/diy/ - Do It Yourself

What's a good cheao SDR? Is it only the RTL? I would love to get my hands on a HackRF, they're so sexy.

reposting for the new thread

>>1459494
I know my basics as far as other circuit elements are concerned, I've just never had to design power supply shit before.
I spend practically all my free time exploring the electronics articles on Wikipedia though

>>1459576
>https://www.aliexpress.com/item/benis-converter-XDD/32885025475.html
oh that's adorable, I'm going to buy a bunch of these.
Now, most of the circuits I make are going to be for synthesizers. I don't want to just eat through batteries, is there a similar product for converting wall AC current to a usable DC current? I'm sure there are, I just don't know what they're called
and speaking of, where would I go to learn about making wall power supplies lol

>>1459950
there are linear and switching AC-DC converters. linear supplies are just a transformer, diode bridge, capacitor, and single-chip linear regulator. switching supplies ("SMPS") are smaller and are used almost everywhere now, but they're far more complicated. you should avoid fucking with them unless you want to make one for its own sake, in which case there's more to cover than I could fit in a post.

you can buy one in the form of a wall wart (1), an industrial supply (2), or a board-mount module (3). the modules are great but they're like $7 minimum. your psu is an smps.

>>1459950
Wall power adapters are to cheap to diy desu. They almost always are switching power supplies with a small transformer to galvanically split the system from wall power. They use a optocoupler system to feedback over the galvanic split. Its complicated...

The old school way is a big fat transformer and rectefier/regulator ic. Thats way easier tho

Anyone doing any guitar pedals etc...?

At the moment Im building a preamp to go with my diy power amp. The preamp has a different than usual order. Anyone interested in a write up of the signal path?

Pic related is the distortion section im working on at the moment.

>>1459967
>They use a optocoupler system to feedback over the galvanic split. Its complicated...

typical example from

http://www.micro-bridge.com/data/CRpowtech/PT4201E.pdf

more info here

http://file.yizimg.com/332467/2010080308450435.pdf

>>1459977
Nice dubbs

With complicated i meant without the 'fancy' chip. I did a project at school building one without ic's, only transistors. But yeah, using a chip in the circuit is easier indeed

>>1459950
returning to the idea of starting with a voltage spec and designing according to it, the usual "standard" for analog synths is a split ±15V supply, with ±12V less common. this is important if you're interested in incorporating others' module designs into your finished system. linear supplies are presumed "cleaner" for audio applications but I'm not sure how much of that is observable and how much of that is audiophile snobbery
dual-supply wall warts are kinda rare, you would probably have to use a brick or (more likely) mount something in the rack. surplus sellers like Halted or All Electronics or Electronics Goldmine are your best bet if you want to try to find a linear for a reasonable price, or try something like this if you want to try a switcher and aren't drawing a hell of a lot of current from it
https://www.aliexpress.com/item/Original-MEAN-WELL-benis/32776801988.html

>>1459967
>way easier
*if* you can source an appropriate transformer, and if you have a place to put all that heat. I will grant you it's easier to source a xfmr for a linear supply than try to find one off-the-shelf for an off-line switcher

>>1459975
the dynamic pullup is kind of interesting. have you had a listen to it yet?

>>1459988

nice dubs yourself.

actually I was agreeing with the complexity part. lol. You're in a different league altogether.

>>1459992
>Original-MEAN-WELL-benis
>benis

>>1459999
welgome to /ohm/ :DDD
gud digids too :DD

>>1459992
>the dynamic pullup
is usually called a current source

>>1460027
if the FET gate were connected directly to the source, I wouldn't question it. the diode has potential to make its behavior a little more interesting, but I'd have to sim it to be sure and I'm kinda lazy

>>1460030
Just keep your signal below ±0.1V and you don't have to worry about diodes. Use a MOSFET as a 2-way analogue switch, JFETs are for nerds with funky power rails.

>>1459841
>https://www.robotshop.com/media/files/pdf/hyperion-g5-50c-3s-1100mah-lipo-battery-User-Guide.pdf

Very good litterature until

> Dispose of the battery in the trash
.
That's right
-
unlike NiMH and
NiCd batteries, LiPos are not hazardous to the environment. They can be
thrown in the garbage with no problem.

WTF

is there any problem with ditching IC sockets and using female machined header pins instead?
does anyone else do this?
it seems to work okay, just fits a bit tight. it's much easier to just break them to the correct length for whatever size IC you have than to keep all different sizes of sockets around.

>>1460073
>using female machined header pins instead?

yes use those they are wonderful

>>1460076
cool, glad to see I'm not crazy

>>1460052
presumably it's the cobalt that makes Li+ batteries environmentally hazardous

>>1460073
ISTR they're superior to cantilever sockets for long-term reliability. it helps to use a tool to insert ICs and also to use a jig of some sort to keep them parallel and properly spaced
you can actually get machine-pin DIP sockets that include the crossmembers for proper spacing

>>1460076
soundtrack: https://www.youtube.com/watch?v=byDiILrNbM4
preferred soundtrack: listening to that guy whine like a little bitch

>>1460079
I just bend the pins straight on my desk, can't be assed to buy or build a jig for it when I don't even use DIP packages often

>>1460081
tbf I just push them against the other side of the socket and hope they're already straight enough to go in cleanly

>>1460076
is this brazil?

>>1460086

something like that.

Is it possible to have an RCA to HDMI convertor with no/minimal lag?

I'm working on a wearable project where the camera output is RCA but the best screen I can use is HDMI. From what I understand the lag becomes a problem when upscaling the image (resolution differences) but what if the camera is capable of outputting video the same or higher resolution than the screen?

Or am I going to be better off in the latency department modifying "personal theater" video glasses which accept RCA signals?

>>1460172
I just noticed that the camera can output AHD signals too. What is the latency like for converters like:
https://www.amazon.com/MakeTheOne-Converter-Adapter-Security-Monitor/dp/B075454YKR/
?

Working on a MIDI to CV module.
tl:dr: converter for digital keyboard/controller interface to digital and analoge outputs to control analog synth gear.

cv = control voltage. gate is just a digital output that goes high to "enable" an oscillator, one per oscillator.

first a built a mono with velocity, pitchbend and modwheel from a project on github.
then i set about adding additional DACs and started programming to get poly.
it also has trigger and midi clock, but i haven't tested it yet.

the original program has an array of all 88 keys, active keys = true. inactive = false.
reprogrammed it to select the lowest three and distribute them to DACs and gate outputs. took me a while because I am a novice programmer.
got 3 poly now. note priority and such needs some work.
gonna make it quad and also add aftertouch CV out.

looks like shit, but this is just the prototype. planning on making some PCBs once i feel i got all i want out of it.

>>1460073
I almost exclusively use them.

pro:
>never run out of one size.
>more room underneath the IC for wiring/decoupling caps

>no indent to mark orienation
not really an issue.
most pcb's have IC orientation on the silkscreen, if they dont use a sharpie or something.

also very useful for other stuff. sockets for all sorts of components. transistors and caps to experiment with guitar pedals etc.

>>1460188
That converter can't do AHD, just SD composite. Expect ~2-3 frames lag.

You can get AHD converters too https://www.amazon.com/HDMI-Converter-Full-1080p-Video/dp/B07DHCN8W1/ref=sr_1_3?ie=UTF8&qid=1536386843&sr=8-3&keywords=ahd+hdmi since it's not deinterlaing expect it to be about a frame faster than SD converters.

You can also get really low latency SD scalers intended for retro vidya use, well under a frame lag, doesn't do fancy deinterlacing though so effectively lower vertical resolution and adds jitter, http://www.retrotink.com/

>>1460221
Thanks for the info. So if I use one of the retro SD scalers will the camera image be scaled down or cut off?

Is there an AHD convertor that has only HDMI output? I only need to pipe the camera view to a screen.

And most importantly is 1-2 or 3 frames noticeable? The project is night vision goggles so you can probably imagine I need to get as close to real time as possible.

How do you power a project with multiple components that each use a different input voltage from just 1 battery?

>>1460237
You pick a battery voltage that can be used by one of the components (typically a motor or heater, something that uses the most power and doesn't really care if the voltage fluctuates by ±20%) and have multiple switched-mode power supplies connected together (like a computer power supply with its +12V, -12V, +5V, 0V rails) and run the small-signal/precision stuff off those. What's the project in mind?

>>1460238
Pic related. I imagine some sort of power regulator will be required for each component along with a big enough battery.

>>1460240
Buck-boost converters are more complicated than just bucks or just boosts, so I'd avoid a 12V battery since it will likely vary on either side of that significantly. ~6V Lead-acid, 1-2S or 4+S lithium ion, by the sounds of the current you'll be drawing a single Li-ion or 3.7V array of parallel Li-ions with two boost converters should be sufficient. How many Wh/J Do you need? How do you plan on charging the thing?

>>1460241
How long do you think a typical RC model's LiPo (~7.4V, 5000mAh) can power it for?

>>1460245
12*0.12 ≈ 1.5W
5*0.1 = 0.5W
P ≈ 2.5W
E = Q*V = 5*7.4 = 37Wh
t = E/P = 37/2.5 = 14.8 hours

>>1460241
>12V
the cam probably has its own step-down internally and doesn't use the 12V for anything
also if security cam type it will probably be fine with a volt or two of drop below nominal 12V, if not more
>12V battery will vary on either side
when Pb-acid batteries start dropping below 2.1V per cell, it's time to retire them

>>1460240
I am guessing you'd do just fine with a 3S 18650 setup and a single step-down board, but you should test the camera with a strong 9V supply or check the spec sheet just to make sure
this may ease your development significantly
https://www.amazon.com/ThinkPow-power-bank-18650-battery/dp/B01N6A7IAD
if the camera is sensitive to low input voltage, there are also 4S battery boxes, which makes headroom for an extra step-down converter for the camera 12V
https://www.amazon.com/Lithium-Battery-Adapter-Output-Switch/dp/B07H24RMYJ

>>1460248
>and doesn't use the 12V for anything
It might for the lens focus motors, particularly if it's stepping it up to a few hundred volts for a piezo motor, but those probably don't need a definite 12V anyhow and will work fine off a varying voltage. Though personally I'd just take it apart and wire my different rails into it directly, unless it already has a high-voltage boost converter.

