/diy/ - Do It Yourself

so I bought pic related on amazon for about $7 and it programs my AVRs fine, but I want to use it for normal communication between a PC and another AVR.

All I had to do was add pic related to the firmware in main.c, and call it from C with a standard usb_control_msg with requestID = 33 and the number of bytes in the appropriate arg.

The think that is weird to me is that it works perfectly for 1 to 254 bytes, but if I request 255 it times out no matter what i put as timeout in the usb msg. The code in pic related can only fetch 255, so I'll modify it a bit to get 256 if I can figure out why it maxes out at 254 bytes. For now it's fine as is, it's just bugging me that it stops at 254 and not 256.

Any USB gurus in the house?

>>1284914
>pic related

oops. the original pic is USBASP version 2.

Making a 16 point moving average filter using a microcontroller to filter out a 500hz signal using different sampling speeds. My current algorithm is:

>output = ( [Sum of previous values] - [Oldest value ] + [newest value] ) / 16

How do I work out the frequency cutoff for different sampling speeds? Is it sampling speed divided by sample amount? E.g. 8khz / 16 points = 500hz cutoff
If thats the case, when I change the sampling speed do I then change the amount of samples to reflect this? E.g. 4khz / 8 = 500hz, 16khz / 32 = 500hz, etc.

>>1284914
254 bytes payload + 1 byte length + 1 byte requestID = 256 bytes total, the longest your length value can accept.

>>1284962
See https://ptolemy.eecs.berkeley.edu/eecs20/week12/freqResponseRA.html

Can't be bothered doing the math, so just looking at just their graph the 3dB cutoff point is roughly at 0.75 radians/sample for a 16 point filter (blue line). Multiply that by your samples/second to get radians/second, divide by 2pi to get frequency.

So for 8kHz sample rate that'd give a roughly 1kHz cutoff. Scales linearly with sample rate.

>>1284964

why would the bytes being returned be affected by length and requestID? the length is one of the four bytes sent in the usb_control_message (that also includes some other stuff) that are frequently called "value" and "index", and can serve any purpose you like.

If one of those four bytes is subtracted from some 256 limit, I'd expect all 4 to be.

In the code up there, ispTransmit uses the SPI bus to get bytes from the AVR and that is all that is returned; the length byte and the requestID are not returned.

Thanks for your comment but it seems wrong to me.

>>1284968

to expand on that: data[] has 8 bytes. data[1] is the requestID, [2..5] are the four bytes that make up the words typically called value and index, and can serve any purpose you like in the usbasp code. the other 3 bytes include the length, so I'm actually sending it in two places and could access it in data[7]:data[6] (which is what I should be doing in the first place).

pic related is some of the existing code that handles larger blocks of data and shows what I believe is the integer usually called buffer_length in control messages.

>>1284962
Matlab gives pic related results.
The cut-off frequency is directly related to the sampling frequency like assumed, but the -3dB point is around 220Hz at 8kHz sampling.
The relative -3dB point moves somewhat if you change the filter length and the amount and position of the nulls change accordingly.

https://ptolemy.eecs.berkeley.edu/eecs20/week12/freqResponseRA.html

>>1284664
>what sort? post schems if you have.
i have a music thing spring drving an old tank from a marshall. cant say I'm very happy with it.
Is the tank compatible? Mine are Music thing modular mk2 the video demo of it sounded fantastic. I’m building mine with the expander module and the reverb brick so I can cross fade between brick and spring, both of the module aren’t working with the spring properly but the brick sounds pretty good. I need to do some troubleshooting as signal isn’t going to the spring although it’s returning fine, as I found out!

>i tried to make my own first, it sounded great but the noise was unacceptable.
Yeah this one is noisy as heck so far. I’ve not moved the spring around yet though so I’m hoping I can mount it somewhere away from interference and it’ll quieten down. It’s part of the reason I built them with the brick as well.
General rule is move the tank as far away from any other electronics or electric cables as possible to reduce noise, I’m guessing you know that already though. I’d imagine a smaller tank might be quieter than the 17” one I ordered. I might order one of those small blue tanks see how that sounds.

>I've been meaning to give it another go though..
You should. What designs schematics were you using? I need to look into this myself, but it might be a bit above my skills at the moment.

>>1284967
>>1284970
I'm terrible at maths so for the 16 point its:
>0.75 radians per sample
>8khz = 8000 samples per second
>8000 * 0.75 = 6000
>6000 / 2pi = 954.93hz
Which means I'll have a 955hz cutoff frequency sampling at 8khz when taking an average of 16 points. Is this correct?
Also, how do I calculate the radians per sample?

Thank you both for helping

>>1284905
Thanks OP.

>>1285014
Your math is correct, but the result is wrong, because the 16 tap filter is actually the red curve, not blue. In other words, that 950Hz would apply to a 4 tap filter.

Do you forget to turn the ventilation on sometimes?

>>1285043
>Do you forget to turn the ventilation on sometimes?
What ventilation?

>>1285044
Do you seriously solder without any ventilation? I hope you don't do it too often or with leadless solder

>>1285043
pepe nooo

>>1285052
Lead isn't the problem, rosin/flux smoke is. Smoke in general is bad for your health.

>>1285055
Don't worry he's safe now

>>1285056
Oh thanks for the precision

>>1285059

fug that was a close one

>>1285024
>Below is a plot of the magnitude of this function for L = 4 (red), 8 (green), and 16 (blue).

Longer filter should give lower frequency no?

>>1285024
If red is the 0.75 radian per sample, does that mean I take the radian per sample value at the end of the first bounce, So that Red is 0.75, Green is 1.6 and blue is 3.1? Because if thats the case then the higher the sample rate I have the smaller sample window I actually need.

Which means that setting a cutoff frequency of roughly 1khz would be:
>0.75 radians with an 8k sample rate
>1.6 radians with a 4k sample rate
>3.1 radians with a 2k sample rate
>etc
So when I've set the sample rate in an interrupt, i need to also change my sampling window to reflect this so that If im sampling at 8k, i must have a sampling window of 0.75 radians in order to eliminate a 1khz wave

>>1285068
>sampling window of 0.75 radians
Radians per sample*

>>1285043
I vastly enjoy this image

>>1285059
you monster!

>>1285083
Thank you, I did it this evening
>even in your shitposts you have to route circuits

>>1285089
Fucking kek

>>1285056
The solution is incredibly simple. Keep temperature at about 250°C (60/40 melts below 200°C) and exhale while soldering.

>>1285095
to understand the circuit you must breathe the circuit

>>1285095
But it's faster and easier at 350°C

>>1285095
>exhale while soldering.
i got into the habit of doing this just because slowly exhaling keeps the hand more steady

>>1285068
> does that mean I take the radian per sample value at the end of the first bounce
No, the blue curve, corresponding to a 4 samples long filter, has its cut-off frequency at 0.75rad. That is, its gain is about 0.7 at that frequency. Similarly, the cut-off for green (8 samples) is around 0.35 and for 16 points it's 0.17 or so. These correspond to frequencies of 950Hz, 450Hz and 220Hz, assuming 8kHz sampling.

>Which means that setting a cutoff frequency of roughly 1khz would be:
>0.75 radians with an 8k sample rate
Yes, but your last example isn't really valid. You need at least 2kHz sample rate to sample 1kHz without aliasing so your "1kHz lowpass" filter more or less passes everything.
Another thing to note is that the filter gain at its cut-off frequency is only 3dB/0.7 times down. In other words, a 1kHz lowpass filter is very bad at attenuating a 1kHz signal. If you want to eliminate 1kHz, you need to set the cut-off much lower. Or if you want to eliminate 1kHz specifically, pick a filter which has a null (zero gain) at 1kHz. 8 and 16 samples long filters do that at 8kHz sampling.

>>1285113
Exactly. Soldering as a zen exercise.

>>1285095
Eh, I use 300C and the rosin doesn't smoke all that much. Or at least that's what the slightly wonky sticker on my "temperature controlled" soldering iron says. I solder near a window and the slight wind from that takes care of the fumes. I was building an extractor fan with a gooseneck but somehow I broke the fan when wiring it up to rectified and filtered 6VAC. I also probably can't take it apart without breaking the PCB, but I'll see if I can turn it into an AC synchronous motor by simply wiring the 6VAC up to two of the four coils.

>>1285117
I think I understand this properly now. Thanks anon

>>1285117
>No, the blue curve, corresponding to a 4 samples long filter, has its cut-off frequency at 0.75rad. That is, its gain is about 0.7 at that frequency. Similarly, the cut-off for green (8 samples) is around 0.35 and for 16 points it's 0.17 or so. These correspond to frequencies of 950Hz, 450Hz and 220Hz, assuming 8kHz sampling.

magical.

>>1285123
>AC synchronous motor
Interesting experiment. Maybe you can use all four windings and a capacitor to asynchronously start it in the right direction.

>>1285052
>Do you seriously solder without any ventilation?
Yes.

I hope you don't do it too often or with leadless solder
Quite often with 60/40 usually
Mostly this
>exhale while soldering.

elder graybeards of /diy/

tell us neophytes what radioshack was like in the days of yore

>>1285147
Yeah, but I might want it to switch off after getting up to speed, which would require circuitry of some kind since it doesn't come with a centrifugal switch. The power-factor part of my electronics physics paper got cut out for some medical imaging, so I'm not too sure how good it would be to leave it running. Ensuring it gets up to speed in the first place would be tough, 50 times per second might prove too fast for a fan.

>>1285043
>ventilation
as a matter of fact, working on building a fume remover out of some fish tank filters. already built the SMD vacuum picker out of the aquarium pump

>>1285150
they had components. a whole wall of them and half an aisle, not just a little broom closet's worth. ICs, sonny! dozens of types. it wasn't bad when in a pinch. also radios and stereos and books and stuff. it still didn't compare to what I could find in Silicon Valley storefronts, pic related

>>1285205
having a well-stocked local electronics store is truly a blessing when you suddenly need a part

>>1285043
>Do you forget to turn the ventilation on sometimes?
I completely gave up on that.
I have to keep the thing way too close for it to actually work and it's always in the fucking way

>>1285155
at which point you may as well replace the brushless driver

>>1285221
it shouldn't be too in the way if smol, pic related

>>1285224
Trying to figure out how that extractor stuck to the handle of the soldering iron wouldn’t constantly be in the way?

>>1285226
looks movable
if using a chisel or conical tip, it probably helps to have it properly aligned

>>1285226
looks movable
if using a screwdriver or bevel tip, it probably helps to have it aligned properly

>>1285253
>screwdriver
Just imagining a Phillips head soldering iron. It might have its uses, same with Robertson, but a Torx head iron would definitely not be useful.

Ayy I know these types of questions aren't that welcome here, but I have this project due in a couple of weeks and I'm wondering if I'm doing it right or wrong.

