface on them

>>1688606
bet it wasn't any better than regs

>>1688642
um, if that's not just the end of tape signal, that's the oxide wearing off of the tape

>>1688676
or get an Arduino Mega, erase the bootloader, and use the chip/board like a grownup with your choice of MPLAB or a vanilla avr-gcc install

>>1688640
>I want to solve an engineering constraint
>solve my engineering constraint for me guise
get used to making shit that doesn't work at all and maybe even blows up in your face because spoon-fed success doesn't teach fuck-all and historical reenactment doesn't teach much at all about how to practice in current year
t.doing this since I was a kid and now old enough to be your daddy

>>1688692
>bet it wasn't any better than regs
I welcome you to try homegrown
I don't buy vegetables from stores because we grew our own as I was growing up.

>>1688309
>>1688384
very useful information. thank you.

I have a small DC motor designed to run off a single lithium cell. Under load it draws around 6A.
I want to power it from a laptop power brick and bring the (average) voltage down with PWM. But can those power supplies deal with that kind of load?

>>1688705
fair enough. the gardener is an important variable, possibly THE most important variable
>/ohm/ - weed general, electronics edition

>>1688737
lol no

>>1688737
Use a buck converter. by a cheap chinese one rated for 8-12 amps and it should handle 6 amps with no fire.
PWM would probably work but a buck converter is a better solution.

>>1688663
I think maybe I could get it down to one if I pick the right varactors, some of the ones I looked at had a ratio of 14. But the same thing applies to the CD4046, which has a completely built-in VCO. A ratio of even 1:3 min:max frequency would be sufficient to cover AM broadcasting, but to also cover all of SW up to 30MHz I would need to switch my inductor. Perhaps I'd be able to have the same oscillator do feedback for FM too, if I lay out my circuit right, but I'm unsure how the real people switch RF signals. Resistors in series with FETs to ground?

>>1688737
If you put an LC low-pass after your PWM circuit then maybe, so long as you limit your duty cycle. But that basically just gives you a

>>1688692
>get used to making shit that doesn't work at all and maybe even blows up in your face
MOSFETs are expensive.
>solve my engineering constraint for me guise
This is not an engineering constraint. This is "I don't fucking know theory" thing.
When it comes to motors and transformers, I'm not that good... And I don't understand how commercial BLDC BEMF controllers work, and I don't even have oscilloscope to see shit with eyes.

But from those diagrams it is obvious they have some delay between zero-crossing and switching...
I don't even know what I should google first

>>1687223
I learned something however I’m not sure i agree with you conclusion. Not saying you’re wrong but in automotive controls I’ll take proper documentation any day.
>or the contractor paid to cleanup the mess will charge you to create their own

>>1688812
>switch RF signals
PIN diodes: the dc bias of the diode determines whether it will also pass or block ac. a b-c junction is almost a PIN diode if you squint hard enough, and 1N4007s have been used as poor man's PIN diodes too

>>1688819
Would those work well enough at a few hundred kHz?

>>1688813
Ok, apparently I can.
So das is easy

How long does it take to get lead poisoning from soldering? At my last job I had to spend ages in a small enclosed room soldering tons of shit.

>>1688669
well this is fun. I can't believe I never heard of this circuit before.
thanks for sharing.

>>1688820
maybe, I haven't personally tried them. forward and reverse capacitances may have considerable effects, and "on" and "off" are not nearly as absolute as with, say, a signal relay. fwiw, for band switching, people usually use signal relays to avoid detuning their band filters, saving diode switching for places where capacitance value is less of a concern.

>>1688829
was it leaded solder?
did you wash your hands and rinse your mouth afterward?

>>1688832
Hands yes, mouth no. I'd never even heard of rinsing your mouth.

>>1688640
>Can I detect zero-crossing in BEMF, and "shift" to next stage.

Not with a "dumb" bang-bang, it won't. there's a delay required after commutating to allow for Zero-crossing to occur "cleanly", which depends on the feedback itself to work properly.

Also, Back EMF coming from a BLDC is notoriously dirty, so signal conditioning and some logic is required.

>>1688813

phase delay between ( polarity off)-(zero cross BEMF) and (zero cross BEMF)- (reverse polarity on) and vice versa in synchronous operation are both basically 30 degrees. ( Polarity off) - (reverse polarity) is 60 degrees, set by your control.

you want synchronous operation. if operation stops being synchronous (i.e. you've loaded your motor), you'll see that polarityoff-zerocross and zerocross-reversepolarity change. you then update the frequency of drive to reestablish syncronous operation. Same goes if you suddenly unload motor.

Loading motor -
>BEMF frequency decreases slightly
> positive off - zero cross offset increases
> zero cross - negative on offset decreases
> decrease drive frequency to match offsets and increase torque to handle load

unloading motor -
>BEMF frequency increases slightly
> positive off - zero cross offset decreases
> zero cross - negative on offset increases
>Increase drive frequency to reestablish synchronous operation.

digest that for a bit and you'll probably see how to do it, at least roughly. a comparator and a timer will be your friend.

Keep in mind that virtually none of this works at low speed or on startup, so bang-bang that sucker blind until you hit a certain speed

>because everyone can make a motor spin with Arduino

Lol no they can't. duty cycle control of a brushed motor or open loop stepper control is one thing. Ad Hoc Arduino for BLDC with feedback is non-trivial.

You either want to solve a discrete logic problem, or you want to solve a programming problem. The phase driving and feedback aren't going away in either situation.