And I think deep-cycles can be run below 2V. (how about calling them Pb-H or Pb-H+ batteries)

If I go with multiple boost convertors will something like this work?

>>1460252
>>1460248
The camera module is actually just the circuit board like this: https://www.aliexpress.com/item/AHD-5MP-4MP-1-2-8-STARVIS-IMX335-CMOS-image-sensor-FH8538-CCTV-camera-module-PCB/32894400322.html

There will be no motor-adjustable lens since I'm hoping to make it wearable.

>>1460246
14.8 hours is pretty good.

>>1460256
Just use one 5V converter not two, anon. Looks like the PCB has its own switching converter or two, nothing that can be done about that. A braver man than I could trace the +12V in to find any traces of it being used for anything else, I guess it's possible that the CMOS array itself (or something else) runs on it because FETs are like that I guess. At the very least there isn't a second large capacitor (similar in size to the 16V cap on the back) for 5V, though the ceramic next to the inductor is sizeable in its own right.

>>1460258
The AHD to HDMI converter looks like it takes 5V 1A DC:
https://www.amazon.com/HDMI-Converter-Full-1080p-Video/dp/B07DHCN8W1/

I'm sure that's more amps than the display needs. To use it with just 1 converter, I guess I should connect the display and AHD/HDMI unit in parallel?

>>1460260
Oh that is a fair bit of power, how wasteful. Might get a little toasty at 5W for something strapped to your noggin. Yes just put them in parallel. If you know what you're doing you can probably find a single converter with both 12V and 5V rails, which would save space since they share a fair few components. Making your own PSU like that is also an option, but likely not one you're willing to take if you were unsure about whether to just put the 5V devices in parallel.

Looking for a 5V CMOS sensor board is also an option, perhaps adafruit/sparkfun/etc. projects use one since it's more convenient for that sort of thing.

>>1460263
I'm intentionally trying to make a digital NVD that does not require an external illuminator so I don't think changing out the sensor module is an option. I'm looking into other options for the AHD-HDMI converter.

>>1460264
So what's so unique/special about the linked sensor?

>>1460073
>it seems to work okay, just fits a bit tight.
Header sockets specifically meant to be used as IC sockets exist, too. There's also a variant where the individual pins fit to the PCB holes, making the socket almost unnoticeable.

>>1460265
It's the latest generation sensor I could find that is high definition enough to be somewhat useful. The sensor is used in a class of security cameras called "starlight cameras" since they are extremely sensitive to infrared in low light conditions like at night. A lot of them don't require much external illumination to function, just the infrared from the moon and the stars are good enough.

>>1460267
If you can find a datasheet for that sensor (and for the other ICs) then you can check if any of them require a steady 12V, but a few minutes of looking can't net me anything. Either test it on like 9V or just run it off another switched-mode power supply, last one is a safe method that will barely be less efficient, aside from the space taken up.

>>1460268
I've looked all over for the datasheets. It's a huge pain in the ass to even find out if the sensor exists and if a better model is out there since the Sony StarVis series not meant to be a typical DIY/maker or consumer product.
The AHD/HDMI unit seems to be the same way. If anything, the fact those are intended to be used in security systems may be why the manufacturers want to make everything a black box where all you have to do is plug things in and "it just werks!".
I think a reason why the AHD/HDMI converter needs such high amperage is because it needs to transmit data over long distances.

>>1460269
Ah, so modding it may be an option. Good luck in any case.

>>1460270
I know a lot of experimentation will be needed. Maybe I can get away with lower voltage or amperage since I will only need to transmit data a few inches, not several hundred yards. Worst comes to worst I can mount the AHD/HDMI converter and batteries onto a helmet.

>>1460272
Even crazier, what happens when multiple sensors get added to the goggles like pic related in an extreme case? At that point I don't know if buying separate AHD/HDMI units for each sensor is feasible anymore and I'm pretty sure I will need to adapt a raspberry pi to place images next to each other.

Would a raspberry pi even be capable of receiving and processing AHD video in near real time?

>>1460273
FPGA?

>>1460272
>amperage
Stopped reading right there

>>1460222
with one of the retro scalers you'll still get the full size image, but only half the effective vertical resolution at a time.

The AHD converters without the AHD out aren't doing anything fancy, it's just the in and out wired together.

>>1460273
>
Would a raspberry pi even be capable of receiving and processing AHD video in near real time?

Not really. You'd want an FPGA + pretty fast ADCs (AHD is an analogue signal remember)

>>1460298
What about something like this?

https://www.amazon.com/dp/B07G4FX4G5

Raspberry pi can recieve data through USB ports, and this thing appears to use much less power than >>1460260 since it takes power from USB.

>>1460192
I've a PCB for a midi to cv converter on the way.
It's this model
https://www.ericasynths.lv/shop/diy/diy-kits-1/diy-midi-cv-ii/

I'm building a MFOS soundlab ultimate and plan to build the MIDI to cv converter into the enclosure with it

>>1460273
at that point you'd be looking into a bespoke design, and a pretty advanced one at that. there are ICs that will digitize AHD and spit out RGB or YCbCr bit streams, which you would then reframe and combine into the desired output image and format, using a high-end system-on-chip or FPGA
otoh the lag from a USB digitizer would probably cause sprained ankles

>>1460484
you gave up? too bad m8
>pick the lowest three
the usual algorithm is to maintain an array of voice outputs and use the key up/down *events* as commands to allocate (on a least-recently-used basis) or free the voices
if you play a chord and lift the lowest key, your other voices aren't disturbed while they're still playing. with the least-recently-used allocation strategy, you preserve as much of the tails of released notes as possible

So why is it a bad idea to put vias on pads?

>>1460562
once liquid, solder will wick down the via and possibly not leave enough solder on the surface where you put it
that effect can be prevented by plugging the vias, but that tends to be expensive

>>1460542
Can you please post an example of such a digitizing chip?

>>1460602
http://www.analog.com/en/parametricsearch/11354#/p1848=Analog%20HD

>>1460604
Holy shit that’s a lot of pins. I take it you can’t just connect the AHD signal wire to one of the pins and it will magically work?

>>1460475
USB 3 can easily supply 1A, up to 2A even. It's just as expensive and probably will take a bit of software shit on the Pi side.

I got this rack off of some old machinery.
It has a switch (that seem to short the outputs), a fine adjustable 7turn pot, and 2 selector switches on the front. The connector on the back has 2 outputs, and 2 unknown voltage comes in that i marked with a battery on the schematic.

What is this?
Is it some voltage selector?

>>1460569
Wouldnt the surface tension of the component lead on the pad suck the same amount of solder in?
Also, cant i just feed solder in until the via is full?

I think ill go through with the via on pad design and see how much trouble it is for myself. Im guessing it is only a problem in automated soldering or reflow.

Figured I'd post this:

I have a Raspberry Pi V3, and a Samsung 32GB EVO Select card. I was compiling a program and the RasPi would freeze. Added 'dtparam=sd_overclock=100' to the /boot/config and it's been working.

>>1460668
old machines did their thing in weird ways. now it's whatever you want it to be *shrug*

>>1460682
>Im guessing it is only a problem in automated soldering or reflow.
correct, where "feeding more in" isn't really an option
for hand soldering QFN, I put a 5-6mm via under the die paddle area and use a screwdriver tip and wire solder from the bottom to solder them to the board, after I've used air and paste or whatever to get the leadless "leads" in place topside
curious, what are you soldering?

Hey goys, sorry but I have another retarded question:

With a DC power supply like pic related, could you charge batteries with this? Like could I hook it up to an 18V drill battery and set it to 1A or whatever and let her rip as long as I keep an eye on the battery voltage?

>>1460778
no, unless you want your batteries to explode

>>1460778

>Bepis
>another retarded question

Literally nobody is surprised.

But, yes, you can charge any cell that uses a CC/CV profile (I can't actually think of one that doesn't, off the top of my head). Be wary, however, of doing this regularly to packs that do not have built-in balancing and require it; lithium in general, it's a good idea with NiMH.

>>1460783
NiMHs use a negative delta-V charge termination algorithm, with current constrained by temperature.

>>1460783
Hmm, ok. So much left to learn about electronics.

So with an 18V lithium pack where you have lots of cells in there, you might end up fucking them up because it could overcharge one cell while the other is still low but a multimeter reading is giving you a voltage that looks like it’s fully charged, right?

But if you had a single 18650, set the power supply to 4.0V, whatever amperage is safe, and keep an eye on it with a meter, it would work?

>>1460782
Gotta start somewhere mang. Used to be so afraid of electronics and I think lots of boomers would be like “Just replace the whole thing!” but it’s not as complicated as expected.

Multimeter, power supply, scope, and some soldering skills and it seems like you can pretty much do anything on the hardware side. Except cell phones. Fuck cell phones. The parts are too small and my fingers are too big. I have taken apart like 4 old iPhones and have yet to get one put back together successfully.

>>1460475
RPi doesn't have USB 3.0

>>1460778
Lithium battery yes.

NiCd/NiMH you can do a slow charge.

>>1460796
So NiMH or NiCd, set the amps real low? Or set the voltage at or just above whatever it’s rated at?

I might need to start exploding some batteries around here.

>>1460788
How about you just fuck off and google "how to make a battery charger" you faggot. You shit up threads all over this board, fucked if you're going to do it here.

>>1460810
Thx for the help bby.

I don’t want a battery charger, I want a power supply because it would be fun to play with. Curious if electricity works that way, and if not, why?

>>1460802
Simple way is set current real low like 0.1/0.05C for NiCd/NiMH respectively, run for ~15/24 hours respectively. They can tolerate that much current overcharging continuously so you don't need to auto detect charging stop.

If you know they've been fully charged can do fast to start then slow down, like ~0.5C for 2 hours, then slow to 0.05C.

The proper way to quickly charge NiCd/NiMH is to watch for voltage dropping a little, or temperature rising.

>>1460833
So if you only need one output and not a crazy amount of voltage or current, are these $70-$100 power supplies anything more than a paperweight? It’s not like I need +/-0.00001V accuracy.