I have this row decoder for my SRAM cells, but for some reason, my outputs are below zero. I was wondering if I'm doing the decoder part right. I'm following my professor's PDF, but I'm not sure if my inputs are supposed to be like this, since if A0 = 0 and A1 = 1, the inverter inside will make them 1 1 and activate two rows.

The thing is, in my professor's PDF, he had A0, A0` and A1, A1`. I'm not sure if I should be inverting outside of the decoder and have 4 inputs? I mean, doesn't a decoder have 2 inputs and 4 outputs if let's say, it's a 2x4?

Thank.

>>1285310
>Torx head iron would definitely not be useful.
It might be, for drag soldering.
>0402 spanner bits

>>1285337
What state is, say, row3 in when phi is low, a0 is high, and a1 is low?
If this is supposed to be CMOS, you need to complement your equation for the low side and place that circuit on the high side. Pic related is a 3-input AND gate. To invert the sense of any leg you can just swap the related PMOS and NMOS.
If this is supposed to be NMOS, don't you need some pullups?

>>1285310#
>Torx head iron would definitely not be useful.
It might be, for drag soldering.
>0402 spanner bits

>>1285337#
What state is, for example, row3 when phi is low, a0 is high, and a1 is low?
If this is supposed to be CMOS, you need to complement your equation for the low side and place that circuit on the high side. Pic related is a 3-input AND gate. To invert the sense of any leg you can just swap the related PMOS and NMOS.
If this is supposed to be NMOS, don't you need some pullups?

>>1285353
It's supposed to be CMOS, these are NMOSs. I was watching a video, and they had something different for 2x4. Pic related

https://www.youtube.com/watch?v=oI4LuCGul-M

I mean, it looks the same, but I have no idea. I also forgot about falstad, lemme see what I can do.

>>1285353
>>1285337

http://www.zachtronics.com/kohctpyktop-engineer-of-the-people/

I found it kind o half-assed but still very good.

>>1285357
oh god what have I done? I'm sure there's something wrong here.
Instead of turning one on, it turns all of them but one lol. I guess I'll read what's in the book. This is weird.

>>1285361
I haven't gotten to the layout part yet, I don't even know if I'll have time if I'm getting stuck on a simple decoder.

>>1285363
I mean heck, I can give up and not follow the PDF my professor gave, and just put 4 NOR gates to draw my 4 rows, but that would be weird.

>>1285364
Look at >>1285353 Pic related again. Answer the question before continuing.
Think of CMOS as two circuits, one which pulls the output high and one which pulls it low, with exactly one applying to the output in any given state. You apply DeMorgan's Theorem to translate one to the other. e.g. for a high-side of phi' AND a0 AND a1, your low side would be phi OR a0' OR a1'.
You could just duplicate Pic related four times and maybe optimize out a few of the high side transistors.

>>1285368
Sorry about that, I read it again. I understand the pull up and down networks, but for some reason, even though I follow the schematic given, none of my rows go high. Maybe I'm missing something extra? Sorry about being a dumbass. I'll read again.
Pic related is the schematic from the book, and to the right is my output.

On the bottom of the output, I guess, since this looks like a dynamic CMOS logic when a specific part is activated, you think I need to add another NMOS with phi under the two NMOS transistors taking A0 and/or A1?

>>1285372
Oh, that little bit under zero? A few hundred millivolts of undershoot won't hurt anything. Mark it zero, next frame.
Unless I'm reading that schematic way wrong, isn't phi(P) going to be active low?

>>1285372
>dynamic CMOS logic
it uses dynamic p-mos (gated) pull-up.

It's hard to tell what is going on your pic.

Try to implement something simple, like NAND gates or whatever.

>>1285149
>exhale while soldering
i inhale but i like the smell. I'll get cancer from 4chan anyway

>>1285380
>not wearing gloves and face mask when using 4chan

>>1285376
Actually the pdf says that phi_p is for precharge, and when it's low, it pulls the word lines (my rows I'm assuming) to high (vdd). And the selected gates with the A inputs, only one line will be activated and the rest get discharged. I guess I have to read the chapter in the book for this. Or just email the professor.

>>1285378
Here's my kms nand gate. At least it's not under zero.

>>1285383
>>not wearing gloves and face mask when using 4chan

even then there's the constant drinking while on 4chan thing going on which is going to kill me

But jokes aside isn't welding for 2 minutes worse than a complete life full of soldering? If it's really the flux fumes soldering is completely irrelevant compared to welding.

>>1285384
>dynamic
So the decoder won't work at all unless you feed it a clock for precharging, which in >>1285372 you are not.

>>1285384
>Here's my kms nand gate. At least it's not under zero.
Just lower clock/stimulus frequency.

http://people.ee.duke.edu/~krish/teaching/Lectures/dynamicCMOS.pdf

>>1285385
I thought I’d read somewhere (maybe forest mims) someone saying that solder fumes were pretty much harmless and if solder fumes were harmful they’d be a lot of sick/dead electrical engineers by now.
Did I just dream that?

>>1285385
hmm, point taken on drinking

>>1285392
stronger fluxes contain acids to help clean the surfaces, which is surely not good to breathe. in the age of no-clean, maybe he's absolutely right. still not wise to huff them all day

>>1285386
I was feeding phi as a pulse before and got the same results. In that output I had phi always high, I tried phi always low now, and my row results changed, because I added another NMOS before going to ground under the two NMOSes but they are still under zero for some reason. Even after adding caps at the output.

Wait, should I be using "bit" instead of "pulse" in virtuoso? (Tried, didn't work right, probably using it wrong)

>>1285387
That's nice, I'll give it a read soon.

I guess I'll work on the column decoder and sense amplifier at the moment and ask the professor about this later. Thanks of the help, good resources and info. But I'll see what happens. I will update when I get it working!

>>1285394
Falling inputs through gate capacitance of an already zeroed node with infinite impedance to ground could bring the outputs below zero. A couple hundred millivolts won't hurt anything and you can consider it as a right zero.
For proper testing I think you need to duplicate the actual conditions of application in your test inputs: present phi_p low on one half of a "cycle", then present row address signals on the other half of the "cycle".
Virtuoso is a bit above my pay grade. Looks neat though.

>>1285387

>>1285398
Just so you know it was HIS nand gate.

Building I contract for just installed a 35kW solar array, one of the owners wants me to mine bitcoins off it as that returnes more than selling back to the grid.
I got my hand on some old hardware from a friend, 8 or so modules, each use 12v 46A.
They've got a built in power supply, but rather than stepping 12v up to 240VAC, then down to 12V, I was wondering if it'd be any more efficient to just have a DC-DC buck/boost converter to take the 11/13 volts from the panels and output a stable 12v.

Only problem is, at minimum, I'd need 45 amps of this, max, 12v 400A.
Is this the point where you say "Fuck it, 90% efficiency on the inverter and 80% on the power supply's good enough"?
Already have the 35kW inverter so that's all good.

>>1285415
>mine bitcoins off it as that returnes more than selling back to the grid
you're about five years too late for that bucko
it doesn't

>>1285415
it really depends on the miners, anon. they probably don't use the 12V for anything without putting it through a regulator. you should look inside and see if that is the case. if so, they should accept a pretty wide input voltage range and ought to be fine with whatever the solar grid gives them
but hey, if it's still profitable to do it the easy way with two intermediating inverters, by all means do so

>>1285417
I have free hardware, and live in Australia meaning you get fuck all back for your energy.
Even with the shit I have, I'd get $1.32 selling back to the grid, or $1.80 mining, and that's per unit.
If I'm happy with the setup, I'd buy a couple grand worth of newer miners, with 10x the efficiency.

It's all silly, I know, but this is just for fun, really.

>>1285418
I was thinking about this, I'll check.
Could be much easier to just add a cutoff that shuts them down if the input voltage goes above or below a certain point, but I always worry about hard shutdowns, especially on fragile silicone.

It'd still be profitable with the inverter+psu combo, but it's all extra noise and shit.

>>1285427
hard shutdowns ain't no thang to silicon as long as any NV storage is protected from anything stupid by some few ms of warning and the supplies shut down in an orderly fashion

>>1285427
Switch options depending on situation:
- mining
- fast trading
- selling to the grid

(Stupid?) question, I have almost no experience with batteries:
I have two 'dead' 12V batteries from a Bosch battery drill. They're NiCD, and charge for ~1hr, then give me about 1 minute of drill time before they go empty, so I guess there's something with the cells or so.

I was wondering:
- Is there some kind of hacky trick I can use to fix this/make it less bad? Get the batteries to keep at least a bit of charge?
- If not, could I just take them apart, somehow separate the 'good' cells from the bad and make these into a general purpose 12V battery for projects, or are they all waste?

>>1285776
NiCd cells are pretty easy to botch, have you charged them or discharged them only partially, i.e. not till completely full or empty? That's what kills them from what I've heard.

>>1285776
There's probably a bad cell or two. You'd have to take the pack apart and find it by measuring the voltage across each cell. When you have, you can search the web for "nicd rejuvenation" and find several ways of resuscitating the cell, depending on what equipment and components you have on hand, or just break the pack apart and use them for general purposes. Energy density is not the greatest, but they're a lot easier to handle and work with than Li+.

Anyone know what sort of temperatures some non-electrically conductive heat-sink compound/paste/grease is good up to? The datasheet only states "Stable at normal temperatures and conditions", but it's sold alongside heat sinks for transistors and such.

I'm going to use it to replace the compound on an induction hob; it uses a diode under each element fixed with thermal paste and a small rubber gland to ensure constant contact for temperature measurement. I had to remove the glass top to replace the fan and found that the existing compound had half-hardened and had to remove it. I'm not sure whether the diode under the glass is expected to get up to the temperature of the pot above; the glass is a pretty good insulator and isn't hot after boiling water for a minute.

But then what is the diode for? Ensuring that the glass isn't getting too hot? Or am I to believe that it approximates the temperature of the pot by measuring the decreased temperature of the glass underneath? There's also a 140°C thermal switch next to the diodes. And don't ask me to trace the PCB to figure it out, there's 6 of them and they all mean business.

>>1285877
the limiting factor will be where the paste/grease turns too runny and drips everywhere, losing your thermal contact. It'l be used as a thermal sensor, if it was attached to the windings it wouldn't be for sensing the glass temperature, it'll be for protecting the coils from overheating, with the thermal fuse next to it as a last resort. Maybe has a separate sensor on the glass, or could just modulate power and expect the user to account for temperature.

What you want is thermal adhesive/epoxy/cement.

>>1285877
Applications commonly call for 150°C. You probably can't go wrong with plain old white heat sink grease.

>>1285888
I would suppose anon would want to replace it with just what was there.