>>1688829
theoretically, more years than your natural lifespan. but don't take medical advice from 4chan! go get the hair test.

>>1688871

>He died at 30? didn't he work in that chincy shop what with the bad ventilation?
>Well, I guess 30 was his natural lifespan.

all jests aside, hair test is the way to go.

>>1688873
It's probably not that bad, because it wasn't my only job - I was only doing it for a week at a time every other month, but I might get the hair test if I do much more soldering in future, because I also probably get some lead exposure from shooting.

I also learned just before I left that they used cadmium connectors, which are apparently really toxic. They probably taught all of this stuff to the normal workshop staff, but I was only there occasionally so I guess they just didn't bother.

>>1688881
t. dead man walking

>>1688881
soldering isn't hot enough to make vapours of lead OR cadmium.
(you can boil mercury with a soldering iron. protip: don't.)

>>1688888
kek's honest truth. but spatter and dust could be a concern if proper control protocols weren't followed. I pity the sad sack running the vacuum cleaner

>>1688692
>um, if that's not just the end of tape signal, that's the oxide wearing off of the tape
very possible, these tapes were not stored in good condition. its possible the whole tape is ruined anyways. but since i cant ffd/rwd yet i have to let it play though all the way to find out. Currnetly looking for new belts to order.

>>1688901
a bad tape could be behind a lot of the mechanical trouble and it will certainly never give a clean picture
you could just play forward for a few minutes which will definitely pass any sort of leader tape that might be present. then look. if it still looks like that, probably a bad idea to load that tape again

>>1688864
>Not with a "dumb" bang-bang, it won't. there's a delay required after commutating to allow for Zero-crossing to occur "cleanly", which depends on the feedback itself to work properly.
But what about delay before zero-crossing? It is obvious from that graphic.
Maybe it is better to detect fall and rise too? I have seen videos of people probing BLDC motor controllers on e-bike, and at full throttle, "ass-shaped" waveform appears, which might confuse rise/fall detection...
>Also, Back EMF coming from a BLDC is notoriously dirty, so signal conditioning and some logic is required.
This sucks, since I don't have a scope, thus I can't make a filter. This would be a problem with PWM current limiting for sure.
But on the other side, I think those motors don't get more than 200-ish Hz, so low-pass should do.
>you then update the frequency of drive to reestablish syncronous operation.
This is the idea of motor control.
You can make them spin in really dumb way, just by going though that 6 states with somewhat reasonable low frequency, but this is not useful at all, you can't ever run fan from such motor.
>Lol no they can't. duty cycle control of a brushed motor or open loop stepper control is one thing. Ad Hoc Arduino for BLDC with feedback is non-trivial.
Proper motor control without hall sensors is hard to make on Arduino. It is too slow to reliably detect zero-crossings.
For that you need to go really low-level and use AVR-C or even asm without any gimmicks. It is possible, because commercial ESCs on Atmega168 and 328 exist.
>You either want to solve a discrete logic problem,
Probably this, much faster than compiling shit, and finding lost ";".
>>1688829
Die young. Fuck living up to 80 years, getting Alzheimer, shitting pants and such. Better die at 40 from lead poisoning combined with cancer from flux fumes.
>>1688888
Checked. Boiling mercury will destroy soldering tip. Don't do this. They expensive

>>1688577
>a ham shack doesn't need calibration in the metrology sense; you don't have a process to control,
actually, i do understand and play with process control as well. i have a fluke 789, a wally box, deadweight tester, etc. i also want to start getting better at electronics by buying broken test equipment and restoring it to life. some of that stuff needs cal bad.

>>1688910
no its not the leader for sure, its defintely the tape. the audio plays clear but the picture is all static and sometimes comes in and fades out again. I have more tapes in storage that I'll get this weekend, seems like this one is just sour after surviving a basement flood and 5 years in a outdoor storage unit lol

>>1688914 here, sorry i hit post on accident.

point is i do have some pneumatic cal and test gear, fluke 789 which does 4-20ma sourceing, and i play with PLCs. im trying to expand my abilties and knowledge though into amateur radio and electronics in general. so i have been digging into electrical and rf test equipment. and as i said, i want to buy cheap test equipment and learn how to restore it to life, and id like to be able to calibrate it to ensure its actually functional as intended.

>>1688829
>How long does it take to get lead poisoning from soldering?
f o r e v e r
https://www.quora.com/Can-I-get-lead-poisoning-from-inhaling-solder-fumes

>>1688829
just stop licking your fingers and you'll be fine, anon

>>1688926
I like licking PCB traces.

>>1688928
RIP in peace anon
F

Hmm, unless I made a dumb mistake, passing the input waveform through a mixer didn't seem to help with envelope following.

>>1688929
I hope so

>>1688913

>But what about delay before zero-crossing? It is obvious from that graphic.

Did you read the rest of my post?

>Maybe it is better to detect fall and rise too?

You can run a differentiator but it really isn't necessary unless you're doing something more sophisticated. the zero crossing detect will naturally tell you high-low vs. low-high if done correctly.

> I have seen videos of people probing BLDC motor controllers on e-bike, and at full throttle, "ass-shaped" waveform appears, which might confuse rise/fall detection...

1.what you have in pic isn't back EMF, that's a drive waveform. E-bike drivers running on simple battery setups generally don't go for true trapezoidal control for a variety of reasons, so they compromise by mixing two sines with a slight offset per motor phase. It's better to do it this way if it's a PMDC motor, anyway.