>>1460837
They're fine if what you're looking for is a bench power supply. If battery charging is your main need, there are better options.

Drill batteries will have multiple cells in series, requiring balancing. Charging them one at a time could work, but it's likely more trouble than it's worth.

>>1460848
drill batteries probably already balance
in any case, balancing boards are available for reasonable prices

What is the detection distance on these generic chink touch sensors? they say you can put them in walls but don't specify max detection distance

>>1460842
I want a toy to play with and send too much power into stuff.

I was watching a video some guy posted reviewing an $80 linear one and since I have no clue about electronics, I have a question-

He’s powering up an LED and if he turns the voltage up too high, the amperage go down automatically or the amps stay the same but the voltage won’t rise anymore. I figured you could just keep feeding it more volts or more amps until the thing burns out. So if I have a 9W LED and I feed 3V-3A, that’s basically the same as feeding 9V-1A or 1V-9A? And then if I have it set at 3V but crank it to 4A, the machine will limit the amperage to 3A. Is this a function of how LEDs work or is this a function of the machine?

As mentioned, I have no clue about electrons but it seems like I learn a lot faster if I have the stuff in front of me and can experiment.

>>1460848
Other anon mentioned that and after a quick google, I understand it because not all of the cells are going to be exactly the same.

>>1460854
>TTP223
you'll have to try it. if you need to, you can make C1 smaller for more sensitivity (datasheet recommends up to 50pF, 0pF is acceptable). search for datasheet

>>1460859
>C1 smaller
not really there is no way i would be able to solder something so small

>>1460861
again, 0pF is acceptable. heat it up and lose it on your soldering sponge or whatever. or get some tweezers and git gud with surface mount

>>1460856
The voltage and current you get will be dependent on the load, the knobs are limits. An LED is a roughly constant voltage load, so you'd usually power it with constant current, so can have the voltage knob way up, set the current you want it to draw with the current knob.

>>1460866
Ahh I see, I’m sure I was misunderstanding something in this goofy guy’s Amazon review video. There was a “Constant Current” and “Constant Voltage” light on the machine, and it seemed like as he tried to turn each of the knob up higher, those lights would eventually come on and stop it from rising any higher and wouldn’t put any more power into the LED.

I’m probably just retarded. For <$100, they sure do look fun to play with.

Switching or Linear? Supposedly linear is better because it puts out less noise, but that probably wouldn’t matter at first as I just screw around with it trying to explode batteries. I wouldn’t want to regret not spending the extra $20 for linear in the future though.

>tfw finally found a cool local electronics place and tried to make it there by 2pm yesterday before the closed, but the girl wanted lunch and crazy cat lady neighbor needed help moving furniture

>>1460686
I have power fets i want to surface mount.
And to handle more current i want to use 2 sides and vias to get to the other side. Like pic related, in the middle.

merchant skills of chinks never cease to amaze me
>1 sensor module costs $1
>1 lot of 5 sensor modules costs $1.3 / piece after a 5% discount

>>1460893
if you're hand soldering via-in-pad should be alright, but I'd just put them directly to all easily accessible sides of the power pads. a couple of millimeters won't kill you, the copper might get warm but it should heat-sink itself readily to whatever it's connected to
not also that some board houses might bitch if your vias are open on one side but masked on the other

>>1460902
Heating wont be a problem the fets are water cooled here. I just dont want to have any little voltage drop, so i need a little more area of copper.

I dunno where to ask this so I'm asking here:
I need a sector->USB adapter to power my raspberry pi 3 B+
On their website they say it needs 2.5A but I wanna get 3A to be safe
Also having 3+ usb ports on the adapter would be good so I can charge phones or whatever
I've found some on amazon but they max out at 2.4A per USB port so it's trash

Can't I get/make a non meme usb adapter? I dunno where to start if I need to make one or if a usb adapter is even the right thing I need

>>1460909
You need a wall adapter.
Or get USB PD adapters, they are a bit harder to find for spec you want but they'll give you enough power to burn down your house.

And I'm pretty sure you don't need 3A for raspberry pi, 2.5A should be enough as I doubt it uses even that much current.

>>1460914
Ok so a dedicated wall adapter would get me what I want. Would a standard 2.4A phone charger work too? Even when running the rpi at 100%?

>>1460925
Find out.

Can we put "Read The Fucking Datasheet" in the OP?

>>1461001
Seconding this suggestion

>>1460925
>Would a standard 2.4A phone charger work too
Depends a lot on charger. Some fast charge chargers will provide only impulses of 2.4A while others provide constant 2.4A. If you rasp pi has constant resets then you'll find out that your charger is shit.

>>1461001
If people can't read datasheets, I'm pretty sure they they can't read op as well.

Here is a little silly traffic light controller I put together to learn some basics on STM8 microcontroller programming. The sequence begins with the press of a button.
The switch is de-bounced with an RC filter and a CD4049 NAND gate with schmitt-trigger inputs. The lights are controlled by 74HC164 shift register and CD40175 (quad)/74HC74 (dual) D-flip-flops.

>>1461063
Here's the rough idea of how the components are connected.

>>1461068
Why with all the logic gates?

>>1460795
Then what about something like this?

https://www.pine64.org/?product=rock64-media-board-computer

is there anything wrong with attaching wires to a pcb with just a screw and nut as shown? terminal blocks are fuckhuge and i don't want to pay for them.

current is 10A, board is HASL (also board is chinese trash).

>>1461161

This is fairly common, though usually there's a soldered-on insert that acts as the nut to make life easier.

No idea what kind of current-carrying capacity you can expect, but I'd think you could get plenty more than 10A out of it.

>>1461167
>soldered-on insert
i actually looked for those on digikey and couldn't find them.

>>1461063
Why not use a 595 so you can ditch the D flip-flops?

>>1461194
or use a uln200x and ditch all of the logic, and then do debouncing in software too.

>>1461196
Well it's not like he doesn't have enough pins to power 5 transistors in the first place, he's literally already using 4 of them.

>>1461044
>>1460925
>If you rasp pi has constant resets then you'll find out that your charger is shit.
plot twist: all the people trying to run Pis off of shit chargers eventually forced the raspbian maintainers to add code that downclocks the processor in the event of undervoltage

>>1461001
OP the faggot here, this is a great idea

>>1461063
>The switch is de-bounced with an RC filter and a CD4049 NAND gate with schmitt-trigger inputs
y tho, do that in software. ditch the cap, sample every 1ms, and only signal a start indication when all of the past 24 samples are button-down
>so many flip flops
y tho, do that in the MCU
still, bretty gud, the STM8 arch looks a bit anachronistic to me and good on you for tackling it. assembler or what C compiler?

>>1461196
or, if he's just powering these little thangs, ditch the ULN and transistors, and use the respectable current sinking capabilities of the GPIOs

>>1461239
Do we have room for "RTFD" next to "Take it to the recycler." in the form of some "Rules of /ohm"? Being able to refer to "Rule 1" or "Rule 2" would be pretty slick. Another rule could pertain to the understanding (or lack therof) of ohm's law and perhaps direct folks to the /sqt/ instead, or to /g/ for more upper-level (PC building, etc.) questions.

>>1461169
the cheap way is just solder a brass nut on the back.

>>1461329
it's pretty tight but I could probably shorten some things

>>1460222
>>1460172
Did a little more reading. I'm probably going to go with an IP camera instead of messing around with analog/digital converters. I'm now looking at these modules:

https://hupuu.com/camera/ipcb-3519vs226-d29-imx226-hi3519v101-starvis-12mp-1-1.7-4000x3000-8mp-4k-1-1.9-4096x2160-ip?sku=CBF-D29
https://hupuu.com/camera/1-2.5-sony-imx274-hi3519v101-4k-8mp-3840x2160-h.265-ip-camera-board

Again I'm going to need to worry about latency.

I've also found a lot conflicting info related to the IMX335 sensor boards' (that only do AHD/TVI outputs) capabilities.

>>1461341
here's hoping latency and buffering won't bite you too badly

>>1461337
Could always add an extra post or two like /csg/'s archive links.

How do those batteries look? https://www.banggood.com/4PCS-MECO-3_7v-4000mAh-Protected-Rechargeable-18650-Li-ion-Battery-p-992723.html?p=O516115442892201607W
I wanna build a homemade charging pack and then a homemade bluetooth speaker which will basically be another homemade charging pack in a bigger box with speakers and a bluetooth player.
Also, what the fuck do I use for a charge controller + USB output? I want a chipset with a good reputation and charge protection, that charges from a 12v or 24v wall-wart (so as to dump the electrons in) instead of a micro-usb input.
https://www.ebay.com/itm/12-24V-Charge-Control-Module-Storage-Lithium-Battery-Protection-Board-XH-M603-/182502866102
This is the closet thing I've found but preferably I'd be able to have a display that isn't attached to the board so I can locate it whereever I want it.

>>1460484
>MFOS soundlab ultimate
that's cool man.

the ultimate has 3 vco's though, and that midi interface only does duophonic.
always cool to slave a sub oscillator though.

>>1461169
PEM inserts, probably not on digikey but you can look up the real name and search for those

>>1460787
>NiMHs use a negative delta-V

I knew this, mentioned NiMH, and STILL forgot, despite having explained multiple times why the memory effect in NiCd/NiMH wasn't a thing and it was actually the chargers being timer-based garbage instead of dV/temp.

Fuck I'm retarded.

>>1461330

That's actually genius, yet I feel retarded for not having thought of it myself.

>>1461399
Brass might not solder too well, watch out.

>>1461377
>(tested by the manufacture maybe can not reach 4000mah)
excellent choice

>>1461377
Well charging at more than 10W is a little greedy anyhow, but I'd look for USB type-C charging boards, since type-C can support *a* higher voltage. Might suit you, might not.

i have a bunch of old HDDs, is there anything of value in them worth salvaging?

>>1461454
if you're into backyard metal casting, you could smelt the frames. otherwise,
>I have junk, what do?
Take it to the recycler.