>>1285888
It's not cement, the diode is sandwiched between a springy piece of rubber that pushes it into the glass to maintain contact. As far as I can see the coil itself is lacking any sort of temperature sensor. Oddly enough, there is no thermal paste to connect the thermal switch to the glass or anything else, it's suspended by the same piece of rubber as the diode but it sits with a 5mm or so gap between it and the glass, so maybe it is to sense the temperature of the coils, or at least their immediate surroundings. Both sensors are situated just above and in the centre of each coil, but there is a ring of plastic between the coil and the rubber.

Just realised that the diode is to tell the temperature of the glass so it can tell you when the element is hot to touch or not, which makes sense, though the thermal switch still doesn't make much sense.

>>1285908
Perhaps the diode protects the cook, while the switch protects the coil?

>>1285911
Well the switch isn't thermally coupled to the glass or the coil in the slightest, but I imagine the 5mm air gap conducts heat better than all the rubber, plastic and air in between the switch and the coil.

At least I'm pretty sure it's a thermal switch, It looks just like ones and I got a mysterious google suggestion when I searched for its datasheet of "thermal switch", but nothing exactly matching it.

can I build a wood frame around a transformer I made (120volts AC to 32 volts DC) and it'll be safe?

let's assume the AC hit the wood somehow (it'll be suspended in the middle of my frame so it won't) but say it touches the wood.. does AC travel through that wood well? if it was 1/4th an inch or 1/2 an inch? just to be 100% safe

I know wood is a bad conductor, just wondering if that is safe

>>1284905
I've got some 3.3v SMDs that I want to use with an 18650 battery.

Would they be fine to run at 4.2V? I tried turning a couple on with one of my batteries and there didn't seem to be any issues, my concern is if this will kill the LEDs in a few weeks instead of immediately.

It's only a 0.9v difference at its peak and the batteries will run down to 2.5v when in use. My primary concern is if it's used with the charger constantly plugged in.

>>1285949
the voltage rating on LED's is the voltage drop across them, you can run them at any voltage, they will always have 3.3V across their terminals
make sure you have a proper current limiting solution though, running them straight off the battery will kill them, which you may have already done

>>1285950
I'll be running the LEDs in parallel, would that still be fine?

This is my first time working with LEDs, and I've noticed a lot of the strips have a resistor on them. Is that all I need?
Why is current limiting required? I've never had this with any other electronics project I've attempted.

>>1285952
>>1285950
Is this what I'll need?

New 640Pcs New Electric Unit DIY Led SMD SMT 1206 3216 Chip 64 Values 1ohm~10Mohm Resistors Assortment Kit 5% Newest Hot Sale
http://s.aliexpress.com/e2EVzUjA
(from AliExpress Android)

>>1285952
If they have built-in resistors, you likely won't need an external one, but check the datasheet to find out what voltage they're meant to be run at, the resistor will provide the right amount of current only at or near that voltage.
You need current limiting because once the LED reaches the voltage it needs to actually turn on, it starts to draw current at a much higher rate as the voltage increases more, far more than it is safe for it to handle (see pic.)

>>1285953
Those will work if you're making your own boards, but I'd get a kit of through-hole resistors for prototyping, and because you'll need them for literally any other electronics project you want to do.

>>1285955
Well I doubt my LEDs would have a built in resistor, they're just generic 5050 SMDs from AliExpress. The cheapest I could find at the time I ordered them.

100Pcs 5050 Warm white SMD SMT LED PLCC-6 3-CHIPS 5050 led 60MA Mid-power 10-18LM 0.2W Free Shipping High quality Wholesale
http://s.aliexpress.com/JbMvI3m2
(from AliExpress Android)

But they did provide the information I need to calculate the right resistor.

Thanks for all your help by the way

>>1285952
>in parallel
You'll probably want a resistor for each set of LEDs (here, 1 per set, more in series for higher supply voltages). Disadvantages of resistors include that they just throw away 25% of your power and, depending on the application, you might not want the light to lose brightness as the battery runs down.
The proper (read: autistic) way to do it is a driver chip such as Pic related.

>>1285963
>>1285968
The best way to run them would be in clusters of 5 series bundles in parallel, and use a boost converter to run these sets of 5 at 18 volts or so.

>>1285917
It's 130degC thermo switch, google "KW thermo switch"
>>1285935
>can I build a wood frame around a transformer I made (120volts AC to 32 volts DC) and it'll be safe?
>I know wood is a bad conductor, just wondering if that is safe
Dry wood is good insulator, you will be safe (unless you literally piss on it or spill something conductive). I would be more concerned about fire hazard (use breaker/thermal fuse)

>>1285995
I covered the entire unit in silicone (the internal part) it's not more than 2 inches I wouldn't say of the PCB I just don't want it to be exposed really.

just trying to make 100% sure and have multiple safety levels

are you recommending a fuse inside of the device? I guess that is smart.
like where my plug in area for teh device have a 10 amp fuse or so?

>>1285995
>130degC

>>1286026
>C

>>1285996
Can you post a pic of the device?

do you guys like circuit bending?

I'm bending a really cheap synthesizer (Stylophone, pic related) right now, and I'm gonna pick up some cheap toys from Goodwill later today if I can find some good ones

>>1286071
Not actually related but
1. saved for my holding-it-wrong collection
2. how's that set of screwdrivers working out? I have one on order and want to know whether I should keep it for myself or go straight to gift wrap

>>1286080
>ow's that set of screwdrivers working out?
it's likely generic crap which is being sold in various packages (at least the bits like very similar to the set I had). Made out of recycled tin cans, not suitable for anything that costs over $50 unless you are ok with marring screws.

>>1286080
It's a cheap set I got for a few bucks at harbor freight. They're alright for small electronics maintenance, but they're not really built to last.
Trying to unscrew small, stuck screws has bent some of the more small screws, but as with most stuff from harbor freight, I can't really complain for the price

>>1286100
>more small screws
meant bits

>>1286100
>Trying to unscrew small, stuck screws has bent some of the more small screws
typical chinesium, had same experience with similar-looking bits.

>>1286107
To be fair, I was manhandling them pretty hard.

My biggest gripe is that the Phillips bits are way too... steep? Narrow? They're too thin and long to make decent contact with the walls of the screw, so I wind up using the flat head bits on some screws

Didn't realized the old thread reached the limit so here I go again.

Hi /ohm/. I recently got an electric guitar and a small amplifier. I have some technique on classical guitar but I'd like to learn to do some cool stuff with pedals and I'd like to make them myself.

Anyone here who has done it can give me some advice where to start? I'm not a scientist or a technician but I have a background knowledge on math and physics and I'm currently learning about basics of electricity and the Forrest Mims book for electronics.

>>1286156
>cool stuff with pedals
Stomptronics is such a wide field because there are countless ways to ruin the linearity of an audio amplifier.

>>1286156
I would start with a gain based stomp box. Nice and simple. Distortion or overdrive, then compression or limiting. Then move on to filters... Tone, wah, ya, EQ... Then I would try time based effects, although at that point you're going to want a programmable ic, and that's another can of worms, since the whole PIC thing will pretty much open up another dimension for you.


But start with gain based effects first, because you will need that knowledge implemented into all the other ones.

>>1286156
I can recommend the bitsbox kits if you are in the U.K?
If not then just buy a couple of the cheapest diy pedal builds you can find and practice on those. Tremolo and distortion pedals seemed to be the cheapest one I could find so maybe try those.

Trying to fix a 2ds with a broken charging port.
Is it viable to just connect the battery pins to a 3.7V power supply it needs, and do I have to bother with the 3rd pin?
If so, can I just connect a phone charger to some diodes/resistors until the voltage drops to from 5.1V to 3.7V or should I look up voltage regulators?
I only started looking into electronics recently, so any help with this would be appreciated.

>>1286254
Grab a tp4056 and stick that in for charging.

>>1286257
Problem is I have no experience with soldering, and the distance between the positive and negative pins is tiny.
That's why I wanted to do it with the battery pins, which are at least big.
And it's a used one, so the battery barely lasts an hour anyways.

>>1286254
no, because it's possible to overcharge a lipo battery
and when you do that they create fire

>>1286268
Ah, my bad I meant the power pins where the battery goes in, just connect those to power directly without a battery, avoiding charging all together?

>>1286271
oh, yeah that'd work fine
you can get a voltage regulator for literally pennies

>>1286271
>>1286275
I'd use an adjustable regulator so you can adjust it to show a fully charged battery, but since there's going to be a voltage regulator in there anyway you could probably just feed it anywhere from 3.2V to 5V. It might also have a boost converter if its circuitry can't run off 3V or so, which might be the case for the LEDs and their switching transistors and such, in which case you shouldn't give it more than 4.2V or so, though I am somewhat unfamiliar with the behaviour of boost converters near their maximum input voltage.

>>1286136
could be JIS vs. american crosspoints?

>>1286345
they'll run at their minimum duty cycle which boosts minimally, or they'll shut down and degenerate to a simple inductor

>>1286368
Shut up u stoopid faggot

>>1286368

>>1286275
>>1286345
Oh, so it would work that way.

Can I get away with just getting a resistor to lower the voltage, or is it unsafe enough that I should read up on adjustable regulator?
I just don't know if I can fit in the whole regulator circuit in the small battery hole.

>>1286372
you can get whole step-down converter modules that fit in that slot ffs. if you can't deal with the small stuff, this task, and in general this field, is not for you.

>>1286372
A resistor will not work. Two resistors making a voltage divider will work but only astoundingly terribly. But a Zener diode + resistor will suffice in a pinch. Either way you'll almost certainly have to buy the parts, for which getting a linear regulator is the way to go. If you can measure the battery voltage that would also be useful; it might not just be a single series cell.

>>1286373
holy fuck you elitist cunt. I'm not who you're replying to but jesus christ not all of us are experts. Get off your fucking high horse you dickhead.

>>1286373
>>1286374
Yeah, fair enough, I misjudged the hole size, I don't have the console on hand.
And I guess I'll stop whining and just get the linear reg.
Thanks for the help.

>>1286370
rood.

>>1286372
A resistor won't work in any meaningful way and it's important you understand why, a circuit that has a fixed operating voltage and current you can calculate the resistor suitable, as soon as the required current increases because your circuit is making a noise or flashing a light the current draw increases, current through the resistor increases and therefore the voltage dropped by the resistor increases. Droopy voltage makes the circuit not work.

A single zener diode in reverse will work as long as you exceed the minimum zener operating current and don't exceed the max current thermal rating. The diode will be burning the voltage x current as heat energy and might get warm if the power is too high!

>>1286373
Holy shit, you sour fuckwad! Go vent somewhere else.

