/diy/ - Do It Yourself

this thread's digits brought to you in part by the venerable CA3081 common-emitter transistor array

What's the best way to get an oscilloscope to sync to a shitty signal?

I can hardly get mine to trigger properly on this.

>>1402649
>Doesn't really help me decide but that was a really cool watch
It's such an underrated youtube channel. Their latest series on Chinese manufacturers is super interesting and well made but their videos have like 2k views

>>1403148
those spikes are the problem.

does the scope have "ext" sync input? if it doesn't have i think is gonna be difficult to fully sync.

why you need perfect sync? do you really need it? that signal looks easy to troubleshoot.

>>1403171
Yeah, it's got an external sync.
I actually was using it with a function generator set to ~near the signal's frequency.

I guess I don't really need a perfect sync, but I would prefer one if I could.
I'm just working on a simple low-pas filter for a PWM based DAC.

>>1403176
read the manual of the scope and use the "ext" input basically you feed the same signal to the scope regular input and to the "ext" input but read the manual because there will be some limitations for the voltage that it can handle and the manual should have a schema of the conecctions.


(i fucking hate those noise generators... i'm talking to you PWM)

>>1403182
I'll look into that.

Also, yeah, I'd agree that PWM is not ideal, but I don't have enough resistors at the moment to make an R-2R DAC.
Since I don't need the voltage to change too terribly quickly for this project, I figured I could just get away with a really heavy-handed lowpass filter, but it's looking like PWM is just too noisy.

>>1403191
you might need a couple or three filter stages to clean that up properly

>>1403204
Yeah, probably.
I think I'll just order a proper DAC and work on the other things in the meantime.

I just finished an accelerated electrical theory/applications course and have no fucking clue how transistors work. Why does doping the different parts different amounts have the effect that it does?
Also, what is actually happening with negative voltage? The schematics we were shown made it look like voltage was just magically being pulled out of the ground into the positive terminal rather than a true polarity reversal of the power source. Is that a common way to show how things work?

>>1403252
always remember that the negative is the source of electrons, the positive "lacks" electrons, this lacking characteristic makes it attract electrons. the "movement" of electrons is from negative to positive.

>>1403081
>YouTube
>not including bigclivedotcom

>>1403081
Is that a jumping spider?

>>1403148
I'd buffer that signal, feed it through a 2nd order low-pass filter, and trigger off that, possibly with a schmitt trigger in there if the EXT input requires a digital signal, I've never used them.

>>1403354
>Is that a jumping spider?

can't be sure
it's a jpg - not a gif

>>1403252
For negative voltages:
Voltage is relative. That means that if you have a negative voltage, then it just means that your ground is at a higher potential than the "negative" voltage source.
Here's a crappy image I made to help illustrate.
Since we declare the middle contact between the two batteries to be ground, anything connected to the lower contact sees a negative potential with respect to our chosen ground.

Can you buy just the square bare metal wire they use in component pins?

>>1403377
I'd like to know this as well. I autistically save all my cut off component legs for when I work with protoboards. They're great because they hold their shape well, they don't "crinkle" like wires, and they're tinned.

>>1403388
>They're great because they hold their shape well, they don't "crinkle" like wires, and they're tinned.
What do you mean crinkle, like crumples/folds when you try to insert into a solderless breadboard?
Are you using something other than tinned solid copper hookup wire?

Question about power usage. I'm currently looking into a appliance that pulls 193kWh on average over the course of a year. Unless I've gone full moran, that is about 528 Wh a day(on average), which at 120VAC, is about 4.4 Ah a day(on average). This translates to 0.183 amps an hour of AC 120V. So if I was running it through a converter, I would need a bit over 2 amps an hour of DC 12V, correct? Or am I not understanding this shit at all?

>>1403046
Are you a moron? Programming has zero consequences when it comes to practice. I seriously hope you are just trolling.

>>1403377

nope, but you can cut the legs off of wirewrap sockets.

>>1403394

if you do it backwards, 2A x 12V x 24 hours x 365 days, gives you 210kWh, so it's correct when you add potential losses.

>>1403393
These, not breadboards. Solid copper hookup wire is much bendier and develops ugly ass kinks.

>>1403403
It's probably cheaper to bulk buy dirt cheap components from China and cut off the legs. But I think OP is looking for longer lengths.

>>1403354
why yes, it is
I'd love to know what was going through its little mind when it saw one of my SOT23-5s in my component tray

>>1403405
Much appreciated. I've been going over the numbers and felt like I was fucking up horribly somewhere.

>>1403406

Show a picture of something you have made using component leads instead of "bendier" copper wire, or show how the bendier copper wire is a problem.

>>1403377
https://www.ebay.com/itm/Pure-Copper-Wire-Round-1-2-Round-Square-12-14-16-18-19-20-21-22-24-26-Gauge-DS/281608645956?hash=item41912eed44:m:m7K_Nd9MrEkctDdUZbHhNhQ

>>1403394
>amps an hour
It's just amps.

>>1403376
While voltage in free space is relative, charge is not. You can measure the charge of an object via an electroscope/electrometer, and it can show you that there are more negative charges than positive or vice-versa. Since everything is capacitively coupled in one way or another, making a (dynamic) relationship between absolute charge and voltage, therefore voltage is not strictly relative. This doesn't matter in everyday use but I imagine it would make a difference in some fringe cases.

>>1403516
what's the most negatively and most positively charged type of matter known?

>>1403252
You will have to read up on semiconductor physics which is outside the scope of plenty of electronics courses, the gist of it is that semiconductors are conditionally conductive and doping affects these conditions. In a BJT transistor the minority carriers from the base allow mobility for the majority carriers between collector and emitter. I think it's functionally same to model an AC circuit so that one terminal is always ground and the other goes from positive to negative as it is to model the terminals to alternate being grounds and positive terminals

>>1403530
it doesn't really depend on matter, in nature a thunderstorm is capable of megavolts charge, that means the clouds have gained a tremendous concentration of negative charge in the order of megavolts with respect to ground. i think this is the biggest natural "difference of potential" that you can encounter.

the ignition system of a car generates a spark thats in the order of 15 kilovolts, theres transmission lines operating at 500 kilovolts or even 1 megavolts.

i dont remember exactly but the charges we encounder in our daily life (getting zapped when get you get out of a car or when robbing a carpet with you shoes etc) its in the order of 1 to 5 kilovolts, this is why we have to be careful when manipulating semiconductors or chips, the charge could be too high for then.