>>1461463
>recycler
hahah nice jokes friend, anything i don't salvage goes straight to the nearest dumpster

I want to design a touch switch which works like this:
two super thin wires run along each other and when a person touches both at the same time it counts as pressing the switch

So i was thinking, i connect one wire to 5V on arduino and the other wire to an analog pin set as input. And when a human touches it he will act as a 100k resistor and will read some >0 voltage on the analog pin and count it as input and then turn on a relay or whatever in the arduino code
Would that work?
Also would there be some simple way to make this work as a toggle without arduino or some other ICs (just with like transistors and shit)?

>>1461485
why are you asking instead of just trying it?

>>1461485
You'll need a pulldown resistor also, probably at least 1MΩ since it's DC. Use AC and you'll get lower resistance through the body. And if you do it well enough you can just use a schmitt trigger and a digital input, no ADC necessary. But there are more efficient methods of making touch buttons/sensors.

>>1461434
there is a little known physics equation that proves that the more yoy don't know whether your battery is likely to become a bomb or not, the better power efficiency you get

>>1461496
Like what?
I tried capacitative sensing but it's trash i get false positives all the time

>>1461508
So that's why I've never had to charge my note 7.

Is there any way to use arduino 1.8.6 to program attiny85s? Seems like all the software was written for 1.6 and earlier.

>>1461582
I'm surprised the software version makes a difference. You could try treating it like a digispark, but we're already 2 layers deep in this faggotry so I wouldn't advise it.

>>1461454
>anything worth salvaging?
rare earth magnets

Good goy! I finally found some solder with real lead in it!

>>1461842

where are you that it is hard to find 60/40?

>>1461853
I could order it online all day, but all of the brick & mortar stores have a very shitty (if any) selection of electrical solder. Still gotta make it to that electronic gadgets store if I can get some motivation to move before 2pm on a Saturday.

That solder has a good amount of rosin in it too. I have 3 other types and never had the rosin leak out of it like that one.

>>1461865
>never had the rosin leak out of it like that one.

i fucking hate rosin leaks.

wait, wut?

>>1461872
Just stating an observation. I have no idea if it’s good or bad, but it seems like it would be good to keep away the oxidation and you can really tell it’s in there unlike the generic little coils in the tubes.

I also melted off a pretty big chunk. I’m sure a tiny spot on a circuit board wouldn’t leave nearly as much.

Personally I'd prefer to get solder without flux core, that way I can choose whether to use no-clean or rosin, since I have both separately anyhow. But both my 60/40 and RoHS have rosin in them, which works fine for most of what I do.

Output impedance of a sallen-key LPF is just the output impedance of the op-amp, right?

>>1461454
sell on ebay as retro tech for big bucks

>>1461508
Hello,
Please delet this.
Sincerely,
Cheng
UltraFire Corp.

>>1461907
It is but only in the passband obviously.

>>1460542
>you gave up? too bad m8

I'm >>1460192 not >>1460484
I'm still working on it.

you are on the money. the issue is as you say.
the priorities in the previous code works out badly. eg releasing a note and the 2 held notes switch dacs and gate outputs.
with analog VCO's that are always a little out of tune, this makes a noticeable jump.

the base code has an array of 88 keys that is scanned with a for loop.
this writes to dacs and gates every loop. regardless of notes changing or not.

I'm thinking of ditching that and go for a procedure that only writes to dacs/gates when a note condition changes.
>note on received (#notevalue)
>check which dac is available(if any)
>write to that dac(notevalue) and gate.

this dac is skipped by the program until the note off for that specific #notevalue is received.

huh, writing this out made it clearer in my head.
I'm not a real programmer, just a circuithead.

>>1461907
yes

>>1461975
output impedance is usually only though of as the series impedance of the output as it acts with the succeeding inputs. any bandwidth/gain is usually ignores.

inb4 transmission lines

>>1461988
>this dac is skipped by the program until the note off for that specific #notevalue is received.
consider a vector clock strategy: keep a large (uint32) master counter, whose value is copied to the selected voice entry whenever you open or close its gate. this "time" is monotonic and doesn't have to be linear, so you can increment it in your main loop instead of using an interrupt
to choose a voice for a new key-down, simply find the one whose gate is closed and whose value is furthest behind the master counter and therefore furthest into release. you could use an analogous strategy to choose which open voice to replace in case of a 4th simultaneous key-down
so your voice struct looks something like
>struct voice {
>uint32_t last_event;
>int8_t note;
>uint8_t gate;
>};
>struct voice voices[VOICE_MAX];
declare your time clock with
>uint32_t now;
increment in the main loop with
>now++;
and select a voice channel for a key-down with something like
>uint32_t age = 0;
>int8_t voice = -1;
>uint8_t i;
>for (i = 0; i < VOICE_MAX; i++) {
> uint32_t vage = now - voices[i].last_event;
> if (vage > age && !voices[i].gate) { voice = i; age = vage; }
>}
>if (voice >= 0) {
> // use it
>} // otherwise no voice available, so ignore event or choose note to evict
just curious, what DACs are you using?

I get the feeling /diy/ needs [code] and [math] tags, or at least just [math] tags.

>>1462056
I get the feeling you're absolutely right
who to bitch to and how?

>>1462014
good idea.
I'm on arduino so i might just use millis() and compare.

for now though, I'm just working on getting it working. it being quadrophonic I'm not that worried about what happens when i press the 5th key just yet.

I'll check back when i got some results.

>>1462014
>just curious, what DACs are you using?
missed that.

just regular old mcp4822.
simple SPI dual 12bit dacs. gets the job done.
the original code used these. I' just expanding it to quad and adding aftertouch.
being dip is also nice for quick prototyping.

original code/project here
https://github.com/elkayem/midi2cv

the 4725 are cheaper and I2C, but you need to order a bunch of different ones becuase they only have 1 address bit broken out, and i need 8 dacs total.

I should mention am a bit of a novice programmer, so the stuff you post only vaguely make sense to me.

>>1462069
thanks for the link, that's a decent design
I wouldn't bother with the I2C DACs especially if I needed more than one or two output channels. those DACs should be just fine. if I were trying to add the pitch wheel into the pitch CV in software, I might want more bits
millis() would be just fine for that clock
I've been doing code and hardware for a long time. you'll pick it up by bits and pieces as you go
>DIP
the 4822 seems to be maybe 50% more expensive in DIP. you might not need them for this project, but those SOIC break out boards sold on ali for ~4¢ each in bulk are really nice to have
it sounds like you got this. definitely check back later!

>>1461508
schrodinger's milliampere hour

Can someone point me in the right direction for making audio play at high quality with a microprocessor. I usually use arduinos and what not but Im good enough at programming c to understand whats going on. But I really dont know where to start here. Like what parts would I need etc. I know a atmega8 isnt going to cut it alone.

>>1462117
start with a micro that has an I2S interface built in, which is more or less the standard interface for 2 channel CD-quality sound. the STM32F103 is a fine example of such a micro. for basic WAV and MP3 it should be enough, but if you want to play MP4 or FLAC audio you might need something faster than the 32F103 and with floating-point arithmetic in hardware, maybe the 32F4xx series
on the output side, get an I2S codec/DAC, there are hundreds or thousands of options
on the input side, it depends what formats you want to play. look up libmad for an integer-oriented MP3 playback library, faad for an MP4 library, or if you have lots of processor power and flash to spare you could look into libavcodec
if you want to synthesize audio onboard, that's a whole other deal. lots of math to that depending on which of the large variety of ways you want to do it

I'm doing some LED strip lighting, 4 separate 12v 6 amp strips, and I want to have them all on separate dimmers. I want to go the DIY route so that I can custom fit it to my space. I've never done much with electronics, so this seems like a good starting point. Do you guys have any tips on how to choose PWM frequency, and how to make the circuit? Do I just do the basic 555 I've seen online?

>>1462120
basically this, but add a MOSFET to boost the power. see
https://www.electronicshub.org/pwm-led-dimmer-using-ne555/
then directly replace the 2N2222 with an IRF540N on a decent heatsink and note that the resistors shown on the LEDs will be built into the strip
>PWM frequency
a few hundred Hz, I would probably start at 250 and multiply it by 1.5 repeatedly as long as I could see objectionable flicker or strobe effects

>>1462120
>12v 6 amp strips
Can you post a link to this product?

>>1462136
https://store.yujiintl.com/collections/high-cri-led-strips-ribbon/products/bc-series-ribbon-120led-2835?variant=16595021511

Sorry, got that wrong, 4.5A/strip. Think my initial math was just total amount to run all I bought divided by the 4 strips I'm putting up, but they're only around 10'/strip.
>>1462130
Cheers, I'll read up more and see if I can begin to understand that.

>>1459841
The battery in my phone (blackberry priv) inflated and looks scary, there aren't any OEM batteries available and I don't trust the shady ebay ones.
How does the charging circuitry in most phones work?
Is all the charging circuitry built in to the battery? I'm wondering if I can just plug any battery+charger of the correct voltage in to the phone and be fine. As long as I find the right connector that should work right?

>>1462238
Thanks for the link. Be aware that PWM can produce annoying noises if the 12V/5A/60W DC source can't handle it properly. Another effect is radio interference. PWM is dirty, simple as that.

>>1462278
Add an inductor and diode to turn it into a constant-current buck converter. That will keep most of the the PWM frequency out of the strip. Add a capacitor just before the switch to limit the area of the loop.

>>1462293
That's a really interesting suggestion, thank you. I know what a 5A inductor would look like and a fast 5A diode.. I'm already back to halogen and iron because experimenting with large 472kHz indoor loop antennas is not compatible with anything that switches.

>>1462120
I have a shitty circuit I made by accident. I was trying to do a PWM with a 555, but I accidentally mixed up the schematic with the VCO. I created this broken ass circuit which changes frequency, and requires tying a relatively low resistor between control voltage and ground, but between 0 and Vcc, you can get a PWM signal for a logic-level mosfet, which works between 400Hz near max percent and towards 1000+ Hz at low percent. It's near linear for brightness and I don't know why it works, but it does.
The CV problem is that the CV is internally biased, but with a high input impedance. If your CV source has a good source but bad sink, off is biased on 50% or 2/3rds Vcc. The lower the input impedance on your source the better the circuit works. At about 1k bleed for off, you get 10% min brightness 1000 kHz. At 5V 100hm input with bleed resistor, you get >90% on. This circuit was tuned more for a pot (1-2k pot) but if you used an op-amp real-zero buffer, you wouldn't need the bleed resistor as long as you can supply/sink around 20mA. I tested with a 250 ohm impedance and no bleed resistor on CV, and you get near 0 and near 100%.