Anyone know how to use a Shunt regulator?
I have a current transformer outputting 0.25 to 5A, I want to take that and drop it to a consistent voltage and then to 12v at 0.7A

I cant find any other way that would suffice and be robust enough to actually implement in the field.

>>1286577
>I have a current transformer outputting 0.25 to 5A,

What? Can you post a pic of the transformer

> I want to take that and drop it to a consistent voltage and then to 12v at 0.7A

>>1286592
> still tripfagging
How are we supposed to suck your cock electronically tho..?

Can one of y'all help point me in the direction of good LED strip lighting? I've spent a couple hours searching, and I am at wits end trying to find one that seems legit. I'm going to run about 40 feet in my room to replace my old halogen lighting, and I'm trying to find something brighter(~>4k lumens total), around 4000k color temperature, and 80+ CRI, in whatever voltage is available. I'm just having trouble finding anything reasonably priced that doesn't seem to be flat out bullshit. I've looked all over Amazon/Aliexpress, but so much of what I find is just too good to be true, or flat out misleading specs(Lists the same thing differently in 3 places so you're never sure which is which, shows obviously the wrong item, etc). The only ones I'm finding I'd trust to be accurate are on overpriced home reno sites wanting $200 for a 15 foot roll! Can you guys help me out?

>>1286599
I've got the 5025 non-waterproof version of this uplighting my living room. There's a special color temperature controller that complements it nicely.
https://www.aliexpress.com/item/5050-Double-Color-Temperature-LED-Strip-One-Chip-with-Warm-White-and-Cold-White-DC12V-60LED/32298725565.html

I came across that one as well, but there's no CRI listed, and I would never use the cold white, only the warm white. Thank you, though.

>>1286592
Its a design constrain given to me my another guy, not an actual transformer I own.

The design is guaranteed to have an input in the range of 0.25-5A. It must then use some system to output a constant 12 volts at at least 0.7A.
I know I could just throw a burden resistor on there and a fly back, but then the low end current range doesnt produce enough power, and I would need a stupid high watt rated resistor.
My next solution was load switching, but mosfets are not stable enough to be implemented in the field according to the manager.
So I'm at square one and kind of out of ideas lol
¯\_(ツ)_/¯

I can't really find any in depth literature on implementing shunt regulator( or even its exact use).
The other idea presented to me is to connect up a two winding voltage transformer up to the Current transformer, then operate it in the saturation range. According to him that would give a constant voltage output.
That's some mad scientist talk, and even he had no idea how to go about doing it hahaha.

Any help is really appreciated

>>1286603
I can only say that my eyes are picky about quality of light and that these look good. Try asking the seller for data.

>>1286607
The 10:1 range is the tricky part. If you can find a switching power supply controller that works well over that range, you're good to go. I've seem them rated over a 5:1 range but not 10:1.
If power factor is totally not a concern, maybe you could use a triac to chop most of the half-cycles off the front, or even skip whole half-cycles, then rectify and filter. Hope your post-regulator has good enough line rejection.
As for shunt regulators, basically you just burn excess current in a hefty power transistor/resistor or the like. The TL431 is a fine example for very small-signal applications. In your case, for a 10:1 input range, plan on burning over 10x, maybe 15x the desired load here. Add MOVs for surge suppression.

>>1286607
Your load is 12V * 0.7A = 8.4 W

at 0.25A you need at least 8.4 W / 0.25 A = 33.6 V drop on CT burden resistor or whatever load you choose. If you are going to have constant 33.6 V drop at 5A you will have to dissipate 33.6 V * 5 A - 8.4 W = 160 W.

Do you really want to use shunt-type regulator? output?

Just got my 50 74HC164 SIPO shift registers from ali, fun times ahead.

>>1286610
Why are you tripping though? You're helpful and all, I just don't understand why you view yourself as above the general board convention, especially now that you're starting to trigger the autists.

>>1286633
>>1284197

>>1286613
What about the 400 LEDs?

>>1286649
>sinks or sources 4mA
indeed, what about them

>>1286633
Because I can and because I'm bored. I don't violate diy board rules so why not?

>>1286660
>Because I can and because I'm bored. I don't violate diy board rules so why not?

So you're feeding your ego and pissing people off because "why not".

gotcha.

>>1286663
If you're so bored by this.. board, then why the fuck are you here?

>>1286660
si tacuisses..

>>1286665
>If you're so bored by this

either you don't understand that my green text was a quote from the jerk I was replying to, or you misunderstood my comment.

>>1286671
>si tacuisses..
>>1286663
>So you're feeding your ego and pissing people off because "why not".
If guns kill people, pencils misspell, spoons make people fat then you are 100% right and I'm "pissing people off".

>>1286675
>If guns kill people, pencils misspell, spoons make people fat then you are 100% right and I'm "pissing people off".

so in addition to your weak ego and being a jerk, you're dense as well.

go on, pal. tell us more about your trainwreck of a life.

>>1286673
Or, he accidentally quoted you instead of the tripcunt you dense fuck..

Jesus Christ you people are getting so butthurt over some faggot with a trip. It's not a big deal ffs.

>>1286689
> anons point out that tripfag is literally the only trip on the board, which has historically been trip free, thus free of all the drama that follows them
> its anons fault, not the tripfags
Nice to see you've finally taken it off. Keep it that way please.

>>1286689
No. Fuck you. There is literally zero reason to use one unless you're OP, and thread hijacks are not a thing on /diy/.

If this faggot wants to make a name for himself he can fuck off back to fakebook.

I'm looking into getting a new oscilloscope. One thing that I want is the logic analyzer capabilities, or at least the ability to display a lot of digital waveforms.

Looking at https://www.rigolna.com/products/digital-oscilloscopes/ the cheapest scope with digital channels is the 1074Z Plus. Anyone know which scopes I can do the usual Rigol hack on to get the logic analyzer abilities of the 1074Z Plus?

Also, anyone got recommendations for a new multimeter? My chinkshit broke and I want something higher quality.

>>1286784
https://www.google.com/search?q=1074Z+Plus+site:www.eevblog.com
>want something higher quality
quality is yellow and starts with an F

>>1284905
Slight sidestep from the usual, where can I get cheap enclosures? Preferably with a subpanel.

>>1286795
.. or the Chinese equivalent of same, 福禄克

>>1286607
Remember the crowbar? Evaluate the principle behind this simple circuit to keep the battery charged that supplies your 12V device.

>>1286795
Those posts seem to mostly be about unlocking the 1074Z plus to something better

Is something like a ZD99 going to be enough for decently delicate soldering, or should I splurge on a more expensive one?

>>1286650
4mA per LED is bright enough.
74HC164 + ULN2803 = TLC5916

>>1286898
Try it. €12.95 is not much for 48W. Splurge later if you really need temperature regulation.

>>1286908
Cheapest 45W soldering station promising "precise temperature regulation". Iron more ergonomic than ZD99 because handle closer to tip.

>>1286594
>>1286633
>>1286686
>>1286689
>>1286690
If you have 4chanX, click the wrench, go to Filters, go to Tripcode, enter this line. This will hide all of his posts from /diy/. Remove ;boards:diy to hide him from everywhere.

/!LeoPard6uk/i;boards:diy

I'm really bored so I opened up an old DVD player. Anything useful or cool I can do with this? Any risks? It says "Danger:visible and invisible laser radiation when open". Does that mean just standing near it is harmful? Am I gonna give myself cancer?

>>1286609
So if a TL431 works at small signal input, what would you suggest we look for? I cant find any examples of other people implementing a shunt regulator in the way I was thinking.
I'm trying to understand if the shunt regulator provides enough power to run the system. And how I would go about actually connecting this up to my CT.
>>1286610
Like said here, the power requirement on it is pretty high, especially since I need to double any watt ratting to prevent current spike damage.
However, If this works, I would be happy to spend as much as needed on heavy duty components.

Another idea now is Pic related.

>>1286898
>Is something like a ZD99 going to be enough for decently delicate soldering, or should I splurge on a more expensive one?
>>1286932
>Cheapest 45W soldering station promising "precise temperature regulation". Iron more ergonomic than ZD99 because handle closer to tip.

I've used similar soldering station, you will not get "precise temperature regulation" as soldering iron temperature changes a lot when internals are heated up. Soldering iron that comes with it isn't the best either, there is gap between the tip and the heater so thermal transfer is sub par.

You should get a Hakko T12 / T15 based station, even chinese ones (built with MCU's) are much better then ZD99 or suggested Stamos S-LS.

I personally use FX951 and very happy with it. I suggested going with chinese Hakko T12 / T15 soldering irons and controllers to few mates of mine and all of them are happy with the choice.

>>1287038
I think the cheapest/easiest solution is running three 12V lead-acid batteries in series, feed by suggested "crowbar"-like circuit
>>1286875
to limit charge voltage to 14.4 * 3 = 43.2 V and then running a DC/DC converter to get 12V.

Batteries could be eliminated and few "big" capacitors could be used instead.

https://www.youtube.com/watch?v=v0o1Uk3-r1A&lc=UggxRwSHptXHXXgCoAEC

These receptacle testers give false readings for reverse polarity if it's reverse polarity _bootleg ground_? Why the fuck?

What if I tried it using a 2-prong extension cord (no ground connection)?

>>1286875
So at low levels the current flows just throught the top line, but at the breakdown voltage of the zener diode it triggers the gate of the thyristor, with a pull down resistor to keep from floating the gate.
I'm unsure what the circuit would do after that.

From what I understand a crowbar is just used for over voltage to protect something likt TTL components. I'm not following how this could be implemented in my problem.
This is the first circuit I've ever built, up till know its all just been theory and math lolb, so please forgive my slow uptake.

>>1287036
That means it has a tiny laser diode inside, nothing dangerous unless you point it in your eye. Those things have some neat servos and linear drive things that could be used to make a bootleg 3D printer or jotter or other light-duty CNC machine. There's probably not much in the way of electronic components you can salvage, but conceivably the caps if they're still good, any crystals, power transistors, and the plugs and sockets and that sort of hardware. Of course the fan and any other of the big pre-assembled stuff can definitely be useful.

>>1287053
>>1287058
Hi there!

You seem to have made a bit of a mistake in your post. Luckily, the users of 4chan are always willing to help you clear this problem right up! You appear to have used a tripcode when posting, but your identity has nothing at all to do with the conversation! Whoops! You should always remember to stop using your tripcode when the thread it was used for is gone, unless another one is started! Posting with a tripcode when it isn't necessary is poor form. You should always try to post anonymously, unless your identity is absolutely vital to the post that you're making!

Now, there's no need to thank me - I'm just doing my bit to help you get used to the anonymous image-board culture!

>>1286784
I don't see any mention online of anyone who's figured out how to unlock the logic functions on the Plus models, if you really need it, you may need to buy the MSO model or the key legitimately, or find someone who knows how the original hacks were generated and try to find the proper option code to enable it.