>>1403530
If i'm not entirely mistaken the dielectric constant defines how well an insulator stores an electric charge

>>1403530
Electrons and anti-electrons? They're lighter than quarks with nearly the same charge, but don't stick together nicely like the strong force does to quarks and their resultant baryons/mesons.

>>1403550
CCTO is some pretty interesting stuff, like the mu-metal equivalent for capacitors. Though technically the insulator doesn't store the charge, rather it mediates the electric field. At least when ignoring the charges in the insulator storing extra energy via electrostatic induction.

>>1403516
I guess what I'm really trying to ask is where the electrons come from in a schematic like pic related. Are they just beinging pulled out of the ground, infinitely, due to electrostatic pressure? or is the ground supposed to be connected to the negative terminal of a power source somewhere else? How can I tell if a ground is just a copper pipe the ground or connected to an actual power source?

>>1403575
>Electrons and anti-electrons?

thats a good one, its wierd how he presented that question.

>Though technically the insulator doesn't store the charge, rather it mediates the electric field...

thats a good aclaration many times its difficult to visualise and grasp what is going on with electrical phenomena.

>>1403590
you are supposed to connect all those to the negative of a power source, why they do it this way instead of making a clear line ? because is just a form of making very simple schematics, you gonna get used to this type of thing, sometimes the negative connection line is clearly drawn.
when you have complex schematics is useful to use just the ground symbol instead of "polluting" the drawing with too many lines (its understood that you need to connect all those grounds)

Hey guys, I need some help with organizing my arduino code.

I'm trying to build an autonomous robot that should go around my garden and detect certain objects and spray them with water. I'm planning on having a RPi be "the brains", while an arduino mega will do the low level management of the motors and sensors. The robot will have two dc motors with encoders to drive around, and two steppers without any control feedback to point the water jet (first question: should I just go ahead and add some position feedback also for the steppers?).
Right now I'm just trying to get the hang of interfacing all the motors and sensors with the arduino, and for that I've downloaded a bunch of libraries (Adafruit's AFmotor, accelstepper, encoder). Of course I can make a simple sketch to get them all running, but it's not exactly useful. I don't really know how I should organize the code in an efficient manner so that whatever I will run on the Pi can be easily integrated and use the arduino to steer the robot.

For example, should the final version of the .ino file contain a lot of code, or just some basic setup and a loop that just says "motor1.run(); motor2.run(); sensor1.read(); ...etc" ?
To start, I was thinking of writing a library that integrates encoder and dc motor commands, something called drive.h that you use to just tell the arduino "drive for 1.5 meters at 0.5 m/s", and maybe another one called aim.h that you'd use once the camera/lidar has identified a target (this will be done by the Pi probably), so "aim at target 4 meters away, direction +56 deg". Then the library would just take care of the rest.

Idk I'm just looking for ideas/pointers

>>1403590
In circuit diagrams the ground symbol almost always implies the grounds are just connected together with no relevance to actual ground. Actually using the ground as return path is done in SWER power transmission. Making an actual ground connection is something that concerns the power circuitry of the appliance where the neutral wire represents ground with possible separate safety ground connection that never carries current in normal operation.

it's not clear from the chink amazon listing
would something like this come with the necessary hardware to make it work out of the box, or would I need to roll my own to make the different displays light up in accordance with the input?

it seems like it's just the LED and PCB but I can't be sure

https://www.aliexpress.com/item/ilikebigpenis/32866583975.html

>>1403611
message the vendor, they normally reply within a couple of days

Why does the source powering a mosfet gate has to be connected to the same ground as the power source that the mosfet is turning on and off?

>>1403635
>Why does the source powering a mosfet gate has to be connected to the same ground as the power source that the mosfet is turning on and off?

No offense intended, but have you ever studied anything about electronics, like Ohm's law, simple dc circuit analysis, voltage vs. current, etc.?

Some things are so basic and fundamental they don't need to be asked if you understand the very basics.

>>1403635
Depends on how autismally you want to see it, but simply put, current can flow only in loops and you need current for charging the mosfet's gate against its source. To form a current loop you have to make two connections between the driver and the mosfet. Using ground for one of those connections is usually the most convenient option.

>>1403642
I just don't get why I can't use two different grounds. As long as the electrons have somewhere to drain to, it should work.

>>1403657
If the grounds are galvanically isolated then you can't charge the gate because there is no path for current to flow.
>As long as the electrons have somewhere to drain to, it should work.
It shouldn't, that's not how electricity works.

>>1403657
Ground is a relative term.

If your two different grounds are at different potentials (voltages), then you get something called a ground loop.
In a ground loop, current flows between the two grounds since they are at different potentials.
This is a bad thing and will absolutely make your circuit behave differently than you expect.

This kind of thing is common in audio equipment.
For example, you have an audio amp on one circuit, and that circuit's ground is actually earthed, so it's pretty truly grounded.
But the audio source, lets say its your phone, could be on an entirely different ground, one at a higher potential.
When you connect the phone to the amp via a aux cable then you get a ground loop.
Current is now flowing from the phone's ground to your amp's ground.
This kind of thing would introduce buzzing or pops and crackles in the amp's output.

>>1403694
That actually makes sense now. Thanks.

>>1403607
maybe you got this already figured out but the problem with this type of "movement programing" is that you eventually will encounter problems with the precision of the robot getting to the target.

i think you gonna have to use other type of mechanics for tracking the movement, in industrial aplications they use wires in the floor to guide the robot.