I don't remember if I considered the input buffer or not, since many digital boards only source a few mA on their analog pins. I also focused on 5V Vcc, 0-5V input, and logic-level mosfet output.

>>1462252
>Is all the charging circuitry built in to the battery?
No. It's in the phone.

>>1462387
A low power basic op-amp I have is the LM358n. It can source 20+ mA, and says it can sink 10-20mA with real-zero. It includes two op-amps in one package, if you wanted to control multiple stripes with different brightnesses. Cheap as hell, too.

The motivation for making the above circuit useful once I had found the behavior: I didn't have zener diodes (often required for full-range 555 oscillators). It only requires caps and resistors. The values were hand-tuned a little bit to make the behavior more linear and predictable.

>>1462387
>>1462404
P.s. I created the actual circuit. The output voltage was around 3.3V, which worked with my mosfet: FDP6035AL or FDB6035AL . At the time these were the cheapest I could find (I was bad a things).

>>1462387
Personally I'd use a DIP-8 dual comparator and make PWM with that instead. Make an onp-amp-style oscillator with one comparator, feed its capacitor ramp "triangle wave" back into its second comparator, with a resistor divider feeding a voltage into the other pin of the second comparator. Probably more passives than the 555 PWM, but it doesn't change the frequency with the duty cycle. You can also tune the ramp wave to be a lower amplitude and more triangular, if you require. Oh but you'll need a ground rail for the thing to oscillate on either side of, which might be a pain. It's actually not a better circuit at all.

>>1462429
I wonder if there's a way you can phase delay two ring oscillators, and have the AND be the output. The PWM would be generated by the overlap of the two 50/50 pulses with regards to phase. maybe you could add a slight delay circuit, so they're always synchronized.

>>1462362
to reduce the current * loop area, you could also try running a 3-wire cable to the lights so that the PWM generator passes a relatively small signal over the cable and the switching FET is near the strip, with plenty of decoupling capacitance on the power leads. also try applying the strips to a piece of aluminum channel which you would then earth, for a loop area as thick as the adhesive layer on the back of the strip *and* side shielding
beware that high-CRI LEDs might only deliver high CRI when run at the rated current, so you might need some sort of PWM to keep the CRI acceptable

This is how to properly desolder, right?

>>1462512
Nah something isn’t right here, I think I need more heat...

Almost desoldered

>>1462523
That solder wick was a waste of $6.

>>1462524
Gas torch wins.

>>1462503
Ring oscillators aren't exactly compact or consistent in their own right, but I assume you mean digital ring oscillators, right? Never got one of those to simulate without reaching a steady state immediately, though I guess if you used schmitt triggers that would be a different story. I think you'd have to have just a single oscillator and through variable delay lines (I don't know what that means) change the overlap. I think you'd want to XNOR then, not AND them since you should also have a [0, 0] = 1 condition.

Also PLLs can probably do the voodoo you're suggesting, but I know less about them than delay lines.

>>1462539
meh, it's all a waste of time anyway since you've still got just as large dI/dt creating EM radiation. I don't think spreading the spectrum will help you all that much

>>1459841
Is it possible to make an handheld SDR that can RX/TX on 900 ISR? How does encryption work on it, and will I be able to make it compatible with other radios?

>>1462560
probably, but if you have to ask, you probably can't
however you want it to
maybe, you have a few thousand pages of reading to do to find out

>>1462539
Ring oscillators are as small as three not gates. Doable with a few external components and an 8-pin dip containing three NOT gates. A phase delay should be as simple as adjusting a resistor to the input of another ring oscillator.

https://en.wikipedia.org/wiki/Ring_oscillator

If you run them near their max frequency they don't switch fully. It would be best to use them at a much lower than max frequency and buffer the output. A cheap low-power schottky chip can exceed 1MHz full transitions if I recall correctly.

>>1462560
>>1462565
Would be a bit of a power hog, probably not worth it. I'd cobble together somewhat of a more analogue setup with some sort of orange pi or microcontroller (or FPGA if you swing that way) handling the encryption while a purpose-built transceiver handles the rest.

Recently picked up a hobby of pulling 18650s out of laptop packs. Of course I have nothing that uses them, but it's fun.
Anyway, I need a charger for them. Should I get an analyzing charger? If I don't need an analyzing charger, how do I bin the cells just test runtime in a device?
I was looking at the Liitokala Lii-500 and the Opus BT-C3100 chargers. Not sure if I need to think about LIFePO4 support for the future, in which case those two are out. Any suggestions?

>>1462503
y tho

>>1462585
>Any suggestions?
If you have a source of 12vdc at the current you wish to charge with.

>>1462600
>If you have a source of 12vdc
If you don't - get the ac model with power built in.

what do the dots on the inputs of the first three gates mean? Negation?

>>1462755

I hope someone knows. I've been designing digital for years and nobody seems to know, so we leave them off.

>>1462755
>Negation

Yes

>>1462760
kek

what should I have the "carry in" input on an adder be connected to? Logisim doesn't like it when I have it connected to nothing, but I get incorrect results otherwise

>>1462856
Well, is there any carry coming in? No? Pull it low.

>>1462765
well I'd say inversion, not negation

noob here.
i need to solder 2 USB wires.yeah, the most basic of all things.And i would like it to survive moderate physical stress.

-how do i insulate the 4 wires soldering from each other? i don't have heat shrinking bands small enough and last time i tried my shitty soldering iron triggered them before i was done. i also don't want to cut and uncover too much length.
i was thinking about cuttin them at different distances and wrapping with tape, but last time i tried electrician tape wasn't really good for that, so looking for suggestions.

-general strength of the joint, esp. pulling and bending.
i was thinking about carving a channel into 2 very small identical wooden blocks then hotgluing + screwing them together to form a small block with the wire inside.
does that sound retarded?

Any electonics experts here?

When comparing power supplies with the same output voltage and current, it seems like the cheaper models have higher leakage current. "Leakage current(mA)".

Any reason for this?

>>1463013
trim off only what you need, then fix them in place and insulate them with hot glue

your block idea could work or you could just make a bending stress reliever out of your electrical tape. Make it thick towards the usb port and have it taper off down the wire.

>>1463018
Quality of design and components vs EMI requirements. Power supplies for medical applications have the lowest leakage. You can feel it.

>tfw replacing my multimeter's 1000V 11A fuse with a 600V 15A fuse

>>1463051
So other than price, using a medial grade smps at home aint a issue?

>>1463020
This, though epoxy also works and might be a better option if you want it to be more rigid. Silicone and silastic are better alternatives to hot glue and are pretty flexible, though more expensive.

Personally I strip maybe 3cm of outer insulation off each of the two cables, and then strip maybe 8mm of insulation off each of the 8 wires, then either put a little bit of heat-shrink down one side of each wire such that it's far enough away from the iron, or solder them without and just wrap each connection with electrical tape. To get a good bond with electrical tape it's important to wrap it around in tension, which can be hard to do with a tiny wire, which is why I've ordered some silicone fusing tape. But the electrical tape doesn't tend to go anywhere once I've shrunk a larger piece of glue-lined 4mm or 6mm heat-shrink over it, so it's probably fine unless you really give it a sensual massage.

If putting a piece of heat-shrink over the whole thing isn't an option due to space constraints I'd just make each joint and cover the entire thing, up to and past the outer insulation, with epoxy/hot snot. Fusing silicone tape may be an option for this too, though I'll have to see how rigid it is once it arrives.

>>1462755
>Negation?
>>1462760
>I hope someone knows.
>>1462765
>Yes
>>1463001
>inversion, not negation

I wanna make a ringmod for my synthesizer.
I need to generate a carrier signal for it to work.
I'm familiar with a carrier from making electronic music on my computer, but as for what they are in electronics, I have no idea what they are or how you generate them or what kind of signal you want to generate for one.

where can I learn about them? Google isn't turning much up.

>>1463074
>under tension, which can be hard to do with a tiny wire
>which is why I've ordered some silicone fusing tape
anon, I have bad news...

>>1463100
a carrier is just any unmodulated signal. you could use the sine wave output of any oscillator

>>1463102
would the sound of the carrier signal affect the sound of the output? It does on the software vocoders I use so I would assume so

>>1463100
Ooh, can you provide a little more info on your project idea/what the mixer is to be used for? I've been fairly interested as of late in ring modulators used for similar signals.

An ideal mixer multiplies two waveforms together, and thanks to mathematics we know that:
>cos(w_1*T+a)*cos(w_2*T+b) = 0.5*cos(t*(w_1-w_2)+a-b) + 0.5*cos(t*(w_1+w_2)+a+b)

where w_1 and w_2 are the angular frequencies of the two sinusoids. Meaning in the frequency domain you'll get two peaks, one at |w_1+w_2|, and one at |w_1-w_2|. Ring mixers need a fair bit of beef in the coil department, more-so for audio as opposed to radio frequencies, but at least with audio I think you can get away with using a magnetic core fo some sort. Be prepared to spend a long time winding coils.

But take care if you put two of the same frequency into such a modulator, since you'll get a DC component unless the phase difference = a-b is equal to n*π/2. [since you'll get 0.5*cos(t*0+(a-b)), and to get cos(a-b) = 0 then a-b must equal n*π/2]. This is one of the roadblocks I ran into trying to add extra harmonics to electronic instruments; the idea was that I'd feed an electric guitar signal to an array of mixers, each feeding into one another in different combinations to give f, 2f, 3f, 4f, etc. with potentiometers on each output before combining them to change their relative amplitudes and get some cool sounding effects. Problem is I'd need to shift the phases a bunch; to avoid the DC I'd need to split each phase apart by varying degrees which sounded like something far easier to do digitally. An ADC and DAC are always an option, good luck.