I have a 1054Z and the digital capabilities over the analogue channels are pretty trash, there's not that many options for protocol decode and it doesn't process the whole capture data, only what's on screen, I don't have high hopes for the MSO models to be any good. For a cheap logic analyzer I'd recommend getting a PC based one.

>>1287100
While I do agree, It's not like tripfagging actively decreases the quality of conversation here, provided nobody mentions it. Just think about how much better off we'd all be if nobody started arguing with the tripfag for being a tripfag. Not quite as good as if nobody was using a tripcode in the first place, but better than arguing over such a petty thing.

>>1287001
>not hiding all trips
replace /!stuff/i with /!/

>>1287080
you are looking at diode bridge, followed by a crow bar, then there is a diode to "decouple" load and crow bar.

it's basically the same principle that's used in most motorcycle regulator/rectifiers.

https://en.wikipedia.org/wiki/Phase-fired_controllers

Basically you have either infinity ohm burden resistor or 0 ohm when thyristor is triggered, and it triggers on overvoltage condition. The cycle is repeated 100 or 120 times a second.

>>1287036

The RF tuner is useful for all the retro hipsters diy'ers out there nowadays, they will start getting difficult to get.

>>1287214
what, pray tell, are they going to tune?

>>1287216
RF

>>1287216
Tuners in VCR sometimes come with micro-power TV transmitters.

There are few VCO's that cover like from 40-60 to 600-800 mHz, some use VCO's to build a sweeper to test filters for example.

You can surely build an analog receiver if you don't like SDR's.

>>1287216

It tunes the video into channels 3 or 4 typically.

And, I'm afraid I have a bit of bad news.
You know Cat6, HDMI, DVI, Firewire, USB 3, Lightning, and all that crap?

All we ever needed was coax, like RG6.
But they took it from you; for many reasons, like (in HDMI's case) DRM reasons.

Professionals still use coax, too. But a lot of that has to do with the fact that you can use runs 10x longer.

>>1287235
This. USB is a conspiracy. I'd much rather screw in a coax and plug in a dc jack to get my serial bus and power.

>>1287238
You don't have to screw it, but thats an option you have though in case you're a pro-grade type of person.

In addition, you can get the signal and the power through one connection as your satellite dish antenna does today (with a DC bias voltage). A laptop, for example, you could connect to the network and power/charge it through one connection.

In fact, older MacBooks used to use RCA phono jacks for their power.

Any ideas what to do with a couple of LM56s, a 555 and a microcontroller?

>>1287258
Use the microcontroller to replace the 555, and make a peltier butter conditioner or something.

Fluke 115 or something else? It's not like I'm frequently around dangerous voltages or anything, but I would like a multimeter that can last me quite some time. I don't particularly need the capacitance, diode, or even resistance modes, I can measure those accurately enough with an LCR meter, but I want something reliable for the voltage and current measurement.

Are UNI-Ts or other Chinese brands long-lasting enough to consider instead? Are DC clamp-meters worth going for instead of a traditional current-shunt meter?

>>1287053
>you will not get "precise temperature regulation"
You will never get 'precise temperature regulation'. You get what you pay for.

>>1287305
What tip was used in the test? Cheap chinese soldering irons with replacable tips are much worse.

you can get "precise" regulation with metcal smartheat or alike.

>>1287240
RCA connectors are shit for high frequency.
The old mac power connector was more like a 3.5mm with a ring around it than a proper RCA, the outside is only a shield.

>>1287235
Four differential serialisers/deserialisers feeding twinax as used by most of those standards is much easier to get down to consumer price points than pumping a single coax at four times the bandwidth.

>>1284905
What's the best place to pirate ebooks?

Really want that Michael Jay Greier book.

>>1287308
>What tip
Same type that fits, don't know the number.
>metcal
I only know the old ps900 where you need to change tips for leaded and lead free soldering. Now you can have two irons and either switch or run both (60W each). Soon.. I like the pen-type handles but that's a subjective preference.

>>1287343
1st or 2nd edition?

>>1287367
Latest edition is probably what I need.

>>1287369
https://ufile.io/9q6ap

>>1284905
I have maybe a stupid question but here goes:

I made a bench power supply from an old PSU, and it has +12v and -12v. I thought I could be smart and use -12v as a ground, then combine it with the +12 to give me a net 24v source. I connected it to a 5v regulator to power a microcontroller, and let the microcontroller control an NPN bipolar transistor to blink an LED off the 24v rail. With 0-12v, eveything works fine, but with -12-+12v the LED just stays on and the microcontroller gets very hot very quickly. What am I doing wrong?

>>1287373
Wow! thanks

>>1287378
Please buy it if you found it really useful. If not then forget about it ;)

>>1287375
Post schematic

>>1287382
Oh shit, I think I just realized what I fucked up... I didn't disconnect the USBasp after programming, so it started pumping current through a channel from 0v (-12v) of the uC to the 0v (0v) of the USBasp. Now that I have disconnected it, everything works fine. Man I'm such a fucking idiot sometimes...

hi guys, suggestions please! i have a grid tied inverter on my solar system but in my dumbshit country the meters measure absolute power so i actually get charged for power i generate! the utilities company refuse to give feed in tariff or anything until the European commission force them to.
i need to monitor the main power line to the house with a clamp meter and determine power and direction. i need that to give me a pwm signal so any power that would be exported can be fed to a 220v dump load to heat my room or water tank etc. this should leave my daytime electric bill at zero and exported power at zero. i really dont want a dc dump, as voltage can vary from 100v to 450v.
i googled but nobody seems to have this set of circumstances.
i am looking for circuits, ideas, or better Google search terms.
i tried dynamic ac dump load but a just found one guy trying to get a gti to work off grid.
pic related. its getting cold here.

>>1287384
Are you using ATX power supply? If so -12V max current is like -0.5A

>>1287388
Yeah I am. I know there isn't much current available, but it's all I have available right now. I'm trying to build a makeshift desulphinator to bring back my lead-gel battery that I accidentally left connected. I read online people had success getting them back up by pulsing at ~500 Hz with a duty cycle of 0.1 at a voltage between 20v-30v. The battery now has a open-circuit voltage of 0.2v, which means it is very dead and the regular trickle-charger I have for it doesn't even turn on. I want to get it back up to a level where the trickle-charger wants to charge it again, which should be somewhere in the 10v range.

>>1287386
Two current transformers, one for measuring your solar inverter current, another for measuring appliances current, each CT loaded with a diode bridge and same burden resistor. Connect "-" of the diode bridges, the difference between "+" terminals should be you net current.

>>1287381
Okay mister Greier

>>1287393
thanks for the reply. will this show the direction of flow?

>>1287400
>thanks for the reply. will this show the direction of flow?
You will get | I_appliances | - | I_solar_inverter |.
| | stands for abs.

>>1287408
ok so that can go in opamp and op can do pwm on a 555. i remember there being some fuckery that gets it to give 0 to 100 percent. that into a triac for much cheapness

>>1287414
>https://en.wikipedia.org/wiki/Phase-fired_controllers
Triacs aren't controlled by pwm. What are your power levels?

>>1287390
Whoop whoop! I think it's working! After about 15 minutes the voltage is already back at ~1v. Let's hope this thing doesn't blow up in my face before it hits 10v.

>>1287416
the inverter max is 5kw but i wont have more than 1.5 at the moment. if possible I would like to have a dump circuit capable of 5kw plus a safety margin to future proof it.

>>1287418
and use a control voltage and diac to vary instead of pwm?

As if trying to learn electronics isn’t hard enough without having to use fucking breadboards.
Fuck my big sausage fingered hands

>>1287510
Tweezers?
I never had a breadboard, I like dotboard.

Currently learning some discrete logic. I managed to create this little sim of a serial flasher for a traffic barrier. I'm planning to breadboard this with:
- 74F374 shift register
- HEF4093 quad NAND gate
- CD4027 dual JK flipflop
- 74HC14 hex NOT gate

In the beginning, the JK flip-flop is set to "1" and the shift register gets filled sequentially with "1" values.
When the register is full, the JK flip-flop is reset to "0" and shift register bits are reset sequentially with "0" values.

Real serial flashers that are used in traffic control systems also have the option to reverse the flashing direction. Maybe to use one bit signal to choose left -> right or right -> left movement. Any ideas how to implement that with logic ICs?

>>1287171
This would work if we were working with a normal input range.
Im working off a current transformer that goes up to 500kv at 1kA
An open circuit, even for a second would arc and destroy the whole thing.

The only way we can see is hooking a normal two winding transformer to the CT and operating it at saturation.

The problem really is that I cannot find a single example of another person doing this.

I believe I'm going to have to get an iron toroid and some speaker wire and make a transformer and just test its qualities.
Idk how I'm going to recreat the conditions I see on the line.

I need to start from somewhere.
Degree in CS, worked on FPGAs, I now need to learn how to get educated on electronics and circuits. I downloaded NI Multisim, what circuits should I make? Is there a book that I can follow? I'd like to avoid the super simple books that teach you soldering and KVL/KCL and the rest of the simple theory.
Any help appreciated.

>>1287519
There are bi-directional shift registers like 74299. They use bunch of logic to route the flip-flop inputs according to the chosen mode.

>breadboard
74F374 does not play well with your other chips.

>>1287419
Triacs are aither on or off, there is nothing in between. How would that work? If you connect 5 kW load and inventer supplies 1.5 kW you will drain extra 3.5 kW from mains. You need some kind of PWM-able device and PI/PID controller.

>>1287528
>An open circuit, even for a second would arc and destroy the whole thing.
>Im working off a current transformer that goes up to 500kv at 1kA

There wont be true "open circuit" as there is always some load (batteries / caps / or a triac in ON state)

I'm looking for cheap laser printer for pcbs any recommendations?

>>1287562
For the toner transfer method? They're all about the same. You'll likely need to calibrate the long axis of the paper in any case.

>starting to get more into circuit bending
>sucessfully bent my Stylophone
>read that people bend children's toys a lot
>alright let's go get some kids' toys at goodwill
>go to goodwill
>buy a few cheap electronic toys from goodwill
>dismantle them at home
>start shorting random leads
>nothing doing
>all of the ICs are covered in a black blob
>turns out almost all modern toys use SMD circuits and cover the IC, pins and all, in epoxy, making them well nigh unbendable

well fuck :(

>>1287626
so all live managed to get done is have some LeapFrog thot yell "Hi! Let's write letters and num- Hi! Let's write letters and num- Hi! Let's write..." at me for 45 minutes straight

So, this is a picture of the board. This might be a rather daft question, but what would be the best way to make an educated guess as to where the timing resistor is on this thing, assuming it has one and it's not just handled by the IC

>>1287626
That's not an IC under the epoxy, it's closer to a raw silicon die, and it's the most common method for making cheap PCBs with few ICs on them.