Will the "butt" show up in back EMF? yeah, hence the filter. if that's all you have to deal with, you should actually be happy - full trapezoidal controls generate some stupid harmonics.

>This is the idea of motor control.

of Synchronous motor control. and even then, only in very, very simple cases.

Plus, if you understand this, then you should understand how to set up what I'm talking about.

> It is too slow to reliably detect zero-crossings.

Interrupts on a Mega can easily pull 10Khz. That gets you north of 4000 RPM. I've seen RC guys use BEMF feedback on an stock Mega before. you don't run the BEMF to the board directly, obviously, but having the Arduino do the majority function and delay calcs for you isn't a bad idea.

>>1688935
I think you got trolled linearly

>>1688942
Well, at least it didn't take too much effort.
I guess it kinda makes sense in retrospect.

I think I'll just slap a filter with a big RC constant on the low frequency envelope followers, not like they have much bandwidth anyways.

>>1688935
use a much higher frequency

Differential probe anon here, after fixing the capacitor problem by using 3 100nFs. No shielding yet. I've tied both inputs together and to the virtual ground with an (unfortunately long) alligator clip, and am measuring a tiny amount of 1.2MHz noise from the boost converter. The noise has a main fuzzy body of ~10mV with ~20mV spikes. If I read correctly, the ripple rejection of my LF347 would imply that this ripple wouldn't be a problem (100dB typ, i.e. V/V = 100,000) but perhaps its Gain BW product of 4MHz means I'm cutting it a little close. Because when I measure the voltage from this virtual ground to the output op-amp, I get in the realm of 80mV of noise. All these measurements are peak-to-peak, with the probe in 10x mode. This same measurement is visible from Vee to the virtual ground, and from Vcc to the virtual ground, and from the virtual ground voltage divider to the virtual ground, and from the output to Vee and to Vcc. The actual noise on the voltage divider compared to Vee is half that between the rails (as expected). So all this noise is coming from this virtual ground circuit, even though my divider has two 100kΩ resistors with two 100nF capacitors, making something approximating a corner frequency of ≤100Hz. The "less than" being me trying to instead account for the input impedance of the buffer itself.

Holding a 1µF electrolytic across the rails makes no impact at all, but putting it from Vee to the output of the virtual ground buffer halves the noise. A 1µF ceramic(?) works slightly better than the electrolytic, but not by too much. The length of the wire being used to connect these may be an issue, holding it in place more precariously lets me get this noise down to about 9mV. I also spy a noticeable 12MHz harmonic, which is highly influenced by my oscope itself. Having caps after my virtual ground feels a bit odd but it's not as bad as putting them after an AC-outputting op-amp. I'll see about chucking a choke in there to make a low-pass or 2.

>>1688996
>try inserting a 100µF while it's running
>sparks
oh shit
I'll see about soldering on these caps later and see how it performs like that. Not too sure where the low-pass elements should go, but a choke or two around the power lines couldn't hurt. I guess I'll also see about measuring for noise that's specifically common-mode or normal-mode, though my second probe is just a shitty dual-banana one and I'm not even sure if the wire before that is coaxial.

>>1688996
>choke
try adding a µF across the pins of the op amp itself
have you ruled out radiated noise?

My normie frends wants me to make him blinking lights that blink with music.
I know how to do everything apart from processing the music itself.
I am getting signal from a headphone output, how do i take that and parse it so i know when to turn the lights on or off?
do i measure voltage or something?

>>1689004
Pic related is the idea: filter off all but the bass to emphasize the beat, then peak-detect what remains, then use the output to set the LED brightness
this is just a conceptual sketch, and is very sensitive to the input level and the genre of music. the outupt detector's time constant will need to be much shorter for bluegrass than trip-hop, start with about 50ms and adjust. the input filter's corner frequency should be roughly 100Hz, for gangsta rap you might replace the input low-pass with a band-pass filter. you might want to replace the passive peak detector with an opamp-based one. also consider a "brightness" pot to adjust the output before the LED controller
protip: pull the filter frequency, "brightness", and output detector TC out to front panel pots

>>1689011
well i am going to use some microprocessor for that like atmega, but my confusion is that if i am getting an audio wave, then the height of it (peak voltage) is volume, simple enough, but how the fuck do i detect things like bass or treble? do i just sample the frequency of the wave for say a millisecond and if i is say under 100hz it's base?
but you can have both base and trebble at the same time in the audio wave too and it's just very confusing to me

>>1689013
oh, you're thinking in the time domain. you need to work in the frequency domain
read up on the most important numerical algorithm of our time https://en.wikipedia.org/wiki/Fast_Fourier_transform
the math gets a bit heavy but the general idea is that you give it a series of samples and get back a series of complex coefficients representing the magnitude and phase of frequency bands in the sample (or vice versa), which you can mix/match/feed to timers as you please
note, you might need a bigger boat than the ATmega because FFT computation still takes a lot of math and RAM, depending on block size. for musical blinkenlights, a 5ksps sample rate is probably enough. whatever micro you use, make sure it has a hardware multiplier

>>1689013
>how the fuck do i detect things like bass or treble?
With filters to seperate them.
A capacitor passes high frequencys better (see:capacitive reactance), so in that anons drawing, the capacitor and resistor on the left act as a "low pass filter" by using the capacitor to pass the higher frequencies to ground, leaving the lower frequencies.