>>1463138
Ring modulators have been used for decades in carrier telephony. You should be able to adapt a pair of amplifier output transformers.
For the diodes, you need low forward voltages. Gold-bonded germanium has the lowest I know of. Like the 1N56a or OA47
You can do the modulation at fairly low power levels and then boost the output with a cheap solid state amp.

>>1463150
Common 50/60Hz centre-tap transformers are probably ok for that stuff, though I haven't tested. Bought some ~1mm thick enamel wire to make some low-turn modulator transformers for a ~30MHz up-converter to listen to lower frequencies with my SDR, though I haven't built it yet. Not sure how many turns I'd actually need, hopefully less than 50 on the primaries. Might be able to use MOSFET ideal diodes with op-amps/comparators if you know what you're doing, but having an extreme winding ratio to make the voltage in the diode ring on the order of 50-100V could work also. Of course that means you'd want another transformer for the carrier.

>>1463105
absolutely. harmonics = more "carriers" that get modulated by the other signal

>>1463150
>not just using some chip with a Gilbert multiplier e.g. the NE602/NE612/NE567
right in front of my salad?

>>1463155
probably less is more
>1mm
>for a receive application
y tho

>>1463157
Just so it's stiff enough to sit off the PCB, I think air-core is the way to go at this frequency, but I might be talking out of my ass.

>>1463158
material 75 would be best for a 1-50MHz transformer

>>1463175
I see a datasheet with "permeability = 3000" on them, could this possibly be absolute and not relative? There is no shown unit for the permeability. Also they're all full of Gauss AND Oersted units, which isn't fun. I don't think I'll be buying one of these any time soon, but knowing that they exist is great; before I'd only seen anecdotal "ferrite is good up to X" etc. comments, not knowing that there where swaths of fancy core materials to choose from out there.

>>1463184
it's relative to free space
>Gauss AND Oersted
since you're not putting all that much power through them, I wouldn't sweat those too much. I mean, you could also cheat and see what other upverter builders have used for magnetics... or you could use a semiconductor mixer, or buy a mixer transformer module e.g. Mini-Circuits

>>1463243
>it's relative to free space
Then that's pretty dandy then, 3000 I thought was in the realm only metglas can get you. Life truly is far more than the wikipedia page lets on. Pic related, it shows both for one stat, then one each for the other two. I can understand that it's convention but surely it would be helpful to share units for intuition's sake.

I'll just write down mixer stats here because I'm lazy:
SA602:
~$0.6 ea (ali)
f_max = 200MHz
Oscillator driver built in

SA612:
~$0.7 ea (ali)
Same other stats

Looks like the NE567 could be used as an IR remote decoder among other things, though with a little more work either of the above would also work fine. 500kHz maximum. Though when receiving a, say 40kHz signal into a mixer, I assume you'd want an oscillator frequency not at 40kHz but maybe around 39.95kHz, so you could feed the output into a low-pass filter, then an amplifier and rectifier if you needed a square wave. If you set the receiving frequency to exactly the same as the transmitter, if the phases matched up wrong you'd get no signal out at all. The 567 on the other hand explicitly states that it directly turns CW into DC pulses. IR remotes are on my to-do list, but for now I'll buy 5 or 10 602/612s since they look damn handy to have around.

>>1463252
to use the SA6x2 for tone decoding, I would probably just use the standard LC bandpass on the output followed by a diode detector
SA6x2 are handy, I keep recognizing this when I don't have them. I keep meaning to order some but for some reason always leave them off the list

>>1463255
You use ali ones? I imagine they're fine since most semiconductors are, and certainly all the one's I've bought have been, but you never know I guess.

>>1463256
haven't tried 'em, frankly, but I would. surplus can be a very random thing. i don't know for sure if I'd buy them for 12¢ each, that might be a bit risky

I was fucking around with an old music keyboard of mine (Casio WK-1630)
I accidentally dropped a lead and shorted some of the leads on the RAM/ROM. The display comes up blank and it no longer responds to any input.

It's safe to assume I wiped the RAM/ROM and bricked it, right? It's fine if I fucked it up, I was going to part it out anyway, but does anyone know of any way it might be salvaged?

>>1463519
It may be possible to reflash the ROM, but I imagine you'd need a copy of what used to be on there, which you could probably only get from another of the same PCBs. May even need a logic analyser to figure out what's written on them, I don't have any experience with ROMs.

If it's RAM then just replace the IC with a blank one, it's volatile memory anyhow, but it doesn't sound like RAM. Look at the datasheet in any case.

>>1463255
wait why do you need the SA602 for decoding if you're already running an LC filter and diode detector? can't you just run the filter + detector on the raw modulated input?

>>1463519
oh but if you're lucky you may find the ROM data on a website somewhere

>>1463519

CMOS chips are very good at surviving that kind of accident, but TTL chips can easily die if you short two outputs. and, of course, signals travel from chip to chip, so the dead chip isnt necessarily the one that received your heavenly gift.

>>1463543
>>1463540
>>1463542

the specific memory module (CY62128BLL-70si) is CMOS based, but it seems to be brown bread. Nothing I'm doing is doing anything, and the chip itself is hard soldered to the board. I might as well just part it.

>>1463540
>SA602 for decoding
Decoding requires software. The 602 can only detect/demodulate.

>>1463554
Fine, "demodulating". It should be fairly obvious that's what the IC itself is being used for, but the context was that it was being used as part of a decoding circuit, in conjunction with a microcontroller. Also there's nothing stopping you from decoding with discrete logic or something like a PAL.

>>1463540
SA6x2: downconversion to an IF (you probably don't want to run this at the frequency of the signal being detected)
LC filter: bandwidth is proportional to frequency at a given Q, which can be optimized for response time vs selectivity according to the needs of the application

>>1463252
I don't know much about synths but I'm assuming you'll be mixing line level audio with whatever your LO is. Bear in mind the SA602/612 is internally biased very close to the positive rail and you can't input large signals... such as line level audio. I think the maximum signal amplitude is around 68mVpk. You can find that info in the NE602 Primer article in ELEKTOR magazine's Jan 1992 article. Sure you can just divide your audio voltage down so that the 602 sees no more than 68mVpk but then you really reduce your SNR and that's probably not good on a synth. I could be wrong but I'd imagine that's not desirable.

>>1463252
there are designs for an LM567 ring modulator out there. might look one up?

When you have a power supply with a dedicated terminal to turn its output on and off. Am I correct to assume the power supply wont take damage if I use that signal to turn it off while loaded.

Quick question about NiMH cells.
I have a single NiMH cell that doesn't charge or discharge properly. Hooking it up to my charger shows it sitting at 1.38-1.4ish Volts, but I can only draw about 10mA out of it using my charger's discharge function set to 500mA. If I try to charge it at 500mA, it'll only do 80mA. What causes this?
None fo my other cells do this.

>>1462278
Is there a better option than PWM for dimming LEDs without affecting the color? Any change in voltage changes the color. Alternately, is there a better way to do the PWM? Is there a relatively easy way to leave the controls for the PWM far away, and then put the actual PWM circuit immediately before the LEDs?

>>1463068

This shit is starting to really piss me off.

Why the FUCK are thermistors such pieces of trash?

I have a digital thermistor that looks exactly like pic and it just keeps showing wrong temperatures.

I basically have a hot metal plate that goes from room temp to about 65°C in about 2 - 3 minutes and i need the thermistor to measure this. The temp changes slow enough for it to keep up no problem but it fucking doesn't. It's off by like 10 - 15 degrees all the time.

I used a piece of kapton tape and taped it down to the hot plate with the flat side facing down so it's there nice and snug with good contact

SO WHY THE FUCK IT DOESN'T WORK PROPERLY

I can't take this faggotry anymore

>>1463846
Add a bit of heat-sink compound or thermal rubber? Thermal contact might just be shit. Are you 3D printer heated-bed anon?

>>1463789
Depends on the power supply, but most likely not unless it's some real garbage.

>>1463819
>better option
No
>Any change in voltage changes the color
The color of an LED depends on the bandgap, not the voltage you're applying
>relatively easy way to leave the controls for the PWM far away, and then put the actual PWM circuit immediately before the LEDs?
Run a very long cable. Otherwise, no.

>>1463819
PWM will affect colour shift less than DC but not by a huge amount, the main cause of the shift is die temperature, not Vf.

There are LEDs with built in PWM drive to do dimming locally like WS2812, when properly decoupled these can be quite a bit better for EMI than PWMing an entire string directly. Also small PWM chips like WS2811 that you can stick next to the LED. Uses a daisy chained serial control system.

>>1463846
>there nice and snug with good contact
I wouldn't bet on that. Also, keep in mind that the back side of the component and the wires act as heat sinks.
Use thermal compound like the other anon said, drill a hole for the sensor or at least use a metal clip (at plate temperature) to keep it in place and arrange the wires to that there's no huge thermal bridge to the ambient.

>>1463877
>The color of an LED depends on the bandgap, not the voltage you're applying
White LEDs often change their color visibly when you change the current.

>>1459841
I have this calculator. it works fine but the screen seems to be broken, it displays numbers badly. Can I replace the screen and keep the calculator or I have to discard the whole thing?

Does anybody know what these kinds of headers are called.

>>1462585
Be carefull with these things, don't ever try to solder them with a normal soldering iron, use either a spot welder or an 18615 holder

>>1463932
http://rarecomponents.com/store/1295

>>1463926
Not really and even if you manage to attach the zebra strip properly, you still need a new display. This means salvaging it from another similar calculator.

>>1463900
>e back side of the component and the wires act as heat sinks.
the thermistor is attached with capton tape and this tape is used on shit like satelites for thermal isolation, this means it won't let almost any heat escape from the back of the thermistor so that can't be it

>>1463944
>this means it won't let almost any heat escape from the back of the thermistor so that can't be it
Yes it can smartass.
Your situation in no way resembles a space environment. You still lose heat through the back, kapton has a way higher thermal conductivity than a piece of insulation like polystyrene, and it's so thin it barely makes a difference in an environment exposed to air. Also you still need heat transfer compound.
Just mount it correctly like all the data sheets on temperature sensors say to do for fuck's sake.

>>1463846
Try leaving it in room temperature and see what it reads? When it settles, hold it in your palms and see how long it takes to settle and/or if temperature correctly rises.