>>1287653
sorry, I actually read about that and I've been using the terms (incorrectly) interchangeably

>>1287650
old toys are bendable. new ones are literally just microcontrollers, which tend to have on-chip calibrated RC oscillators these days instead of cheap off-chip RC. I fear there's not much to bend anymore, but C1/R8/C4 or C7/R6/R7 might have potential, depending on what's on the other side.

>>1287650
I'm guessing R6 and R7, since they're near some capacitors and don't look to be involved with I/O. You could also try messing about with various digital inputs to see if there are any invalid or unexpected outputs you can give it, but that probably won't give you much.

>>1287662
C1/C8 made visible sparks when I shorted them with other contacts earlier

>>1287664
messing with R6 and R7 seems to crash the thing and I need to do a hard reset.
Or it causes it to loop the voice

I'll experiment more with these and post back

>>1287519
Get an FPGA.

>>1287536
Why are you becoming an engineer in reverse?

>>1287668
Don't short the caps to other pins, put other caps in parallel with them, same with the resistors. Putting a 1k in parallel with the 33k/333 might have interesting results. If you have a multimeter go probing the board for any voltages that aren't digital, since analogue voltages are pretty much the only signals you'll have any luck bending, save maybe putting single-transistor inverters here and there.

>>1287664
Here's the other side of the board, just an led array and some power supply stuff.
The leads directly above c1/c8 are capacitors

>>1287679
You could always try grabbing a bunch of digital outputs and feeding them into a DAC to see what happens.

>>1287679
prehaps you should take this as an opportunity to realise that circuit bending is pointless and gay?
don't get me wrong i'm heavy into reverse engineering but:
1) making toy keyboard make a different noise is worthless
2) why would a computer use an rc network as a clock, it wouldn't, they are for things like toy pianos.
3)if you want to get into anything built after about 1999 then go heavy or go home you will need appropriate tools to fuck with uc.

the led matrix should have been a giveaway that this wasn't going to be easy.
if you want to prove to yourself one way or another about the ability to mess with this thing then simply draw out a schematic of the entire board and work out exactly what everything does.
you can't expect to mess with things unless you understand how they work.
if you want to waste 5 to 7 minutes of your life email the manufacturer and ask for a service manual.

oddly enough, stroking R7 with one of my multimeter probes causes the pitch to warble

>>1287687
I'm doing it because I make electronic music, I was originally gonna stick to electric pianos and such but I read about people bending small toys too and it was too good to resist

>>1287687
this anon gets it

>>1287690
then add a resistor of approximately equal value from R7 to somewhere else and see what happens

>>1287672

Probably got tired of needing a minimum of a Raspberry Pi to flash a LED.

>>1287386

Crude simple solution. A clamp on current transformer on your service entrance. Whatever sort of ucontroller circuit you want to build your regulator - an arduino would be fine. Your microcontroller sends a simple analog and on/off signal to a vfd powering a motor driving a pump circulating water within your heat dump thermal reservoir. Use a flow restrictor on the outlet of the pump to keep the head high. All the energy you put into the flow is going to ultimately turn into heat. You can bleed from the high pressure outlet ahead of the restrictor to drive heat exchangers elsewhere.

The advantage of this solution is it requires no custom high power electronics. You could pick up a vfd and a pump/motor combination super cheap from surplus/ebay. Use a couple old water heater tanks as the thermal reservoir. The 'lost energy' from the dump is going to be from the vfd itself, and from the body of the motor (~90% efficient).

This is basically a mechanical engineering solution to your electronics regulation problem, but would work fine imho.

>>1287733

oh yeah, stick the pump,vfd and (well insulated) tanks in your basement. That way any 'lost' heat just goes into the space you are heating. great way to hit 100% efficiency.

>>1287536

You should start by helping
>>1287519

design his barrier flasher (whatever that is, sounds like a UK thing targeted towards motorized rollinghams) by using no semiconductors except for transistors and/or diodes.

>>1287733
>>1287733
>A clamp on current transformer on your service entrance
What kind of current transformer are you referring to?

Generic current transformer will output AC and you will have no way to detect current flow w/o AC voltage reference. I.e. if you are pumping 5 A into the mains or burning 5 A you'll get 5 A output.

>>1287536
Build something you are interested in. Audio amplifier, transceiver, VFD/ESC for e-bike, 10-50 kW charger for E-car, device to aid with executing fish, 100 W WiFi amplifier, tesla coil, etc

>>1287386

Reading the thread further, the power you want to dump is pretty limited. Why not add use an incandescent light dimmer (rated to several kW) powering a heater. You will need to modify the heater such the fan is driven by full 110 (or 220).

Just turn the knob of the light dimmer with a servo driven by your arduino / uC. (maybe a ss potentiometer upgrade in the future). The dump frequency response of this is going to be low (response time ~.5sec). That would be plenty to handle shifting environmental conditions. Add some hysteresis to the servo output to minimize wear on the pot.

The input would be the same current transformer at your service entrance. Code a super simple controller trying to keep net output current at <= 0.

This would be a super cheap adaptive power dump. You might even have enough stuff sitting around to prototype this. Almost zero power electronics work needed.

>>1287744

take the raw output of the current transformer, and put it across some sort of shunt resistor to read the transformer's secondary (output) current. Sample that with the arduino to capture the AC current waveform. You have plenty of horsepower for that.

I don't understand your question, unless you are trying to handle power factor.

this guy has a dc output even:
https://www.mcmaster.com/#6583k26/=1ah2qgn

but a normal guy should work with some sort of shunt resistor on the output.https://www.automationdirect.com/adc/Shopping/Catalog/Sensors_-z-_Encoders/Current_-a-_Voltage_Sensors_(AC_-a-_DC)/Current_Transformers/Split_Core_AC_Current_Transformers/CTF-0200-5-SA

might have to do a little op-amp signal conditioning to keep the arduino's input happy.

>>1287744
>>1287768

oh, fuck, sorry - retardation on my part. You will have to sample the voltage too. I see your point.

you can do that crudely with a massive voltage divider straight across the mains. Think mega-ohm resistors. You would have to sample that too.

>>1287771

this would give you instantaneous power flow into /out of the house - basically handling power factor shifts too.

>>1287536
How adept are you in the frequency domain? If good, probably start with semiconductors. The comprehensive reading is still good, just skip over the super basics.

>>1287771
anon's meter reads net production as consumption and wants to divert net produced energy somewhere else instead of paying for it.
I'm a bit surprised the inverter doesn't offer this sort of information through a handy data port.

>>1287386
Mine crypto with it?

>>1287386
• Are the inverter mosfets connected directly to the service line?
• Are you sure they're not fining you for the crappy square wave back-emi you're sending into the grid?
• How many watts are we talking about here?
• Assuming 220v RMS, the actual voltage is going to be (peak to peak) something like 600v so I wouldn't worry about the 450v.

It seems to me, if you can already get your daytime electric bill down to zero[1], the problem is already mostly solved by
the inverter itself -- has the capability to switch-in more power from the service line when needed.

And, if it has such sophisticated capabilities, then we (or you) would need a schematic of it. There is undoubtedly some tap points on a PCB which would give us almost exactly what we'd want.

Also, I would think a pump motor running continuously would wear out, as well as having other inherent problems.

---
[1] Yes, I'm ignoring the exported power problem for now.

>>1287760
>Why not add use an incandescent light dimmer (rated to several kW) powering a heater.
Dimmers use phase-fired controllers. Current will vary a lot during one half-way cycle and op will pay for that no matter what.

>>1287780

That's actually a good idea! Plus a lot of crypto mining also inherently generates heat for the winter too. Win-win-win.

Posted about a Wien bridge design a few weeks ago, finally got around to making it!

Attached pics of board and lissajous figures made with it.

>>1287794
cheers, just wish we could see it better

>>1287799
Close up on the perfboard.

>>1287788
>[1] Yes, I'm ignoring the exported power problem for now.
That's the problem that anon is trying to solve!

>>1287800
>>1287794

>>1287794
noice.

>>1287789

It's going to play with the power factor for sure, but currents won't ever be uncrontrolled/crazy due to the restive load.

It's only a couple kW too. That's nothing. Sure something exotic could be created, but payback would never happen.

>>1287903
Looks like you don't understand the problem.

If you use dimmer and phase firing controllers you will be NEVER able to zero generated / drained from power.

Because you have fixed load and that load is either on or off. Catch my drift?

>>1285056
He said leadless is bad, not leaded. Leadless has worse fumes.

>>1287912

Perhaps.

There is a question of what to do with 1.5kW of power. I don't understand what your trouble is with a 99% solution.

If he can push his 'zero' current down to a couple amps, that's fantastic. Designing some sort of active pfc load dump is not easy, and even then you don't know what all the other devices inside the house are doing, so any sort of parasitic dump is going to be limited on performance. It still sounds worthwhile to do.

I've got these Chinese generic 10w LEDs (pic related) and they're pretty good, except they seem incapable of doing 10w.

At 9.3v it's pulling 0.18A, at 12v it's pulling 0.7A, but at 12V it gets so hot it burns itself unless attached to a heatsink.

If I run it at between 9-10v, do I still need a current limiter?
It seems fine, it doesn't even get warm after a couple of minutes.

They are rated for 9-12V but the listing advises a resistor if used at 12v...

Apologies if this is a stupid question.

>>1287912
>>1287932

oh yeah, and to hell with true zero anyway. stick with a slight positive load, so at least you are getting something for your money from the utility. Whatever the error is, bias by that.

>>1287934
>>1287932
Use two current transformers to get "net" power
Get a 2-5kW variac and connect a heater to its output.
Hook up current transformer output (difference) to a DC-motor servo that drives the variac.
Ooops, we got a solution.

>>1287519
>Currently learning some discrete logic
I don't see how this can be done with your 374 D-FF construct (maybe it can) but if you use a standard shift register like the 74HC164 there should be a way to change the direction by chaining a second register and wiring the lamps (e.g. LEDs) _between_ the outputs of the first and the second register. Like Q0A - Q7B, Q1A - Q6B .. Q7A - Q0B. First all outputs of register B are L and one after another lamp comes up until Q7A goes H. Now register B begins to fill and one after another lamp goes off (starting with Q7A - Q0B because both are now H) until all outputs (A and B) are H and all lamps are off. Then you clear both registers and the next cycle begins. I'm hoping this is complicated enough..

The obviously much simpler solution is either the universal shift register (74HC299) mentioned earlier or: 4 bit up/down counter (74HC4516) -> 4-to-16 decoder (74HC4514/15) + some control circuitry = up to 16 lamps, back and forth.