To detect higher freq's use a "high pass filter" that blocks lower freq's

>>1689001
>have you ruled out radiated noise?
No, but probing the inductor of the boost converter gives me noise that looks identical but at ~15Vpk-pk, so it's reasonable to assume that the booster is where my noise is originating from. What is harder to imagine however is how that's being passed through the IC's ripple rejection, let alone amplified through a buffer amplifier. I can only assume that it's subpar frequency response near the 1.2MHz is to blame. The 1µF across the rails does nothing of note. Adding a ferrite bead on the Vcc rail doesn't improve anything significantly. A 220µH in series with the rail didn't make a difference either, not even with a 100µF cap afterwards. I'm guessing it's at too high a frequency for the larger inductor's core.
Adding load resistors does nothing. Putting a 1µF between the output and one of the rails causes the rail to turn from a noisy (50mV) approximate DC to a fairly high-amplitude 32kHz triangular/sinusoidal wave. Since it's dependant on capacitance I can only assume I accidentally made an oscillator with that extra feedback path.
Throwing an RC filter after the virtual ground seems to do some good, so I'll likely do that (maybe with an RC pi filter), with a corner frequency chosen such that it won't affect common operation. 200kHz perhaps.
I'm considering throwing an LM317 in there to regulate the voltage down a little just to get that ripple rejection and see where my problem really lies, perhaps tomorrow.

>>1689004
>My normie frends
Are you their abnormie?
low-pass + band-pass + high-pass = red + yellow + blue
The file is called AudioEqualizerPresentation.pdf

>Press chip into socket
>press socket into thumb

AAAAAA

>>1689011
If you put several filters in series, does the each filter double the filtering effect? so if the first cuts out say 50% of the low frequency, does the next filter cut out 50% out of that first 50%?
I tried to run some simulations and the filtering seems to be annoyingly gradual instead of a flat line going into a nice deep ravine so i hope that by adding one more or two more filters in a row this will yield better results

>>1689169
So looks like there is no point to use more than two

>>1688937
Took me a while to understand...
I need to measure BEMF between phases, not neutral, otherwise I have 30 degree (+/-) shift. Feels really stupid.
> PMDC motor, anyway
Pmsm? BTW, what is the difference between pmsm and bldc?
>delay calcs for you isn't a bad idea.
This is not needed, if you measure between phases. Because you will have same key point as hall sensors. And with halls you don't need any smarts, just simple logic.

>>1689177

doesnt work that way, my negro. if you just cascade RC filters, what happens is the next stage detunes the previous stage unless it's isolated by (say) an op-amp. that's why some smart folks had to come up with Salen-Key, Butterworlth, Woolworth, Chebyshev, etc, type filters.

Solar panels lose a lot of their power output at high temperatures. Would it be possible/feasible to transfer the excess heat from the panels and utilize it, for instance, heating utility water?

>>1689190
I thought about this. I even wanted to patent it

>>1689185
R·C -- aR·C/a -- a^2R·C/a^2

>>1689026
>No, but probing the inductor of the boost converter gives me noise that looks identical but at ~15Vpk-pk, so it's reasonable to assume that the booster is where my noise is originating from
yes. now bend that (((aesthetic))) shit flat and get the inductor's flux lines not pointing directly at your low-noise circuit

>>1689190
>it has wires, therefore it's electronics
fuck off back to >>>/diy/sqt/ and take your mechanical shit with you

>>1689204
Solar cells are inherently more electrical than mechanical you fucking dork.

>>1689190
common hybrid module

>>1689206
cooling is mechanical. now fuck off you wheedling trollnigger

>>1689202
amazing invention, anon

>>1689214
Yes, but when I realized such systems exist, I became sad.

>>1689216
the less you know about a field the easier it is to come up with great solutions that "no one has ever thought about before".

>>1689218
/diy/ the post

>>1689180

>BLDC

the "family" name. Includes PMDC, SRM, Doubly-Excited, and even some forms of Squirrel-Cage/Induction motor.

Depending on the industry you're in, will often get used interchangeably with PMDC.

>PMDC

Permanent magnet DC motor


>PMSM

a subset of PMDC designed to be driven specifically with sinusoids, not with trapezoidal control. What you're looking at doing is really general PMDC, not PMSM.

>This is not needed, if you measure between phases.

Variable delay is used on most BEMF setups with very precise speed or position management. Earlier anon talked about 60 deg. dead time between polarity switches in a single phase - only really true at a stable condition. precise control of time-varying loads and time-varying input will generally add or subtract small delays on either side of phase dead time depending on control target. that's probably beyond what you're doing, though.

>I need to measure BEMF between phases, not neutral, otherwise I have 30 degree (+/-) shift.

clarify this? You might be on the right track, I'm just trying to make sure.

>>1689190

look up Thermosiphons. what you're describing is already done on some setups, but it does take some thinking and optimization to do well. it's really not that hard to get going though, so if you're wanting to do it for your own solar rig, go nuts.

You could heat-pipe it out to a managed TEG or something, but a thermosiphon is likely gonna be easier and more efficient at the delta-T's you're looking at.

>>1689220
>\n
>quote
>\n
>\n
>sentence
>\n
>\n
>sentence
>\n
>\n
>quote
>\n
>\n
>sentence
>\n
>\n
>\n
>quote
please cease

>>1689228

I really didn't space it out like that, I don't even know how it ended up with double spaces. I thought hirosimoot axed that automatically.