>no datasheet
>no photos of assembly
>just complaining
I bet you use arduino too

>>1463939
>>1463926
>even if you manage to attach the zebra strip properly, you still need a new display.

nope, 99% of the time, the connector is the ONLY problem. the glue has dried and a few connections are no longer touching. there are youtube videos on replacing the connector, but sourcing the part is probably gonna be impossible unless it's a very well known model, like a TI-92.

what cheap heatgun would you guys recommend to for desoldering SMT stuff

>>1463803
it's a 'dry' cell

>>1463877
>The color of an LED depends on the bandgap, not the voltage you're applying
not when phosphorescent coatings are involved, such as what turns blue LEDs into white LEDs

>>1463819
yeah, you can run a three-wire cable with power, ground, and signal. put the MOSFET switch at the lamp ground and put the rest at your remote control station. decouple power+ground with ceramic caps + electrolytics just at the lamp. as I wrote here >>1462508

>>1463879
lolduino go home

>>1463846
because thermistors are inherently low-accuracy devices and you really wanted a thermocouple
>off by 10-15 degrees
y u no calibrate

>>1463932
machine pin headers

>>1464070
anything from ali that won't catch fire

Any of you guys play around with making leads for meters or scopes?

I saw the silicone wire on Amazon and a 25-50ft spool of 14AWG or 16AWG should work, right? But what the hell am I looking for as far as the plugs that go into the meter? Is it 4mm? 4.5mm? I searched Amazon for “Banana Plugs” and it gives me audio stuff and then if I add “Multimeter” to it, it’s full leads. Anybody got a good source for just the plug?

I’m about to collect a bunch of those HF coupons to get free cheap meters, buy $1 zip ties at like 6 different stores and cut the plugs off of there.

>>1464159
>But what the hell am I looking for as far as the plugs that go into the meter?
There is this new thing called 'Google' which is said to be great for helping answer questions on the interwebs.

https://blog.muellerelectric.com/what-size-banana-plug-are-you-using

>>1464168
That’s what I was thinking but then it brings me to a bunch of audio plugs and they don’t look the same as the multimeter lead plugs.

I better start printing out HF coupons and map a route to the 12 nearest stores so I can rip apart those leads.

>>1464168
>>1464173
Would the audio ones like pic related work?

>>1464179
>Would the audio ones like pic related work?
Yes. That's the way they all looked in the past.
It was only recently (to me - 75yo) that the plugs were covered.
This was done as a safety measure.
Say you have the leads connected semi-permanently (alligator clips) to a voltage source.
If a banana plug gets pulled from the meter it may be the 'hot' lead that's still connected to the voltage source.
They'll work fine - just be careful with them.

>>1464201
Ahh ok cool I’ll try it out. Now I need to figure out if I can find that fine strand silicone wire here or if I need to get it on Amazon...

>>1464201
based ancientfag

>>1464206
Ehh, no flexy silicone wire at the Home Depot. I was thinking about speaker wire but not sure about what kind of voltages those can withstand. I got this 16ga lamp cord wire until I get a real reason to order the good stuff from Amazon. I can split the lamp cord at the ends and I’m guessing it can withstand a little bit of power. I just want to play around with electricals.

>>1464227
https://us.banggood.com/wh_buy/Usa/Silicone-wire.html

>>1464259
for a dollar or two more:
https://us.banggood.com/Wholesale-Warehouse-DANIU-1-Pair-1000V-20A-Banana-Universal-Multimeter-Test-Probe-Leads-Cable-wp-Usa-91653.html?rmmds=search

>>1464264
Nah I have regular probes. I wanted to be able to make custom length ones with alligator clips or whatever on the end of the leads.

I ordered the one here >>1464227 and it should be here Sunday. It’s 25ft of each. Test fit with the lamp cord worked, but that flexible thin strand silicone wire will be way better.

>>1463968
>I bet you use arduino too
i do because my time is valuable and i won't hinder myself just because some hipster scoffs at a marvel of technology just because it's popular

>>1464420
If you need the extra space, extra speed, better power efficiency, or some unsupported feature you'll change your mind, but otherwise there isn't really an issue, especially for prototyping with an ATMega. But with my ATTiny the bootloader takes up a third of the memory already, so when you upgrade to microcontroller ICs picked to have minimum specs for the situation you'll want to program in something more space efficient. If it's running off batteries you'll also want to use something more efficient.

But he's just making likely assumptions about your personality based on your "take the easy way" mentality.

So I've got a transformer I want to use for a project, but it's about half an inch too big. Can I cut the transformer laminations down a bit and re-weld them together, or did I miss a crucial part of Phys 122?

>>1464537
Transformer cores are normally designed to operate as close to saturation as possible. So, removing more than a little bit of metal makes them saturate. Saturated transformer runs hot and the output voltage is lower than it should be.

>>1464547
Shoot. Alright, thanks for that.

>>1464547
Is there any way of seeing how close to saturation the transformer is or am I SOL?

>>1464562
Got a variac? Increase primary voltage until current goes way up.

>>1464579
Theoretically, could I make an i7 equivalent using breadboards and a shitton of logic gates?

Theoretically, could I make an i7 equivalent using breadboards and a shitton of logic gates?

>>1464612
no, the i7 is the i7 because it uses transistors a scant few nanometers in size. The size and distance between these transistors is what gives the i7 its speed.

>>1464612
In theory, yes (apart from speed; the speed of light is only 30cm/nanosecond, which limits the speed of complex circuits made from discrete components).

In practice, it would never work. You're talking about something with billions of transistors, so it would have millions of connected PCBs, and the probability of having every single PCB and interconnect functioning correctly is minuscule.

>>1464615
>>1464624
So should I cancel my amazon order?

>>1464626
haha what

>>1464579
>variac
Fuck this, you need to increase the load current or use a variable voltage inverter, not chop up the waveform like some demented electron butcher. Capacitive loads might work.

>>1464825
Hooo, boy...

>>1464825
ironically, going all "demented electron butcher" on the waveform is exactly what he needs to do to use a smaller transformer

>>1464626
>>1464612
It would have been smarter to use an FPGA but the answer would have been the same.

>>1464825
>>1464879
Fortunately, if I turn it 90 degrees, chop the old mounting plate off and take of the terminal blocks it's *just* small enough.
90VDC 10A is really hard to get ahold of.

>>1464893
You could still probably more efficiently use a SMPS, even a home-made one. You'll need a much smaller output capacitor as opposed to with a transformer anyhow.

>>1464893
>90V 10A
damn son

>>1464915
I hate to cry 'feed me', but I've never done my own SPMS and all the ones I can find are too big. Any sort of things I should be looking for?

>>1464957
ignore him. he's a shitposter that overcomplicates literally everything and doesn't know what he's talking about
you'll need to know how to work with high voltages safely (400V) and also wind your own magnetics. it may not be worth the trouble and risk. just buyfag this one, if all you need is a single supply
https://www.aliexpress.com/item/NEW-Single-Output-1000W-Switching-Benis/32813327376.html
what are you supplying?

>>1464964
Some power supplies. Specifically some Rui-Deng 85V 5A digital ones. Given the transformer *just* fits and I've got some caps and a bridge rectifier, i'mma just finish this out.

I keep forgetting that the soldering iron gets hot.

How do you sort your resistors?
Can't find these containers on ebay/ali

>>1465006
I use SMD parts kept in a sample book.

How do I learn how to make design circuits like this. Ive always had to look up the designs for a chip to use but wanted to know how the creators knew what capacitors, resistors, etc to put in and how.
I feel like if I really understood what was going on I could only add the relevant parts and start designing more personalized and condense circuits.

>>1465006

i got a new box of these at the thrift store for $2. use foam or potato for corks.

>>1465017

trial and error. e.g. for each cap, you start off at 1F and reduce step by step dow, to 1pF until it works.

>>1465017

trial and error. e.g. for each cap, you start off at 1F and reduce step by step down to 1pF until it works.

>>1465017
Datasheet and Application Notes for the chips you're looking at are a good starting point.

>>1465020
>potato

>>1465017
>>>/diy/ohm

>>1465006
that is a bait tray, which I also use to sort some of my components
the little vials are kinda cute tho
the last batch of THT resistors I got was on cut tape sections of 10 each. I placed them into outer bags by decade and stacked those in an old Aliexpress shipping boxes which are somewhere around 6cmx10cmx8cm. I do the same with my THT caps, which arrived individually sub-bagged by value

>>1465010
sample books are patrician-tier. I should order some more of these

I found this over at 3dpg, how much energy can it make?
I am trying to decide if it's worth the effort making it
https://pinshape.com/items/27184-3d-printed-self-powered-motorgenerator-arduino-speedvoltage-controled

>>1465045
>I found this over at 3dpg, how much energy can it make?

It loses energy.

>I am trying to decide if it's worth the effort making it

Do you like youtube tinfoil hattery and/or making complex spinny things?

>>1465047
>hurr durr i am so smart earth is round coz they told me it is
ok

>>1465045
Likely not worth doing, you'd need a solid core of silicon steel laminations to make it comparable in efficiency to a commercial motor. The best homemade motors will use a fairly small internal rotor and hence minimal need for electrical steel, personally I'd go for a permanent magnet brushless synchronous motor.

That 3D printed thing has lots of 5 coils in a row when it's only necessary to have one, the design is in general pretty poor. Modern motors will have 3+ coils that go the full length of the armature, which will often overlap, depending on the design of the motor. PLA or ABS is probably far too flexible to hold its shape well for such a use, provided you want it to spin in the first place.

From what I understand the "voltage converter" is just a rotary transformer/inverter: put DC across the stator, move the rotor, get AC out at a different voltage. Just use a transformer, or buck or boost converter if you want a different voltage. With all that filament wasted you won't have much to show for it, at least research motor construction before blindly following the design of someone who clearly hasn't done their research.

>>1459947
>8bit adc/dac
>8 rf switches between antenna and frontend
lad.....

>>1465006
i have one of these. it's a flimsy piece of trash like everything off amazon but it works fine.

Where do I get a cheap power adapter that won't burn down my house?
Need a 12V/0,5A EU adapter to power a nixie clock kit.