>>1403607
Use limit switches so you can recalibrate the steppers when necessary (after losing power). The whole point of steppers is that you can steer them without feedback.

Why encoders on the drive motors? Do you expect wheels getting stuck and/or wheel slip?

With a preempt-RT kernel you might be able to get away with doing everything on the Pi, saves you some headaches.

Good luck, seems like a massive project to get something remotely useful.

>>1403715
>wires in the floor
interesting, I didn't know. It's not really the direction I want to go though. Could you give more specifics on those problems you mentioned?

>>1403718
>steppers
I'm just worried about skipped steps. I was also thinking about using EEPROM to save the position of the steppers after power-down, but I guess I'd have to save pretty often because if power is lost suddenly (eg battery is yanked out) I'm fucked.
>encoders
Isn't it the normal way to go with DC motors? So you can tell how far you've driven (+kalman filter with other sensors eg IMU)?
>everything on the Pi
I heard it's not optimal because it can't time things as well as a microcontroller, so things like encoder interrupts wouldn't work as well.

>>1403732
>I was also thinking about using EEPROM to save the position of the steppers after power-down,

think for a minute. If you trust the stepper to be in a known position, why save to eeprom? just drive the stepper to one particular place every time you shut down.

or do like the other guy said, and on power up, run to a limit switch or detector, which is what works best for steppers. Or do the old floppy drive "save 20 cents" routine, and run the stepper into a hard stop for 100 steps or whatever it takes to know where it is.

your plan to use steppers to aim the nozzle is fine, and you don't need feedback if you design and debug your system. Using cameras or whatever type of sensor to navigate the garden is all the feedback you need.

I know this is a long shot, but did anyone here design a board using an STM32F3 and a SMD oscillator?

I'm looking at a 8MHz abracon ABM3 smd crystal, and while it satisfies the minimum transconductance requirement (gm/gmcrit is ~11, while >5 is required) with its load capacitors and ESR, I'm not sure whether or not I'll need to add an external resistor and therefore reduce the transonductance ratio if the Iq is too high. I cannot know this until I get it and measure it though. Anyone try it with this or a similar crystal?

>>1403753
Yep that's true, I hadn't thought about doing it that way. Though it doesn't really solve the sudden power off or skipped steps problems, but I guess I'll test it and if it's no good then I'll add a switch or two.

But anyway my original question was more about how to organize the code, though of course I still appreciate all input.

>>1403458
It's not so much of a problem but I find component leads nicer to work with when making jumpers because they hold their shape better, are stiffer, have a tighter bend radius (thinner gauge), and they just look nicer. I also greatly prefer component leads when doing dead bug soldering for the same reasons.

>>1403835

sounds like you're using the wrong kind of wire. 24 gauge insulated solid ''telephone'' wire is so much better in all respects (except stiffness, and i got too much of that already). 30 gauge is better for denser designs.

Anyone got nice idea for esp32?

>>1403933
IOT buttplug with humidity and gas sensing

I have two-cell 12Ah (7.4 v) lipo battery that is going to be drained up to 10-15 amps.

Is that safe to also run the 5v electronics from it via a DC/DC? It has capacitors in it.

I mean 15amps is a lot in my book, would it fuck my raspberry pi somehow?

>Pic unrelated

>>1403933
see FAQ #3

>>1403941
does your home branch circuit jam 15 amps through everything that's plugged into it? no? then stop asking this question

>>1403943
What? I am building a robot with lots of servos. They’re going to create a lot of spikes of high current, so I legit need to know, give me a brake.

>>1403945
my bad, wrong question
it'll be fine, anon, the dc-dc should handle regulation quite well enough to keep the logic happy

>>1403959
>it'll be fine, anon

the correct answer is 'try it and see'. there's no way to analyze something like this based on zero technical details.

>>1403590
Metals by definition have free electrons in their conduction band, and these electrons are free to move from metal atom to metal atom, we refer to it as a sea of electrons. Insulators do not have this sea of free electrons and semiconductors are somewhere in between. Say you have a generator that's just a permanent-magnet rotor with coils in the stator. By spinning the rotor you cause the free electrons in the windings to experience an electromotive force, which pushes the electrons down one wire and sucks them up the other wire. In reality the electrons actually move pretty slowly and the wire is just propagating the electric field very strongly. But nowhere is the wire actually connected to the Earth.

In modern power delivery systems, 3-phase power is delivered to substations and streets, with a phase or two being fed to each house. They balance it such that a similar amount of power is used by each phase, but any extra is either fed back to the generator via SWER or a small-gauge neutral line. With an ideal 3 equal loads on 3-phase, the current flows back and forth in between the three wires without needing a seperate neutral.

>>1403989
Well standard DC-DC converters have input caps, so if you put a diode before the caps the voltage spikes won't drain them at all. Big enough caps will eliminate the issue entirely, but chokes here and there would probably help.

>>1403882
>30 gauge is better for denser designs.
correct!
Pic related: my Lattice iCE40HX1K "dev board", built on a chinkshit adapter plate. 1.2V VCC supply at center, 25MHz MEM(E)S oscillator at center-right. note use of pre-formed trade-secret MicroBodge™ 30AWG hooks all over because no 1206 caps in stock
also, an example of how Kapton tape is a life-changer

>>1404026
top view. short 30AWG wire-wrapped connections for 3.3V I/O wiring won't interfere too very much with secure terminal attachment, see white wire at top right, and wires can be undone later without too much swearing should I need to set VCCIO differently for one or the other bank
(the bottom-right edge of the chip is where I finally got the hang of drag soldering)

>>1404033
>doesn't pic related
fucking roads

How do I completely disassemble two old CRT monitors without electrocuting myself to death?

>>1404058
http://www.instructables.com/id/How-to-dismantle-a-CRT-monitor/

How do I create a nice-looking, customizable HUD? Something like pic related.

Anyone have an opinion on whether it's worth it to get the $500 siglent 4ch/100mhz scope over a $350 ds1054z? Been waiting to get a proper DSO for so long, making due with an old analog one, but it's time to make a purchase, my current project demands it.