>>1689218
describe /diy/ in 25 words or less

>>1689230
gookmoot is probably trying to make reddit users feel more at home

>>1689234

Kek

Need to add in cargo containers and tiny houses some how.

>>1689234
>describe /diy/
why? to PROVE myself? /ohm/ and /arduino/ are the only threads I choose to visit

>>1689240

bro, step back the autism, he's saying your first post was
>describe /diy/ in 25 words or less

>>1689242
>le autism meme

>>1689218
>>1689234

>the less you know about a field the easier it is to come up with great solutions that involve cargo container bunkers 50' below your tiny house filled with crab meat

>>1689242
oh I'm not >>1689218 got it wrong, sorry

>>1689248
I really wish we had ID's on this board at times.
Other times I'm glad we don't.

>>1689157
>soldering pin header
>hold opposite side straight with finger
AAAAAAA

>>1689220
>clarify this? You might be on the right track, I'm just trying to make sure.
I suck at explaining with words. Graphic is better than a million words.
This is partially the reason why I was confused, because everyone described shit in table, or with words. (Words in science and engineering must be banned, prove me wrong)
You know what, I think I know even better solution. Bottom pic is related. Why can't I just use optocouplers for this? All you need is to get 2-ish volt EMF and there you go. With 20V motor it is not a big deal...
>dead time between polarity switches in a single phase - only really true at a stable condition. precise control of time-varying loads and time-varying input will generally add or subtract small delays on either side of phase dead time depending on control target. that's probably beyond what you're doing, though.
I don't think it matters that much. In any case I can add delay any time I want.

>>1689272
Weak. I clean soldering iron with my fingers.

the headphone audio is great but the mic sound level is piss-weak on this $8 usb sound-card. does anyone have a clue why that could be the case (beyond the low-hanging fruit "bc it's a $8 sound-card")?

>>1689274

ah, no. that's really not going to do what you want it to do.

1. you don't want phases relative to each other, you want BEMF on phase in dead time relative to " neutral" , that's the only place where a relative zero-crossing measurement will be meaningful.
2. Other anon talked about "filtering" and "majority function." He also needed to talk about masking and reconstruction. Pic related. These are actual drive and BEMF wave-forms. it's absolutely monstrous.
Honestly, this is sort of the problem with tables and figures pulled from application notes - they're extremely idealized. I'm gonna do another post with the "Hard" topology they usually use to accomplish BEMF feedback. it's pretty straightforward - honestly even more so than what you're planning.

>>1689297
>>1689274

Post 2 - see pic related.

Here are some really good application notes on the down and dirty of this all, one in video form. -

http://ww1.microchip.com/downloads/en/appnotes/01160a.pdf
https://www.youtube.com/watch?v=unhQ2ToXZO0

>>1689297
>1. you don't want phases relative to each other, you want BEMF on phase in dead time relative to " neutral" , that's the only place where a relative zero-crossing measurement will be meaningful.
But why?
>2. Other anon talked about "filtering" and "majority function." He also needed to talk about masking and reconstruction. Pic related. These are actual drive and BEMF wave-forms. it's absolutely monstrous.
Chinese e-bike controller has some simple RC filter (and some weird bottom comparator for something). But it uses neutral point (and I'm 100% sure chinks tried to make it as cheap as possible, so there must be a reason why everyone is using neutral point, not 2 phases).
Anyway, I though of rise/fall, because there is a delay, between zero crossing (phase-neutral) and actual switch engagement. I don't like hard-coded delays (different motor load, different delay, ew), because sensored motor controller work fine without those (6 states, everything is easy, and you can get decent results even by using Arduino slow abstraction thing). This is why I though for a moment that finding zero cross of free phase is a good idea, because it should correspond to hall sensor high/low, thus it should pretty easy (at least I thought so) to emulate hall sensors.
>>1689303
This is too complex for now. I don't understand basic operation in ideal world without PWM and other noise.

>>1689318
>But why
>so there must be a reason why everyone is using neutral point

Because BEMF is offset by and relative to the "neutral" point. It doesn't zero-cross negative; it doesn't zero-cross positive. it probably won't even reach those voltages in a small motor. the neutral point is somewhere halfway in between. you can actually use +V/2 as a reference "neutral" if you don't want to do anything complicated.

If you don't want to screw with the filter, don't - but the LM339 Comparator method is very straightforward and gives you a decent hall sensor-like output with a little bit of fiddling.

Maybe this is a better video -

https://www.youtube.com/watch?v=NXkLydhRvS0

>>1689290
try to find out if it needs a particular type of microphone and make sure yours is of that type.
Try a different mic.

>>1689324
>the neutral point is somewhere halfway in between. you can actually use +V/2 as a reference "neutral" if you don't want to do anything complicated.
Virtual neutral point is easy, just 3 resistors and there is your point, you can use as ref. for BEMF.
>Because BEMF is offset by and relative to the "neutral" point. It doesn't zero-cross negative;
I think I know why I'm confused. I don't really understand how BLDC motors work in first place .

So. There is a delay between zero-crossing (neutral-phase). 30 degrees in ideal conditions (how the fuck I'm supposed to measure degrees, if motor can change its speed?). What will happen if I switch earlier than zero-crossing? Motor will go faster, or I will blow MOSFETs?

>>1689334

Now that I have a better feel for what you're asking -

>What will happen if I switch earlier than zero-crossing? Motor will go faster, or I will blow MOSFETs?