>>1465006
They're all tangled together in a plastic bag under my desk, half of them have fallen off the paper strips on either side, and a dozen of them are stuck in the carpet. Each time I reach for a new one more get detached from their paper strips in the process of untangling and another one or two falls on the floor. All their leads are tin-plated iron and the tin-plating is wearing off all over the place so they're slowly rusting. Aside from that it's pretty good I guess.

>>1465107
my brother

>>1459947
>HackRF
bit of a mistake on my part. don't do this unless you're working with transmitters within spitting distance

>>1465092
>6W
are you sure?
meh buy cheap, buy two different ones, and pop them open to see which one looks most sound. then use the other one for something where you'll be watching it

>>1465127
>are you sure?
Yeah, why?
>pop them open to see which one looks most sound
What should I be on the lookout for?

>>1465129
comfortable clearance between the mains and isolated copper, clearance between windings on the transformer (they're not separated by nothing but enamel and kapton tape), a MOV and fuse/fusible resistor, and ideally a named controller with fault detection.

>>1465129
a quick check:
a CE symbol is usually a good sign
compare distance between line side and load side (wider is better)
compare any cutouts between line side and load side (cutouts are better)
inspect for obviously dodgy wiring or soldering
ensure no solder balls running free inside (yes, really)

>>1465134
also this

I need some help. I have an old VFD display from a vending machine, and it uses the 10937p-50 IC to control the display. The IC uses synchronous serial for communication, but I'm using arduino, and I don't know how to add the clock pin. I tried SPI, which is the one that gave me better results, because I can adjust the frecuency (the IC uses 66kHz), but I think it misses some bits, i don't know. I also tried shiftOut(), but all I get in the display is @, which in the datasheet is basically 00000000 sended.

https://www.hobbielektronika.hu/forum/getfile.php?id=243736
This is the datasheet I'm using.

>>1465134
>a CE symbol is usually a good sign
No, but lack of it is a bad sign.
Signs of actual test laboratories are a good sign, except that the chinese and the other scammers are just as able to print those than anything else.

>>1465137
That's just bullshit.

>>1465138
A pic oh the VFD using SPI. The other boards are for supply and stuff.

>>1465138
you can honestly just bit bang it. something like this:

>void write(char)
>code to shift from ascii to the char assignments in table 3 into a cmd byte
>for (i=0;i<8;i++)
> set clock high
> set data pin to the first bit of your byte
> set clock low
> byte <<= 1
>set clock low

add nops as necessary.

>>1465137
fucking KEK

>>1465138
pretty
that chip expects data on the falling edge of the clock. maybe try one of the other four SPI modes (mode 1 or 2 looks good)

>>1465146
I've tried every mode, and mode 0 is the one that "better results" is giving to me.
>>1465145
That's what I thought first, and what I didn't wanted to do, but I'll try. I can make a nop between a clock pulse and a data pulse. It's a heavy way to write something, but okay.

I don't know a lot into these displays, so correct me if i'm wrong. If I want to write something, after I disable the POR pin, the characters will write from left to right for each data I send? And then, at sixteenth character, i should stop sending data, and if I want to write again, a pulse to the POR pin, and all again? This seems like the obvious way to me, but I may be wrong.

>>1465149
>a pulse to the POR pin, and all again
no, look at Table 1: Control Data Words. you move the data pointer back home (0xAF?) and write your new chars

Based Amazon free Sunday delivery. Got muh silicone wires for makin test leads and burning more fingers with the soldering iron.

>>1465228
meanwhile my chink boards will be here by 2022

I'd like to make a logic circuit for a lighthouse. At first I thought about using a simple 4017 counter driven by a 4049 oscillator to light up one of ten leds sequentially, emulating the rotating light beam of a light house.

But how could I make a version where the on/off transitions of the leds would have gradual slope instead of distinct on/off states? That would give a much nicer effect. Any ideas? Is this feasible to implement with logic ics?

>>1465238
i'm not entirely sure what you want, do you want like 5 leds in a row that get dimmer and brighter progressively like in your pic? since i think real lighthouse lights just stay completely on.

>>1465240
ignore this i somehow missed half your post

>>1465238
let's try this again. i see three options:

>use a microcontroller
please do this

>use pic
each shift register tap drives a few leds, with its central led being driven the strongest by using the lowest value resistor. choosing the resistor value lets you account for nonlinear brightness. the shift register feeds back into itself. you need n^2 resistors so you'd probably just want to do 5 leds or so. i think this is the minimum amount of actual logic you'd need. you only really need two ics (sr + 555).

>use a pwm circuit for each led
you'd need a shift register as shown but instead of n^2 resistors it would be n pwn circuits with fade-in and out on the voltage control. i don't think it'll be practical.

>>1465265
fuck me

>>1465238
What about using a 555 for PWM

>>1465234
I’m way to impatient to order anything from AliExpress. I would rather pay the extra 30¢ and have it 3 months sooner.

Maybe this is more of a question for the sparkies, but with pic related, are the long cutter-crimper tools significantly better to use than the flat all in one type tools on the bottom?

Mostly for crimping connectors, but it also seems appealing to have worthwhile cutters on the same tool.

How can I prevent a cheapshit chinese 5v AC/DC from making a high pitched noise?

If I have a 3-phase and ground setup and I need to run a single phase fan off it can I put it between one of the phases and ground?

>>1465278
light load whining is caused by the flyback circuit inside your chinkshit brick setting itself to a low frequency (as opposed to 100kHz+ operating) to improve efficiency. if it's badly designed that frequency will be low enough to be audible from piezoceramics or the transformer vibrating. so reverse engineer the control circuit, hope there's some passives that control the switcher's minimum frequency (if you can even get a datasheet), and change the passives to make it higher. otherwise there's really nothing you can do except to get a better brick.

>>1465310
>brick
>control circuit
you underestimate my frugality
If I were to get a different one how could I ensure it didn't whine, make sure it switched at a higher frequency? Would the minimum freq be noted in a datasheet?

>>1465319
>Would the minimum freq be noted in a datasheet?
only for the controller, not the converter as a whole. by control circuit i meant the internals of the converter. my post wasn't entirely serious because it would be impractical to do what i described. all you can really do is find a quality UL listed device and hope.

>>1465278
You could try hot-gluing everything steady, perhaps desoldering the transformer so you can get some hot snot between it and the PCB. I think the transformer will be what's creating the vibrations, not any capacitors. I'd be doing the same thing with mine if it didn't happen to be the apple all-enclosed glued-shut style with no non-destructive way to open it. Also you could try switching it at the wall (automatically?) when not in use.

>>1465238
yes, microcontroller this. if your choice of micro is short on PWM channels, check out the PCA9685 LED controller

>>1465238

this has been done a billion times before. search ''knight rider circuit'' in google images and pick a design you like. my favorite simple method is just to add caps across the LEDs or lamps.

>>1465578
But then they don't take time to charge up, only down. An LC low-pass might work well however. Single inverter looping a SIPO shift register with each output powering an LED+filter one one side, and then to an inverter that powers the LED+filter on the opposite side. It's a fair few chokes however, and does kind of shoehorn you into having 16+ LEDs unless you want to use discrete D flip-flops.

>>1465586
>But then they don't take time to charge up, only down

not true. LEDs almost always have current limiting resistors. that creates an RC charge curve, followed by a discharge curve. i.e. brightness ramps up then ramps down.

>>1465586
>But then they don't take time to charge up, only down

not true. LEDs almost always have current limiting resistors. that creates an RC charge curve, followed by a discharge curve. i.e. brightness ramps up, then ramps down.

>>1465652
I was assuming you'd put the resistor right next to the LED, but that does make sense to use an RC filter with the existing resistor. Though I wonder what that curve would look like with the shockely diode equation taken into account.

>>1465663
>I wonder
Spice it, then

>>1465674
I did, it wasn't that exciting

Found this oscillator in my $3.50 network cable tester, works pretty well. Anyone recognise the topology?

>>1465719

looks like a variation of this.

>>1465719
The thing has a Johnson counter that directly feeds 8 LEDs through into the 8 RJ-45 pins. On the other side it has 8 LEDs pointing out of the socket, into a common rail to 3 regular diodes, pointing to pins 0, 2, 3 (so in antiparallel with 3 of the LEDs) where the current flows back through 3 more antiparallel diodes on the same pins back to the counter. The counter is clocked by the wacky oscillator. The thing requires a 9V battery but has no resistors in series with the LEDs.

I suppose the aim is to have multiple return paths, but until I buy a battery I'll have to wait to see how it functions.

>>1465727
That's what I was thinking, but since it only has one capacitor I'm lead to believe it acts a little more like a 555 or comparator multivibrator. The output wave doesn't simulate to be very square nor very close to 0V. The IC is a CD4017.

>>1465728
CD4017 has a Schmitt trigger clock input

test

>>1465747
gratz

>>1465238
Micro with software PWM to do it is ezpz.

>>1465819
>PWM
>blinking
disgusting

>>1465284
Yes. Between L1 or L2 or L3 and N.
Pic: 3~/400/N/PE Distribution.

If i have 5V power source DC and i need 3V i can use one resistor to drop 2V.
So why do people use voltage dividers with two resistors instead?

For anyone who cares the Rock64Pro single board computers with 4GB RAM are back in stock. Basically it's a raspberry pi on steroids with more RAM and CPU (+64 bit architecture).

https://www.pine64.org/?product=rockpro64-4gb-single-board-computer

Just ordered myself one for my digital NV project.

>>1465999

dont do this, nigga. using 1 resistor is just as retarded as using 2. coz your load is almost certainly variable, then the resistor value has to change dynamically to adjust. resistors dont change dynamically: voltage regulators do.

at the very least, use 3 diodes in series to drop a fairly steady 1.8V to 2.1V.

>>1466015
>PCI-E slot

>>1465999
>why do people use voltage dividers with two resistors instead?
because they only need a voltage, and the load they're using doesn't draw a significant or well-defined current from it. thus, we apply the significant and well-defined load of the bottom resistor

NEW THREAD

>>1466111
>>1466111
>>1466111

>>1465940
PWM for the gradual slope.

>>1465284
3 phase + neutral + earth?

What voltage?