>>1404026
>>1404035
>$5 FPGA
Interesting. What do you guys do with these?

I i have a relay that needs 10v to turn on and i have 10v power source, how do i turn on two relays?
Do i connect them in series or parallel?

Dunno if this is the best place for this but /ohm/ is my home so:
I ordered something from amazon.de but I had a small issue with it, contacted the seller (China-based) in English and they replied in German asking for my delivery address. Is this legit? Their reply looks like an automatic message, and it's all done through amazon so I'm pretty sure it's not phishing. But still, why would they ask for my address, don't they have it already?

>>1404166
parallel

>>1404169
Forgot to mention they want my address to send me "a new article for free"

>>1404170
>>1404166
parallel but make sure your power source can supply enough current for both relays.

>>1404171
They relay only says "10V to trigger" how do I find out how much current is that?

>>1404175
Try to find the datasheet or measure it

>>1404169
>don't they have it already?
No, Amazon has it but they don't want to handle the issue and risk data protection problems. "a new article for free" sounds like a replacement of the article you had problems with.

>>1404175
If you have a multimeter you could simply measure the resistance in the relay coil. Is it 10V DC or AC? if it's DC then what you measure is the resistance and you just use ohm's law and you're done. If it's AC then you will only measure the real part of the impedance meaning that directly measuring the current drawn during operation would be more accurate.

>>1404179
Yea I read somewhere else that they may be asking me directly for the address to avoid paying other fees to amazon, basically to cut out the middleman. It just seemed a bit sketchy.

I'm looking at a digital stereo amp, which has two outputs. Looking at this chart do you reckon RL is the load per output (2 8 ohm speakers) or the total load of the combined speakers? it doesn't really make a difference, just curious

>>1404204

>>1403611
Are you US based? I just searched the model number and found it for less on Amazon w/ prime shipping https://www.amazon.com/level-Spectrum-indicator-Amplifier-Adjustable/dp/B01N1RSRPM

It has all the necessary components. You'll have to assemble it yourself, but it should be plug-and-play once you're done.

>>1404073
Do you have any relevant skills, or are we starting from nothing here? You need to give give us at least a little detail if you really want help...

>>1404204
Per output would be normal.

I need something to
>raise the dc level of a differential signal
>turn the differential signal into singular signal (Vout=V+ - V-)

I know this screams of something like an instrumentation amplifier, but the slew rate is FUCKED since I need to measure square-ish signals

Anything else I can look into, since I don't need to amplify the signal, just do the dc offset and subtraction?

>>1404135
I'll be using a smaller FPGA from this series (that costs more like one fiddy) to add the funky command signals for "normal" LED matrix drivers to an SPI data stream, but I didn't have any QFN32 breakouts handy
I may end up using this level of chip to drive a 48x8 matrix with 10-bit PWM, if I can't find a better, cheaper driver chip

>>1404266
how fast are we talking, here? any opamp should be able to handle diff to single-ended conversion. maybe you just need a faster one

Currently fixing a friends audio mixer. The PSU seems to be broken. I replaced 2 bulged caps but that did not fix it. Now I found a dark, "burnt" spot on the underside of the PCB. On the other side are 2 diodes, 2 resistors and 2 of those orange round caps. I checked the diodes and one of them does not block the current in either direction, so I took that out.
How do I know which one I can use to replace it?
It says: "HER303 121V" and a logo that looks like "pec"

>>1404297
did you even try to google?
http://www.rectron.com/data_sheets/her301-308.pdf
https://www.digikey.com/product-detail/en/micro-commercial-co/HER303G-TP/HER303G-TPMSCT-ND/2650912

>>1404300
that does not really help me. those are just 2 of 5000 _different_ HER303 diodes, and i can not find any HER303s for sale in Germany. Is there not a way to find an "equivalent" that might be more available? I don't feel like waiting a month.

>>1404295
>how fast are we talking, here?
Ideally around 500kHz, at the moment I'm using a 200kHz signal and the slew rate (4V/us on my INA128) basically eats one fifth of my period, with the signal in range of 0 to 3 volts

Jacking it up to 500kHz would make it a triangle wave basically

>>1404307
Like the other anon said, opamps can be pretty fast these days.
You could also use a transformer.

>>1404302
>those are just 2 of 5000 _different_ HER303 diodes,

what? I looked at two more and they are the same as his first one. 5000 different???

show ONE that is much different from pic related. and as for germany, if they sell diodes (laugh) there, they sell one that is close enough.

when you say you are "fixing" an audio mixer, that would imply that you know a bit about electronics.

>>1404302
any diode in the 1N4003 to 1N4007 series would work, in a pinch. does the diode in question still read short both ways when removed from the circuit?

>>1404307
the LMH6611 would be overkill for that, I think, but there are economical and available opamps in the few tens of MHz GBW range that you could either use as a simple diff amp, or assemble three into an instrumentation amp if the offset voltage of a simple diff amp makes for too much inaccuracy for your application

Have there ever been attempts to make robots that move like humans (ie all movement being based on pulling mechanisms)? Once I got some more spare money I'd like to get into robotics

>>1404319
>>>/g/

>>1404318
>any diode in the 1N4003 to 1N4007 series would work, in a pinch.

are you comfortable replacing one rated for 3 amps average forward current with one rated 1 amp, especially one that burned out (if It did).

>>1404323
fug, I missed that part. engineer - coffee = catastrophe, every time

>>1404323
Also fast recovery diode vs. slowest possible turd.

>>1404209
Assume I have all required skills, I guess. I'm willing to work or find someone to help with this, no matter how complex it seems.