No, nothing. it's just janky. I was answering previously for larger, nastier motors, but given that you want to run an small e-bike, don't worry about it. just use zero crossing to switch to next step in the 6-step cycle. you can add a fixed delay via cap and resistor in there if you want.

It will limit your top speed and operation will be slippy, but that's probably fine.

>>1689339
But first I will try fixing BEMF on my current controller. Some chinks had installed 1 uF ceramics there, instead of nano-farad ones.

>>1689244
5/10 not only are you 6 words over, you forgot about the very important immitation property of the crab meat

>>1689272
he laughs at your pain.. and at you not using a fingernail to hold the middle of the pin header body flush to the board while tacking one or both end pins

>>1689276
I clean my fingers with my soldering iron, you vaginer

>>1689290
OS mixer and level settings, probably. the mic power switch is sometimes found here. or, possibly, it's a line in and not a mic in, can't provide power to a mic, and therefore you're fucked

>captcha: select all squares with a traffic light outside of an NXP office

>>1689334
> I think I know why I'm confused. I don't really understand how BLDC motors work in first place.
A "brushless DC" motor is a synchronous AC motor (usually 2- or 3-phase) designed to be driven via an electronic speed controller (rather than the sinusoidal AC you get from mains). You can drive them "blind" (without any sensing of the rotor position), but you get better performance if you synchronize the drive to the rotor position.

At any point in the cycle, one of the six possible combinations (of which terminal is connected to which rail) will maximise the torque/current ratio (two will give reduced torque, three will give torque in the wrong direction). You cycle through those combinations in a given order (depending upon direction of rotation). At the point you switch combinations, the efficiency of the current combination is falling while that of the next combination is increasing. You aim to switch at the crossover point to maximise performance.

> how the fuck I'm supposed to measure degrees, if motor can change its speed?
The speed doesn't change instantaneously. Particularly if the motor is driving something with non-negligible inertia. You know the speed from the rate at which zero crossings occur, and the speed isn't going to change very much in 1/6th of a rotation.

> What will happen if I switch earlier than zero-crossing? Motor will go faster, or I will blow MOSFETs?
If the timing's off, you'll get lower performance (torque, speed, efficiency) than if it's spot on.

>>1689362
>If the timing's off, you'll get lower performance (torque, speed, efficiency) than if it's spot on.
How much lower, and what is redneck way of tuning it right?

Can anybody explain this.
>E-bike controller, hall sensors
Works fine, goes up to max speed (25 km/h).
>E-bike controller, BEMF
Works fine, until it reaches walking speed, after which motor starts making weird noises and not accelerating.
Without load it reaches top speed no problem

After some disassembly, I found out that they used 1003 SMD resistor (100k), and ceramic cap. My tester identified it as 0,5 uF, which would give low-pass filter, filtering everything above 3-ish Hz, which is fucking too low, for wheel that has 30 magnets, and spins at 600 rpm (10 rps), which means frequency is roundabout 300 Hz.
I don't understand this.

Where can I find 10-30 nF caps?

>>1689204
>now bend that (((aesthetic))) shit flat
It was flat while I was taking measurements. But I'll still probably want some form of shielding as a part of the case since it will likely be folded at that angle in the end, suggestions welcome. I have some old 2-edged razor blades that feel thin enough to fold and cut into a shield of some kind.

>>1689372
sure that's a series resistor not parallel?

>>1689375
well, fug
I'd start working on shielding, maybe putting the converter across the workbench from the amplifier to see how much noise is conducted vs. radiated. I'd also try replacing the inductor on the converter with a shielded inductor of like value and sufficient current capacity (because they clearly don't give a fuck). yeah it's another 10 burgercents or whatever but it may help

>>1689385
Sure. Three 1003 SMD resistors in proximity and 3 capacitors. Middle point goes directly to MCU.
Same controller, but older, worked fine...
Just for shit and giggles soldered 103 ceramic through hole caps (which was hard, and I ripped trace, as usual).
Tested this shit, didn't workie workie. Same effect but not without load. Which means, there is another resistor somewhere else I can't find. So I need more capacity there. More than stock.

>>1689394
Got scope?

>>1689397
Sadly no.

>>1689388
>replacing the inductor on the converter
I was pondering that, but it's an SMT inductor operating at 1.2MHz. Now I've got no shortage of magnet wire, and measuring its inductance wouldn't be terribly difficult, but would a common ferrite bead work as a core? No clue what material it claims to be. The smallest actual toroid I have is about an inch in OD.

>>1689404
>>1689394
Anyway, here is a theory.
With stock capacitors, it filters out real waveform as scooter gets faster.
With current capacitors, it doesn't filter junk which confuses the micro.

Niggaz, is computer sound input fast enough for those waveforms?

Will this fry the MOSFET?
Gate-source voltage is 16 volt max. Lamp is basically a dead short when it is cold.