Also, consider cleaning up your mixed bag of logic families. All your functions are available as 74HCxx, e.g. 74HC374/574 (574 only has a different pinout).

>>1287933
All these LED's modules MUST be mounted to a heatsink.

>>1287800
berry gud

>>1287933
always always use a current limiter. the tempco of Vf is negative and thermal runaway is bad. for something big like that you probably want a switching driver.
also consider wattage an approximation. pump the specified current through and let the diodes inside worry about voltage.

>>1287939

smacks of rube goldberg - not a terrible thing though. Twisting the rotor of a 5kW variac under control is going to be a lot more difficult than dinking a pot around.

I can't imagine a panel layout that is going to give you the right place to stick that 'appliance' current xformer.

The two transformers assumes you have a

>>1287950
>Twisting the rotor of a 5kW variac under control is going to be a lot more difficult than dinking a pot around.

Get a DC servo or wiper motor or side window motor out of a car.

With variac you can control net power, with a dimmer - you can't.

>>1287794
Congrats, well done. How was the lissajous music?

>>1287969
>With variac you can control net power, with a dimmer - you can't.

For the purposes of OP's problem, I claim that your are full of shit, good sir. And your mum is a whore.

Regardless of what current measurement technique OP's utility meter is doing, you can create a current waveform through the dump device controlled with scr/triac devices that is going to read effectively zero current by the utility meter. Yes, I know the power factor is going to be fucked, but the meter is not going to be smart enough to know that.

Do you dispute this? (2nd paragraph anyway)

>>1287944
That wasn't part of my question. I would have mounted it regardless.

>>1287947
So I still have to use the limiter if my voltage is close to the activation voltage of the LEDs and it's not drawing enough current to have noticeable heat output?
It says forward current 900mA but my LED was only drawing 180-190mA.
Is there any particular reason? Safety? It seems to be regulating itself just fine in my opinion but I'd still like to know why.
http://s.aliexpress.com/AZJnqARV

>>1288253
Also I was using an mt3608 boost converter to step up from a 5v power source.

Could the boost converter be regulating the current? It does appear to have resistors on board.

>>1284905

holy shit look at that fat kid

>>1288254
> Could the boost converter be regulating the current?
Yes, just as any adjustable voltage source. You're going to loose Vfb/(Vfb+Vf) efficiency, so extra 5% of the energy will just get wasted in your case.
That's why current sources have very low Vfb (e.g., for my favourite step-down LED2000 it's 0.1V)

>>1288253
>It says forward current 900mA but my LED was only drawing 180-190mA.
>Is there any particular reason? Safety? It seems to be regulating itself just fine in my opinion but I'd still like to know why.
The reason is that high-power LEDs (featuring especially LED 'laser' diodes) don't have stable voltage-to-current characteristic.
Temperature changes, ageing etc. can impact it. And suddenly the diode at seems-to-be-safe voltage will source too much current and burn.

>>1288253
When it's drawing its full 10W (approx), there's no way you wouldn't be noticing heat.
The temperature coefficient of Vf of an LED is about -2mV/°C. For each degree Celsius of die temperature increase, the forward voltage for a particular forward current drops 2mV. Lower Vf means Pic related's curves shift left, and current increases when voltage is held steady, thus power increases, thus die temperature increases, and so on until it burns.
Notice also the slope of Pic related curves, that small voltage changes effect large current changes. Thus you also need to figure the tempcos of the regulator circuit into this. It's just not a reliable way of driving an LED, notwithstanding that chinkshit does this embarrassingly often.

>>1288254
It could if it weren't connected as a voltage regulator.

>>1288301
>>1288322
I'm getting conflicting messages here, so the resistance of the boost converter would or wouldn't do the job?

>>1288322
>>1288306
So ideally I'd need resistors like pic related?
The link for the LED recommends a 2k ohm resistor when running at 12v.

>>1288329
>resistance of the boost converter
What resistance?
>i see resistors on the board
No, they're not part of the power path. Not even close to the right size for that duty.
If anything is regulating the output, it's the current limit of the boost converter you're using and/or the saturation point of the inductor on the module.
>conflicting messages
A boost converter, generally speaking, COULD be configured to regulate the current. That board you're using is not configured to regulate the output current, only the switch current.

https://www.youtube.com/watch?time_continue=10&v=Ol3mV8Vqg6A

I made this to learn how to use KiCad. It turned out pretty well and OSH Park is awesome. It's crazy how easy it is to do this stuff now and days.

>>1288345
bretty gud :D
should've used a switcher instead of the LM340 though :D

How do I choose a transformer?
I need something like a 5v in to 30V out transformer that is rated to 5A, and It has to saturate at like 0.25A. The VA rating has to be 5 times whatever the input voltage is, so a low input voltage is best. The output rating needs to be at least X large where X*saturation point= at least 7Watts. I said 30 because at a 0.25 saturation point the 30V constant output would give a power of 30*0.25=7.5W

If its not possible to buy, how hard would it be to build that?

>>1288254
>Could the boost converter be regulating the current?
Yes. The MT3608 has a reference voltage of 588..600..612 mV. Typical value is 600mV. If you want about 900mA you would need a low side current sensing resistor of 2/3 Ω. If you use 3x 2.2Ω in parallel = 0.733Ω you get 802mA at 588mV, 818mA at 600mV and 835mA at 612mV. You will not notice the difference to 900mA and your LED will live longer. The correct voltage across the LED group will be set automatically.

The procedure is this: Supply the MT3608 module with a voltage lower than 9V (e.g. from a 5V 3A adaptor) and set the output voltage to 9V. Switch off. Connect the (+) terminal of the LED to VOUT+, the (-) terminal of the LED to the current sensing resistor and the other side of the resistor to VOUT-. Switch on and check that the LED comes up. Now additionally connect the (-) terminal of the LED to the 222 resistor (red arrow points to where). The LED should get brighter and the current is now regulated.

Note that the MT3608 is a boost converter which means VIN must always be lower than VOUT. You cannot supply it from 12V other than using a step-down (buck) converter set to an output voltage lower than 9V.

>>1288369
You talk to a transformer vendor about your design, since it is highly likely you'll need a custom item anyway. Transformer vendors are somewhat set up for that, so not a real biggie. You might even be able to skip the middleman and get a safety-certified current transformer that saturates where you need it to saturate.

>>1288369
>rated to 5A (.8A sec)
>saturates at 0.25A
i'm sorry? are you trying to make a mag amp?
the formula for primary turn count is Vacinmax/(4.44*max flux density*min frequency*min area). the secondary is just your voltage ratio. wire gauge is based on allowable losses and is more of a heat transfer question than an electrical one if you just don't want anything to melt. losses are optimized when they're equal on each coil so you can pick gauges mathematically that way. transformers don't saturate due to current when loaded, they saturate due to voltage. unloaded transformers are mostly just inductors so can saturate the same way but i don't remember that formula.

it's easy to build transformers. you can get kapton tape and magnet wire (not particularly cheap) online. not sure about laminated cores. if you have to wind 1000 turns for your primary then give up now.

>>1288108

>For the purposes of OP's problem, I claim that your are full of shit, good sir. And your mum is a whore.

wrong

>Regardless of what current measurement technique OP's utility meter is doing, you can create a current waveform through the dump device controlled with scr/triac devices that is going to read effectively zero current by the utility meter. Yes, I know the power factor is going to be fucked, but the meter is not going to be smart enough to know that.

You are absolutely wrong again. The thing is you have no idea what power factor is and how the meter is working. Not to mention you are dumb enough to understand how would current waveform look with the use of SCR/Triac controller dump device (even if the dump device is pure resistive).

>>1288264
Don't believe everything you think. This is an age old morph template.

>>1288322
>>1288322
>The temperature coefficient of Vf of an LED is about -2mV/°C. For each degree Celsius of die temperature increase, the forward voltage for a particular forward current drops 2mV.

It's -1mV/°C ... - 10mV/°C and depends on LED's

>>1288377
Thanks a ton
Who would you suggest going with?

>>1288379
We are trying to take in a current between 0.25A and 5A out from a CT. Having the transformer saturate gives us a constant power output. In theory this should work. However going from theory to practice is a bit harder haha

>>1288384
Thanks for the clarification.

>>1288387
Triad Magnetics is a well-reputed firm that's been doing custom magnetics forever. They might also have insight into your larger problem of how to rob power from a high-voltage mains.

>>1288387
>We are trying to take in a current between 0.25A and 5A out from a CT. Having the transformer saturate gives us a constant power output.

You will not get constant power output. You will get clipped AC voltage (clipped sine wave).

Transformers do not saturate due to excessive current (excessive current actually decreases magnetic flux), they saturate when due to excessive voltage (which produces excessive magnetic flux density).

Looks like you need a knowledgeable mentor or you are deemed to fail.

>>1288387
>In theory this should work
The image is unreadable. I would really like to see your theory.

Anyone here own a HIOKI multimeter? They're Japanese and look pretty good.

Does flux make soldering 'easier'?

As in does it make it heat up faster or spread heat better or is it just for 'better' solder joints?

>>1288431
YES. It won't noticeably make the solder heat up quicker, but it will make the solder flow together really damn nicely and make nice shiny clean beads on each pin/component leg, and leave you with clean, strong solder joints. It's a must have.

>>1288438
ty famalam.

Should I find solder without flux or is it ok to use flux with flux solder?

>>1288442
Well, by flux is a "must have" I meant either the flux inside solder wire and external flux. I find just using flux-core solder is fine most of the time, but sometimes I need a little extra, for which I use a tub of jelly-like rosin that I can apply to the board with a brush or spare component lead, but I've heard good things about flux pens. If you want to use no-clean flux, which those pens apparently use, then you probably will want flux-free solder, if only so that you keep the no-clean property.

>>1288406
So it acts a lot more like an MOV? The current would just be sized dependent on the N1/N2 transformer equation.

I am pretty confused about what brings a transformer into saturation. The datasheets led me to believe it was at the point when the input met the saturation current.
If that is not it, what?
My mentor barely knows what to do and shot down every idea besides this. THIS is the way he wants it done, no other way.
I really need some help lol

>>1288442

What >>1288444 said, also if you evaporate all the flux due to taking too long, the solder won't stick to anything

>>1288444
many different fluxes are available in flux pens, from no-clean to really strong shit. RMA only needs to be cleaned when it's undesirable to deal with the sticky residue. anything stronger than that should probably be cleaned anyway. different formulations of flux are generally compatible.

>>1288448
>My mentor barely knows what to do
>THIS is the way he wants it done, no other way.
i've had the fortune of always working under reasonably competent people but if this is your job then i'd start looking for a new one.

>>1288516
Maybe there's something he wants you to learn Anon. Ever seen karate kid?