I'm looking for something for a motorcycle helmet that will simply give speed, fuel level, and pre-configured graphics (e.g. An X when fuel is low along with an exclamation point if tire pressure is bad). How do I get started

>>1404341
it seems safe to assume there isn't anything both high-powered and fast inside an audio mixer

>>1404358
>all required skills
kek, like how to divide a problem into its component parts?
to get started, get your HUD working on a small area of plain old LCD monitor, which is a straightforward lolduino/rasPi class of multiple data acquisition/display problem that's been done to death already, then worry about getting le movie meme graphics onto the inside of your helmet
in the meantime, just decide what you want to do, google it like the rest of us do when we're curious about something, and repeat until satisfied

>>1404380
> then worry about getting le movie meme graphics onto the inside of your helmet
You're correct, the actual display part is trivial. The hard part is getting them onto the inside of a helmet. I could probably (maybe) build my own helmet to get the parts to fit inside the helmet (or possibly embed them inside an existing helmet), but the main problem is the projection part. A good, nondistracting HUD requires infinite focus distance (like with pilot flight simulation HUDs), and I haven't been able to find a way to achieve that without making a helmet with a LOT of empty space in it.

>>1404405
the only thing I can think of is https://www.ti.com/dlp-technology/about/how-it-works.html and that fails to solve the infinite-focus "requirement"
other than that, optics ≠ electronics, maybe >>>/sci/ is better

I have a cd4017 decade counter
I only have 5 LEDs I want to send output to, so I figured I could just wire, say pin 0 and 5, 1 and 6, etc so that the counter would continue uninterupted

I'm wondering now, could it damage the IC if I send power to a pin that's supposed to be an output? do I need to use diodes to prevent damaging the thing?

>>1404457
>could it damage the IC if I send power to a pin that's supposed to be an output?
Yes. You have two practical options:
- combine the each pair of outputs with diodes
- truncate the counting sequence by connecting the 6th output to the reset input
Of these, the latter is simpler and more popular option.

Can anyone here recommend me a chinese EPROM programmer? I'm looking for something that can cover a good range of chips (EPROM, EEPROM, Flash). Currently need it for 27128's but i want to get something that won't just be for a single use case.

Are there any manufacturers that still make tunnel diodes (bonus if the prices are reasonable) or is the only choice to buy new old stock on ebay?

could someone glance a this assortment of components and tell me if they're all worth $35, or at least close?
I know buying them individually would be cheaper but I also don't really want to wait a a long time for shipping

>>1404484
you'd have enough "fun" tracking down the sellers and paying them each their shipping to make it worth your while to go for the kit. the box alone is worth at least $3, but let me call my expert on plastic boxes to be sure

>>1404295
Is lattice as much of a bitch to program as I've been told? I've done microsemi, xilinx and altera but no lattice.

>>1404489

Any of the larger eBay/Amazon sellers would have everything there or close to it. Everything there is exceedingly common. Except for the Uno, I'd be surprised if the guy putting these together is spending more than $4-5 in parts. You'd pay more than that since you aren't buying by the thousand, of course, but not all that much more.

>>1404522
nothing can be worse than altera

>>1404522
I can only speak to the iCE40 series rn. iCEcube the integrated environment is a hot mess, but Verilog is pretty much Verilog anywhere, and the chips are decent, if basic. if you don't like iCEcube, there's an open-source toolchain iCEstorm for some members of the iCE series

>>1404525
meh, Xilinx pissed me off before with their Platform Flash crap (before the whole rest of the volatile FPGA world settled on commodity flash and dragged them kicking and screaming along with) and still piss me off with the super-proprietary Platform Cable shenanigans. I'll take the $3 USB Blaster clone tyvm

>>1404266
Why not just use a discrete transistor based diff amp? Fast transistors with gain bandwidth products of several hundred MHz to over 1GHz are cheap. This amplifier isn't all that different to the differential amp in the front end of an op-amp the only real difference with the major difference being it lacks the power amplifier integrated in the output stage of most op-amps so it can't source much current. Don't try to drive low impedance loads with it. The reason it performs so much better than an op-amp is due to the lack of dominant pole compensation which makes an op-amp more stable but severely reduces the bandwidth. Another trade off is because it's all discrete components matching between components will be pretty poor.

Anyway this circuit seems to work okay in simulation though you might need to tweak it a bit for your specific application. I'd buffer the inputs and output too since I didn't take much care to make the inputs particularly high impedance nor the output particularly low impedance. It should work with DC as well as AC and seems to handle 500kHz sine and square waves alright.

I finally got my full-analog LED chaser/flasher thing done
all I need to do is add a second 555 to independently control the cd4017

This is the most complex thing I've done. I have no idea how I'm going to wrangle this clusterfuck and put it on a protoboard and have it still work. Any tips?

>>1404570
fuck the protoboard. order a pcb for $5+sh.

>>1404405
surely a parabolic semi-mirrored surface as the visor itself wouldn't be too bulky, it could be pretty small if you really jam the screen above your scalp.

>>1404524
>$4-5
Mate the breadboard alone is half of that. I'd rate it at closer to $10-15, but that's as someone who hasn't bought any of those modules before.

>>1404624
>Mate the breadboard alone is half of that.

I know these aren't fixed prices, but:

https://www.alibaba.com/product-detail/830-ponits-Big-Breadboard-MB-102_1880926751.html?spm=a2700.7724838.2017115.61.20f04c5eRndkhc

Injection-molded plastic and stamped sheet metal parts cost next to nothing in significant quantity. Period. While, yes, you'd pay more than that as a random buyer of a single piece, I still don't see that kit as being worth it. Realistically, the discretes are cheap to the point of being negligible, and the oddball breakouts, breadboards, and etcetera don't represent much value, either.

Don't get me wrong, you could do worse for $35, but I'm the kind of tightass who would rather spend another hour and a half browsing through the dark corners of the web than spend an extra 17 cents on a $17 part. That being said, I can't see that particular collection being worth more than $20-25, even considering the convenience.

>>1404570
if you're feeling masochistic, consider a spool of 30AWG wire wrapping wire

Would it be possible to build something like this (https://www.lightmodehelmets.com/neutron-s-kit/grid)) but for cheaper? $130 seems overpriced since I doubt this is his own tech. What's the underlying technlogy here?