>>1689441
lost pic

>>1689441
read
the
datasheet

you're looking for information on maximum instantaneous current, or perhaps a maximum current / duration graph
measure the cold resistance of the lamp too

>>1689425
>SMT inductor
size 4D28 or even 1206 Ls should still be big enough for a relatively clumsy person to put on those pads with controlled hot air + flux paste + shitty tweezers + a pinky on the table for balance. hand-soldering works well too, once you've gotten the old one and its mass of solder off of the board by whatever means
the shielding, whether by iron-impregnated epoxy or solid metal, is the important part
>ferrite bead
material mix is often engineered to be lossy and you wouldn't be able to estimate saturation current very well without data. I wouldn't, except in an emergency
>measure inductance
nah, the highlighted portions of Pic related are universal for step-up converters (assume Vsat ~= 1V, Rsc and Ct are irrelevant). choose a realistic Iout(max) of your actual application, the opamp's max recommended output current + a couple of mA for the voltage dividers + 30% margin would be a good start. check against inductor's Isat from datasheet, or calculate flux density and all that lovely stuff vs. graphs in the material data sheet if winding your own
di/dt is inversely proportional to inductance. less di/dt is better because less delta dv/dt i.e. output ripple, but you still want to get enough current up there to charge the cap fully in a timely fashion while supplying the loads as they come up

>>1689446
stronk Rosskiy man does not need korrupt Western physics ideology or datasheet to chase bear on scooter. "safe" operating area is not relevant to bear on scooter

What is the maximum frequency that an ordinary diode can handle?

>>1689454
how heavy is an ordinary rock?

>>1689454
Look at the reverse recovery time.

>>1689451
>hot air
not here jeff
should I have bought that 50W Aoyue at the junk shop to dual-wield?

>>1689458
hmm, yeah, you probably should have
I have managed to get those little sons of bitches off the board with a knife tip across both pads at once. it wasn't the best party

>>1689454
slow, compared to extraordinary diodes

>>1689177
FUCK
turns out the pass filters don't work with half sines, which is what i get when i amplify the audio signal with an op amp because the op amp has only 0 - 5V output range
Is there some easy way to get negative voltage from my 5v wall wart so the op amp can output an entire sine?

>>1689466
oh? that's cute
it's easier to get a 1/2Vcc voltage

>>1689472
where do you get the negative voltage with that?
I was hoping there would be some simplay way to do it. With two batteries for example getting negative voltage is incredibly simple so i was thinking maybe using two buck convertors instead of batteries, powered by the wall wart?

>>1689472

but then it can only work with signals half as big. better idea is to get a second wall-wart to create -5V (or -12V, doesnt have to by symmetrical.)

>>1689475
aw shit again, the power supply must be floating so bucks won't work since the ground isn't isolated.
So i will have to use two wall warts instead.. not ideal, but i will work

>>1689475
can't, buck converters aren't isolating. maybe if you use two separate wall warts but I wouldn't

>>1689476
>half as big
it's for input into a micro's ADC for blinky lights, not a 28-bit audio codec. 1Vp-p should be plenty to do some FFT and decide how the lights should blinky

>>1689479
Any idea if chinks sell some magical chip than can invert voltage?
I don't need big currents like 20ma or so will do

I know about programming microcontrollers, but not much about actual electronics. If I just buy a ton of parts and try to put a little device with screen and wifi together, how likely am I to succeed within an attempt or two?

>>1689475
there's no reason for all that headroom, what you're trying to do will work just fine on a 5V single supply. the supply splitter to create a 1/2Vcc internal reference is the usual way this is done

>>1689480
why yes, they do, ICL7660
fortunately yours isn't a particularly noise-sensitive application

>>1689481
the design in your picrel, or your own work? the first question is how's your soldering? if you have any doubt (and you're asking so you probably do) I suggest picking up a few SMD soldering practice kits such as https://www.aliexpress.com/item/32812908140.html and get confident first
next question is how's your soldering iron?

>>1689482
>ICL7660
seems simple enough to use and they can do up to 20ma which is perfect, i am going to order a lot of them from ping pong and be done with it

>>1689485
own work, but similar-ish concept. I have a little bit of experience soldering, but not much. should definitely practice first.

>next question is how's your soldering iron?
don't have one right now lol, still need to get a lot of equipment

>>1689488
right. you're gonna want to work up to it and do a few smaller projects from start to finish first. I recommend looking at lots of other simple low-power microcontroller designs, so you can get a broader taste for the relevant concerns and ways to address them in a small portable design
meanwhile, the wide variety of "for arduino" breakout boards on ali can help you sort out design and code before you try to integrate it onto a board. you should also learn KiCAD or other pcb layout software, and you should definitely practice navigating manufacturer literature such as application notes, datasheets, reference designs, etc.
when the glorious day comes, order enough parts for 5 boards because you probably will mess the first couple of boards up. displays etc. can be easily removed and reused for the next copy so just a couple or three would be enough. also expect to have missed something important on your first rev of the layout. it's pretty usual

>>1689494
that seems sane, thanks. would it make sense to start out with a raspberry pi, and move down to arduino, then standalone micros, just to ease into things?

>>1689498
oh "for arduino" is really just a keyword that chinese sellers use to identify the sort of bits and pieces that might be useful to someone prototyping a system
how much do you know about programming microcontrollers? ever used an RTOS?

>>1689499
I messed around with atmel avr chips before, but never with more complicated attachments than buttons and leds. Haven't used an RTOS before, no. (But do programming in general all the time, so at least have that covered.)