>>1288521
>My mentor barely knows what to do
couple this with the knowledge that they've tried to get someone with minimal knowledge of electromagnetics to do something unconventional (and still hazy) with a transformer

>>1288523
That is obvious, and the mentor is definitely overshooting, but there's gotta be something the mentor wants confused Anon to learn. I'm sure he has learned something from this, and that could be the only goal for this exercise.

>>1288526
I've learned a lot. Thats the goal.
This isn't a job, its my capstone senior project. Its why I'm asking for help from more knowlegeable anons. This is my first step into the real world of design and I need as much help as I can get.
I have a sneaking suspicion that the project sponsor knows exactly what he is doing and is just playing dumb to let us discov3r it ourselves.
It's still and extremely strange way to use a transformer lol

He mentioned something about a 100+ year old book written by Westinghouse that used this.

>>1288448
>So it acts a lot more like an MOV?
Kind of.
>The current would just be sized dependent on the N1/N2 transformer equation.
What?

>I am pretty confused about what brings a transformer into saturation. The datasheets led me to believe it was at the point when the input met the saturation current.
>If that is not it, what?
Volt-second product brings transformer core to saturation.

I love my circuits, and I love soldering.... But I just got a few Arduinos, and the simplicity is winning me over. I used to talk shit about tarduino fags but, holy shit! I can input 10 lines of code and skip breadboarding, 5-10 calculations, fishing around for specific components, and all I have to worry about is a proper power supply circuit. Please help me... I love my circuits...

>>1288539
>>The current would just be sized dependent on the N1/N2 transformer equation.
I mean like with a 5:30 transformer a 6A input would output 1A just like normal. So under saturation the current input output ratio doesn't change.

What specific datasheet type parameter does that equate to to achieve something like I need?

I was assuming it was the current saturation parameter.

From what the sponsor said in leading us toward this it sounds like he thinks transformer will automatically saturate when connected to the CT.
The 5v input was decided because he said that dictates the power rating of the transformer. That times the max current input is our VA rating.
The 30V was chosen becasue it would output a max 30V when saturated.

Fun trick of the day: digitally controlling any four-terminal adjustable voltage regulator by current injection.
MCP4725 is a rail-to-rail voltage DAC and MCP6001 is a rail-to-rail CMOS op amp, both selected for size and low cost. Iout connects to the feedback node of the regulator along with the usual resistive voltage divider.
The upper right portion of the schematic is a voltage controlled current source referenced to Vcc. The gave is driven to yield Iout = (Vcc - Vdac) / R1, or 50µA/V. Iout replaces some part of the current flowing through R4. The output voltage is therefore Vfb + ((Vfb / R5) - Iout)(R4). With the 100µA Ifb in this example and the Vfb of 1.25V, the input has (3.3V - 1.25V - Vds(on)) ~= 2V of useful width, and so only the uppermost 2/3.3V = 2482 codes are usable, yielding an output range of 1.25-12.5V and an output resolution of 4mV.
Three-terminal regulators might be usable with this scheme as well, if the current source is inverted to the lower rail and bias current is accounted for.

>>1288538
If nothing else, magnetic regulation seems to be a good way to keep the regulator robust against strong electric fields, lightning strikes, and other hostile events, and to prevent destruction of the attached equipment.

>>1288545
Ditch the Q1, connect R5 lower terminal to op-amp output, increase op-amp gain

Now you get better regulation & less parts.

Anyway nice idea. Not really useful imho.

>>1288543
>I mean like with a 5:30 transformer a 6A input would output 1A just like normal. So under saturation the current input output ratio doesn't change.

Under saturation input voltage and output voltage are 5:30 (disregarding tranformer regulation).

>What specific datasheet type parameter does that equate to to achieve something like I need?

I don't get the question.

>From what the sponsor said in leading us toward this it sounds like he thinks transformer will automatically saturate when connected to the CT.

It depends, if the transformer load is high enough it wont saturate.

>The 5v input was decided because he said that dictates the power rating of the transformer.
>That times the max current input is our VA rating.

Saturated transformer will have higher losses due to increased losses in magnetic core.

Having 5 V saturation voltage and 0.25 A minimum CT output will give you less than 1.25 W of usable power. Is that enough for you?

>>1288553
Seriously though, why the fuck are you tripping? Why do you insist on standing above the rest of us you egotistical fuck? I'm not even one of the other faggots who have been banging the drum, but I am starting to see their point.

>>1288551
You don't even need that opamp. You can connect the DAC to the regulator's voltage divider via a resistor.
You can use the same scheme with various regulators so that you can have anything from a programmable PSU to a constant current LED driver. Actually, I saw a while ago a similar circuit used in a projector for controlling the lamp (LED) brightness.

>>1288575
Tripfagging is the cheapest way to get attention and "online reputation" here. End of story.

New to electronics here.

I’m going to build a voltage regulator. Quick question, here Is some description

>In the schematic below, you can see that the LM78L05ACZ is the voltage regulator surrounded by two filter capacitors. These are simple power supply reservoirs that fill in the gaps when power supply’s are shaky or a circuit demands a lot of current. An additional capacitor of about 0.01-0.1uF is often put next to the 10uF capacitor. These capacitors should be positioned in close proximity to power supply.

Could someone pls explain this bit as it’s not explained as to what the smaller cap is doing

>An additional capacitor of about 0.01-0.1uF is often put next to the 10uF capacitor. These capacitors should be positioned in close

>>1288610
Here’s a pic of the page too

>>1288610
It's doing the same as the bigger one, but for higher frequencies. This is due to the real-world limitations of the capacitors.

>>1288617
Could you maybe explain a bit more in detail?

So basically the caps take the input power and fill up so they can supply a steady voltage and are hi and low pass filters?

If so why are there specifically 3 for that circuit? Are they working together to form 1 filter or are they working separately to filter out a set of certain frequency’s?

>>1288628
7805 does not really require any capacitors. All the capacitors are added to improve performance or make the circuit handle bad input better.
The largish input capacitor is there to handle the situation where the load suddenly draws lots of current. If the supply feeding the regulator has high resistance, inductance or whatever, the regulator's input voltage could drop below the minimum required voltage (6 to 7 volts), causing the output voltage to drop as well. C1 prevents the input from sagging too much so that the output remains stable. It is also needed when the regulator is fed from rectified (but unfiltered) AC.
C2 is added, because the regulator's regulation speed is limited: the regulator needs some time to react to the changing load and the faster the load changes, the larger the transient error will be. The capacitor supplies current during this time, reducing the voltage sag. Practical 10µF (electrolytic) capacitors have relatively high impedance at higher frequencies, so an additional small capacitor can be used to improve the situation at frequencies where the larger capacitor is less effective.

Even though all the capacitors are ultimately there to reduce the output voltage sag during sudden load changes, you should note that C2 is there to handle the regulator's imperfections and C1 is added to handle the input voltage problems. If the 7805 was perfect, C2 would be unnecessary while C1 wouldn't.

>>1288658
Thank you anon. This is exactly what I was hoping for. Thank you for your detailed explaination

>>1288380

I wrote the sim. There are conditions under which the rheostat based power dump does drop net current under conditions of solar generation. That is what you said could not happen. That said, it is much more sensitive to parameters than I expected.

the base home load is 1.54kW
at time 0.5 sec we start solar generating at solar_low (1.1kWkW)
at 1sec we move to solar high (2.2kW)
and at 1.5 sec we start the power controller.

This one calculated instantaneous power in and out of the house and attempts to minimize that.

There is a 'perfect solution' that does not involve a rheostat and needs only a single current transformer using a fuzzy/self-learning controller.

OTHER DUDES:
where is the preferred place I can post the sim code? ~250 lines c++. someplace comfy & anon.

the (controlled) actual power into the house - sliding average over 5 cycles iirc.

>>1288738

this technique drops utility meter measured rms current from 3 to ~2A. Yeah the net power is converging to zero - but not the utility measure current. Improved yes.

>>1288746

I do see your point though. I cannot escape the immense faggotry and expense of your concept's controlled multi-kw variac.

I also don't see a clean way to instrument a panel with the 2 current transformers as described above.

To solve the variable dump: use heaters in power of 2 wattages (something close is all that matters) I chose
62.5W at 220V
125W at 220V
250W etc
500W
1kW
2kW

as op wanted, you can extend the top end as far as you want incrementally. The 62W probably isn't that necessary, but my fuzzy controller sucks ass, and having the increased command resolution helped it.

to measure the house power, you make only a single rms current measurement - no voltage, with a clamp on current sensor. The fuzzy controller runs a series of experiments to figure out how to minimize the measured rms current. It increases the load, and if that helps, keeps increasing the load. If it increases the load and the house's rms current goes up, it starts dropping the resistave load.

Nothing fancy for the heater controls either - simple zero crossing SSR's for each.

this is super low cost, super simple, and requires no HV access (in the panel).

this technique drops the houses' measured rms current to 0.3A, in spite of the shitty controller (easy to fix - I lost interest)

>>1288759

here is a bit of what the fuzzy controller is doing.

Note, I'm probably using the term wrong, but who gives a shit, the concept seems to work.

>>1287536
US navy has a course you can find around called NEETS, use that if you don't know the basics.

>>1288553
..yeah, thanks and all, but I'm not much interested in what you have to say when you so clearly have a motive behind being so "helpful", especially considering how much we bore you. Polite filter and sage.

>>1288551
>>1288594
>tfw design reviews
Thanks anons. I swiped the concept from the LM10010. The beautiful straight line that came out at the end was just too sexy to pass up.

>>1288628
>>1288658
Fortunately we have 10µF ceramic caps these days. Now, just to complicate things more because I'm a little bitch like that, ceramic caps have a series inductance that depends on the physical geometry of the capacitor, so occasionally, in hundreds-of-MHz circuits, caps of different case sizes are seen in parallel.

>>1288733
pastebin?

>>1288761

code:
https://pastebin.com/QwHfBV2C

>>1288913
thanks anon.

NEW BREAD

>>1288971
>>1288971
>>1288971

NEW BREAD

>>1285043
Low T cucks don't know how to blow the smoke off as you solder for optimal precision. Not to mention your hands are most stable when breathing out / holding your breath.

>>1287308
Jesus Christ is that what they're teaching you guys now? Its a fucking pointy tip, flat tip, or round tip.

T. Avionics technician

>>1288997
Does t. mean troll or tard?

>>1289028
terkkuja, "yours truly" in Finnish
lurk more

>>1288913
>caps of different case sizes are seen in parallel.
Same basic idea.
Some people suggest paralleling different size ceramics and some don't. Those who don't claim that in addition to the wanted lower impedance at higher frequencies, you'll also get parasitic parallel resonances at lower frequencies.
I don't know who's correct.

>>1289031
To be fair, that would require lurking on some other board.

>>1289031
It's terveisin, not tekkuja.