>>1404660
It's electroluminescent tape.

https://www.kickstarter.com/projects/lightmode/illuminate-your-helmet-lightmode-kits-redesigned

The custom shaped tape itself is probably very cheap in volume (shops sell the stuff in many shapes already). Selling a complete set for much less than that would be hard though.

>>1404484
Looks pretty decent brah
I doubt you'll use everything, but by the time you get to the point you know what parts you need for a project you'll have outgrown the kit

Pull the trigger and by the thing

>>1404481
AFAIK the bulk of Germanium Power Devices (who actually made tunnel diodes) is now in bankruptcy.
M-pulse Microwave is still alive, but I'm not sure if they actually manufacture tunnel diodes or just rebrand/sell them. Then there are some companies which basically sell new old stock as their own "production".
Apparently making them is pretty diy-friendly, much more so than the other semiconductor devices.

Now I'm curious. Why do you need so many of them that a couple of random diodes from eBay is not enough?

FUCK
i bought the wrong mosfet
how do i turn 5V into 10V so i can turn it on?

>>1404702

simple transistor inverter. if inverting is a problem, use 2 transistors.

>>1404628
Holy shit, the individual markup is that high?

>>1404660
While he's probably using EL tape for that you could see about using an OLED panel instead. Not sure if you can cut them to custom sizes; whether they're a bunch of discrete LEDs or not, but it's something to look into because they'd be much easier to drive than EL strips. They're certainly newer to the market and subsequently harder to find however.

>>1404702
A small-signal BJT/FET and a 10V PSU?

How would you hook up this 16V1F supercapacitor pack?

>>1404790
With wires and connectors.

>>1404790
Well I'd start by tracing the circuit to see what it does, if you don't have the PCB's pinout. The circuitry present could be for balancing, over-voltage protection, over-current protection, or all three, but either way I'd just treat the whole assembly like a single 16V 1F supercap. What are you using it for?

>>1404794
To help my powersupply at start up of a 12v heavy fan. The start up current is quite high but once its spinning it takes only a fraction of that.

>>1404795
Not sure how well those things keep their charge, but either way you may want to add some sort of soft-start circuit to limit the capacitor charge current, for the well-being of both the caps and the PSU. Not sure how the capacitor board handles over-current, if at all. But at that point you're probably better off using a soft-start circuit for the fan itself.

If the caps don't stay charged you'll have to turn on the PSU to charge up the caps, then turn on the fan a little later. If it's an external PSU that you just leave plugged in with the fan itself off this should be fine since the PSU will remain on and the caps charged, but you might want some sort of relay+comparator in there to stop the fan from turning on before the caps have charged if you try to turn it on too soon.

If i have one of those big supercapacitators and charge it to say 5v, can it zap me? I would think no since 5v isnn't enough to tunnel through the body resistance but i am scared

>>1404835
>If i have one of those big supercapacitators and charge it to say 5v, can it zap me?

You can touch 5v. when you were a kid, didn't you ever touch a 9v battery to your tongue?

The issue with something like a super cap is that if it has enough capacity and you accidentally short it, you have a bit of an arc welder going, which anyone who's ever screwed up while working on a car can verify.

>>1404838
Can i legit weld with a capacitator? That would be pretty cool

>>1404835
There's a standard (SELV), which says that voltages under 48VDC can be considered safe and don't need extra insulation. 48V can give a minor shock though, particularly if your hands are wet. Another (less official) criterion is 30V and yet another 24V. You're unlikely to feel anything if you touch 30V or 24V with dry hands.
While one can speculate about sticking wires to open wounds and shit, the primary hazard here is probably the spark produced by the capacitor if you short it.

>>1404844
1. there are no potatoes in capacitor
2. https://www.youtube.com/watch?v=oCOD1NBFYmk

I got this badboy from ebay. Heats a 8mm rod just fine, but a 10mm rod pulls so much current from a 24V supply that it turns off.

How can i limit the current it would pull at bigger loads? I tried adding more chokes but it didnt help.

>>1404570
I made a diagram of the circuit to try and help me plan out the protoboard or begin designing the custom PCB

it's nowhere near as complex as I'd thought, all the jumpers just made it look a whole lot more intimidating than it is

>>1404570
I made a diagram of the circuit to try and help me plan out the protoboard or begin designing the custom PCB

it's nowhere near as complex as I'd thought, all the jumpers just made it look a whole lot more intimidating than it is.

What's the difference between a pad and a via really?

Will a via be covered by solder mask while the pad won't?

>>1404905
Pads are always on the outer layer of the board and not necessarily drilled through with SMT components, vias are interconnects between the layers and can go through a variety of layers not necessarily being exposed to the outermost layer. Vias should be covered with solder resist.

>>1404674
So in theory I could buy it from China? What do you mean by "complete set"? I can just use the EL tape and hook it up to the system I already have to control when it's turned on/off, no?

Is it possible to convert dBV or dBm to dBSPL?

The voltage across a speaker is obviously correlated to the loudness of the sound (as is by extension the power dissipated by it) but I'm just not sure exactly how they're correlated. I know distance from the speaker is a factor in loudness which might be what's complicating this. Let's assume I'm wearing headphones and the distance between the speaker and ear there is no loss over the small distance between them. In this case how many dBSPL is it per 1Vrms audio level?

>>1404945
Just making a guess here, but let's say the speaker works with a certain efficiency at turning power into sound pressure and then there's a certain damping factor from the distance, let's call this x. This would imply that the conversion is as simple as x*dbm = dbSPL, if this is true and you have access to a decibel meter (smartphone?) you could simply measure the sound pressure level at a couple of different output power levels and plot it in excel. With dbV you have to assume constant resistance as well and calculate the power from there which should be correlated with the square of the voltage, too tired to think about how it works with dbV

>>1404889

so the clock pin, 14, goes to ground?

Which is a better option and why?

Anyone know what this type of cable is named?