>>1689466
>>1689177
The easiest way would be to filter the direct audio wave which is a sine and then after you use the opamp to boost the signal coming from the filter, which on average will be couple dozen millivolts if you use 2 stages, but that is plenty enough for the op amp to multiply
that way you need no negative voltage at all

>>1689549
>>1689177
alright i did it! i am getting a nice clean strong filtered and rectified signal out of the opamp exactly the way i need it and this signal will go into an analog pin on an arduino
but i just realized something
the arduino is powered by an isolated wall wart and the filtered signal is supplied by audio jack from some random audio device like a laptop, or desktop computer sound card.
I know i must connect the grounds between the wall wart ground and the audio jack ground, but can i really do that? the wall wart supplies DC and the audio jack supplies AC, that seems like a recipe for disaster

>>1689446
Let's assume current is OK.
But. At moment you turn on the MOSFET, you get 20+10V on gate (compares to source pin), which is above max allowed voltage on gate.
>>1689451
>stronk Rosskiy man does not need korrupt Western physics ideology or datasheet to chase bear on scooter.
MOSFETs are western lies. BJT is way to go...
No, in reality, there were only handful of transistors in soviet union, and everybody knew datasheets by heart, because there were no internet back then, and searching in books was slow.


Ok, here is an interesting question.
I found this router in recycle dumpster, and it works fine. Sure, it is old first rev DIR-300, which is like 2007 technology, but for some reason it survived.
But. Explain me this. I have 330SMD resistor (33 Ohms?) just in parallel to power jack. Why? I removed it, because it was getting hot, and I didn't had beefy power supply... What is the purpose?
Of course router kept working after removal just fine.

>>1689660
And yeah, there are no vias around that resistorb too

>>1689501
ok. well, if you've ever programmed a threaded application in whatever language, you probably get the general idea of what something like FreeRTOS can do for you. and you will probably want to learn well how to use the internal peripherals on a micro, especially interrupts and timers, which will prove invaluable for many purposes, especially handling things for you while the power-hungrier core is sleeping
if you want trainer mode, try the STM32. bluepills are $2 each

>>1689641
>mixing ac and dc is disaster
not quite totally wrong. engineers do it all the time. but in these cases there is a concern when the input goes outside the power supply rails. the ESD-protection diodes built into most ICs these days will turn on and they aren't really dimensioned to carry significant current for any significant amount of time. they also introduce odd harmonics like any clamping circuit. that's why I've been recommending ac-coupling with a cap and superimposing a dc bias on your side of the cap. but if you are willing to toss the bottom half of the signal you can just ensure there's enough input series resistance that a speaker amp's not pushing several hundred µA through the protection diodes
>plug into headphone jack of line-connected equipment
oh, then you'll want to add a cap between the common terminal of the jack and circuit ground to thwart noise and possible damage due to ground offsets

>>1689660
someone must have seen something while developing the project and popped that in there to fix it. they could be covering for a warehouse full of shitty 5V power supplies they already bought that run way over-voltage when this router is idle. or maybe those are connected to the pass-thru switch on the jack to dis/charge those caps with no spark. if you don't have a psychic friends network you'll need more context which means more tracing, scope probing, and possibly x-ray images of the board. beware, capitalist pig dog designs sometimes run by mastic and baling wire

>>1689663

allows manufacturer to claim a higher power spec. more power = longer distance.

oldtime stereo makers would add a lot of lamps in the unit to increase power spec as well. govt had to legislate watts-RMS-per channel coz capitalism is inherently crooked. today you can get 40W computer speakers running from a 4W transformer. i imagine legislation has been withdrawn, or computer stuff is exempt. and then there are 1kW rated 5.1 AV receivers. that cant be real.

>>1689663
put a blank plug into the power jack and see if the caps are still connected to the rest of the system ground

>>1689678
>making shit up
found THE Ukrainian

Hello I am working on a project for a tabletop game where there is object detection on a grid most likely using IR cameras. I want to send that location information to multiple android phones(up to 5) simultaneously and in real time. What would be the best wireless protocol for doing this?

>>1689675
>someone must have seen something while developing the project and popped that in there to fix it. they could be covering for a warehouse full of shitty 5V power supplies they already bought that run way over-voltage when this router is idle.
You know, this is possible, because those shitty routers were failing every fucking day, because of caps exploding.
> or maybe those are connected to the pass-thru switch on the jack to dis/charge those caps with no spark.
It is not connected anywhere, besides power jack. I ripped pads (too lazy for desoldering), and there were no vias, no nothing.
>>1689678
Kek.
Imagine amount of energy wasted by those things over 12 years. This is like roundabout 80 kW-h. Multiply this by amount of those sold (back in the day DIR-300 and router were synonyms, at least here).
This is why global warming is happening.
>>1689684
>put a blank plug into the power jack and see if the caps are still connected to the rest of the system ground
They are connected through another layer.

>>1689685
host a dynamic webpage on a raspi instead

>>1689685
Wi-Fi Direct seems like the correct, compatible answer

>>1689341
Huh, I didn't notice second resistor. Where can I find 1 uF caps?

>>1689721
>where can I find
read
the
OP
dumbass

>>1689686
>Imagine amount of energy wasted by those things
Kind of what you do when you quote entire passages

>>1689731
fun fact of the day:
an idle adult human dissipates 876 kWh per year

>>1689729
I'm not interested in stores, I wonder where it can be used.
I found three needful caps in LED bulb.
1 uF is not enough, still noise confuses the brain. I would add 2 uF, and call it a day.
>>1689731
?

>>1689738
Is there a way to harwest that energi?

>>1689743

yes, it's called slavery.

>>1689743

>>1689754
/pol/ the post

>>1689741
they're probably the 2nd or 3rd most popular ceramic cap value for decoupling. the problem you'll be having is getting one with an adequate voltage rating, which can't be read off the package unfortunately. at these levels of power, I would much prefer to know for sure what I'm working with
THT 1µF ceramics are not unheard of but they'll be hard to find