>>1405044
something like 6(AWG)/3(conductors) w/ground stranded copper

>>1404982
it's confusing, I should clear that up more.

>>1404674
>>1404706
$130 seems overpriced for electroluminescent tape in a nice-looking shape. If I can get run-of-the-mill EL tape off adafruit.com for 10 bucks, then surely I could get some tape that's in a cool design for not much more, no?

I want EL material, but I also want the Tron-esque meme design. Since cutting EL tape can be difficult and complicated (not to mention asymmetrical, in all probability), how could I achieve that EL tape design? Or perhaps a better question is how do I begin the process of ordering a Chinese knockoff?

>>1404945
speaker sensitivity is commonly expressed in dBSPL/W at 1m. ~80-85 is fairly common for bookshelf-type speakers
it's straightforward to go from dBm to dBW, thence to dBSPL at 1m, then subtract 3dB for every doubling of distance to the listener

>>1405089
none of your friends have laser cutters? you will be forever alone

>>1404912
>Vias should be covered with solder resist.
Does this also mean I can put silk screen text over a via? If they're not exposed metal

>>1405095
>laser cutters
Is it safe to laser cut EL tape? I don't see anything about someone successfully doing that when I google it.

>>1404945
The sensitivity of earphones is usually specified as xdBspl/1mW. Since sound pressure is a field quantity it follows voltage, not power. You can find the corresponding voltage via U=sqrt(P*R). If your headphone produces 100dBspl from 1mW and has a resistance of 40Ω it would need a voltage of 200mV. If you feed it 1V, the spl would increase by 20log(5)dB to about 114dBspl and the power would increase to 25mW. Please verify, I'm in UTC+2 and I'm getting tired..

Rate babby's first PCB design

>>1405133
What is dip? That's a single row.
Pick a net (usually ground) and flood fill the board with it. The copper gets etched from the board,not suck on according to your design, so less copper is waste because it uses more etchant and produces more waste. I have had cheap China boards where they do the flood fill themselves to save on chemicals apparently.which is cheeky if you really wanted them etched....

>>1405133
it'll be perfectly functional. pcbs are art though. with that in mind it's not very good. upsize your traces to use some more of that free space fix all the spacing and alignment. if you're ordering this board then add a ground pour just because it looks better.

pic related is a board i'm testing now. i think it's pretty.

>>1405044
>>1405023

I have a multispeed fan that quit working. How do I go about testing to see if the switch or motor is busted?

>>1405153
if you don't have a multimeter you could try opening it up and shorting the common and one of the speed taps on the switch with a screwdriver. depends if the mechanical design of the fan makes that possible.

>>1405156
I do have a multimeter. I was thinking if I connected it to the plug and switched between off and 1, 2, or 3, shouldn't I see a change in the resistance?

>>1405165
yes. you can also just use the continuity test mode, if your harbor freight special has that.

>>1405167
Rude. Its a Fluke asshole. Thanks for the advice though.

>>1405169
gib fluke

>>1405133
manufacturable/10
the second side's almost always free, personally I'd make a solid ground plane out of it and leave that much less to squirm around in (see LED- for example)
>needing a module for MAX485
are you sure Y/N
personally, I'd skip the daughterboards and just build everything onto the baseboard... but I'm not you

>>1405140
>flood fill themselves
WEW

>>1405142
it's kinda tryhard

>>1405011
That board looks like a half-bridge/H-bridge, possibly with inductive feedback circuitry, which is only useful if you want to run it in reverse or if it's a BLDC.

>>1405089
If you print out a template and cut to it with a craft knife it should be accurate enough. I wouldn't laser cut one though.

I don't uderstand the voltage splitters. so you have two resistors.
source goes on the left, ground on the right, and the device that wants lower voltage in the middle
But what is the point of wasting the energy into the ground? why not just use a stronger resistor and completely omit the second resistor and ground?

>>1405279
Well both are flawed, but say you're powering an IC that flashes an LED. When the LED is on it's drawing 25mA, and when the LED is off it's drawing only 5mA. A factor of 5 through a resistor means the voltage being dropped by the resistor has changed by a factor of 5 regardless of the value of the resistor. But with a voltage divider, as long as the current being grounded is significantly higher than the current being pulled out of the middle, the voltage will not fluctuate much. While sinking hundreds of mA to ground would be foolish, often we just want a voltage reference for a comparator/opamp, which has an input current of a few µA or possible much less. So we can use a 10k potentiometer (variable voltage divider) to sink less than 1mA to ground while the voltage at the output doesn't fluctuate. Picking a single resistor for something that only pulls 10nA would mean it has to be in the range of hundreds of MΩ, meaning any oils left by your fingers on the outside of the resistor could allow a path of least resistance.

But for power supplies to ICs and such it's common to use a zener diode or linear regulator instead, which allow a high-ish current to be drawn with low ripple voltage.

>>1405065

if you follow that line, it leads to ground.

>>1405089

side-emitting fiber-optic wire seems like a much better choice.
https://www.youtube.com/watch?v=eOEykSKhh6U
(Glow Laser Wire VS LED and Electroluminescent )

>>1405279

dunno where people get the idea that voltage dividers are suitable for powering gadgets; must be some new high school thing. no one uses that in the real world coz it's stupid.

anyway, theoretically speaking, the reason you use 2 resistors instead of just one is that it gives you a FIXED division ratio. if you use just one, then the voltage output will vary with the current consumed by the device. so imagine a radio, that as you turn the volume up, the power supply drops, and so the volume drops instead.

I'm not sure how to concisely describe this, so it's hard for me to search and get good results for it.

I want a bluetooth mic+speaker adapter that can link with a mobile device, desktop computer, etc. However, I don't want the speaker and mic to be build in; I just want the ports.

I found this (https://www.amazon.com/gp/product/B01FRY1GC4/ref=as_li_ss_tl?ie=UTF8&psc=1&linkCode=sl1&tag=dand0f-20&linkId=8f0ea050f531ee6d694770d9bc9ba864)), but I want something simpler (e.g. no "call" button), preferably just a device with a bluetooth sync button, an audio jack, and a mic jack.

Where can I find something like this?

>>1405279
That'll work fine if your load is always constant but it is not suitable for a load that changes. If the impedance of your load goes up the voltage across the load rises and if it goes down it drops.

It's also not suitable for high impedance loads. Say you wanna divide 12V down to 6V for as the reference for the non-inverting input of a single supply inverting amplifier. If you didn't use a voltage divider you'd need to match the input impedance of the op-amp which is usually very high. In the case of older op-amps with bipolar inputs you're looking at 10M or more. For FET input op-amps it can be gigaohms or teraohms. If you even managed to get resistor that high in value it'd be a.) super expensive, b.) extremely shit tolerance c.) very noisy (Johnson–Nyquist noise). There's no reason not to use a divider in this case with reasonably large resistors like 100k they will be burning next to no power so it's not a big deal.

>>1405291
Know how they get the copper inside the fibre?

>>1405293
I can't imagine there's much demand for such a device. We're on /diy/ after all, why not get something similar to what you posted and just solder the jacks in? With a new enclosure I guess.

>>1405301
It looks like they do indeed have copper in them, pic related, but I went to the site of the guys who make their fibre optics (the fibres are Corning's "fibrance") but there was nothing about having copper in the middle there at all.

Sorry for the stupid question, I'm not into electronics, just need to know this to repair something.
What are components 15 and 16 supposed to be? Also why are there two fuses one after another in the positive coming from the battery?
Thanks

>>1405494
15 is the switches in the steering wheel you press.
there are 4 in parallel because they are spread around the wheel.
14 is a clockspring, a coiled up wire used to connect the wheel switches. it needs a clockspring because the wheel turns around and around.

>>1405494
What is the provenance of that painting? Are you the author? It doesn't look like a technical document, more like a graphical artist's impression.

Whats the best way to cut holes into this plastic enclosure (2mm thick) so I can mount switches in it? I have some dremmel bits, but I don't have an actual dremmel, can I still use the bits in a regular powerdrill for this?

>>1405529
You never seen car schematics before?

>>1405530
Drill the rough shape out with a drill (lots of holes around the perimeter), cut the rest with side cutters, finish with a sharp knife.

>>1405530
Drill holes near corners, cut it close to the line with a coping saw, then finish with a flat file

>>1405537
I don't have a coping saw, I can borrow a hacksaw from a friend but the blade is quite tall so it's need a pretty bit hole to fit through. Is using the dremmel bits in a regular drill not an option?

>>1405524
Thanks so much man. Also, do you if it's common for relays to go bad?

>>1405554
took me a few minutes until i read it was for a horn!
its possible, as with anything.
beg borrow steal a multimeter and find out where the problem is instead of guessing if you cant afford to throw parts at it

>>1405533
Inspirational Collection Of Car Wiring Diagrams
Document Template Designs
BEST PLACE TO FIND DOCUMENT INSPIRATION
http://www.fixour.us/car-wiring-diagrams/
(first hit, many more template sources for the same image)
Your fake was way too childish and your follow-up post a dead giveaway.

I want to create a glove with buttons on the fingers. When pressed, I want to send which button combinations are being pressed to an Arduino which will transfer this info over Bluetooth.

However, a setup like this (https://blog.adafruit.com/2009/10/20/example-code-for-multi-button-checker-with-debouncing/)) implanted in the glove is way too bulky due to the breadboard.

Do you any of you have an idea of how I might be able to achieve this end goal without using such a large breadboard?

>>1405623
don't use a breadboard? Breadboards are just for prototyping. Depending on how much you care about the project, the final design should be on a pcb. Then, given how simple the hardware of your project is, you can make it as small as you want. The limiting factors will be the buttons themselves (which you can find some pretty damn small ones), and the arduino (again if you go offroad there are some very small arduino clones).

>>1405623
just get a smaller breadboard and solder directly to it, this is only 5 x 3.5cm, it should fit on the back of the glove like a powerglove
https://www.mouser.co.uk/ProductDetail/Chip-Quik/SBB170?qs=sGAEpiMZZMtgbBHFKsFQgqvxd2XGUZHj2DNQLeQqrWPJVwMjmRByFw%3d%3d

You might want to look a a smaller arduino too, e.g. a D1 Mini
https://wiki.wemos.cc/products:d1:d1_mini

>>1405669
D1 doesn't do bluetooth by itself
there are, however, some nearly as small ESP32 boards that might fit the bill. this one's the same size but a little more expensive https://www.aliexpress.com/item/32-bit-fisting-glove-bluetooth-wifi/32819107932.html

I'm trying to fix my bass amp and I think i've narrowed the problem down to a diode in the power supply (reading 0.5MOhm in one direction and 0.8 MOhm in the opposite).
I can just about read the number 1N 3448 0203 on it but can't find this in particular. The best i have found is a non-datasheet on 1N3448 (https://www.semicon-data.com/diode/dz/1n/1N3448.html)) but that is all there is for that number, all it gives me is a Uz value of 15V and N of 2W, neither of which I understand what they stand for.
How do I figure out a replacement for this? I have the following diodes from a pack :1N4148 1N4007 1N5819 1N5399 1N5408 1N5822 FR107 FR207
are any of those suitable?

>>1405734
it's a Zener diode, 15V, (according to the non-datasheet) 2W or (according to other sources) 1W
what makes you think that part went bad? what voltage were you reading across it when powered up?
>testing diodes with an ohmmeter
just, no. your meter will not usefully test a 15V Zener diode

Can you use a regular multimeter to test an electric fence?

>>1405769
sure, as long as it's rated for the volts/amps (it probably is).

>>1405769
>>1405777
http://lmgtfy.com/?q=voltage+of+electric+fence
hahaha no

>>1405769
actually, it's more reliable to just whip it out and pee on it

>>1405769
>>1405777
>>1405782
>>1405790

>>1405791
https://en.wikipedia.org/wiki/Wire_of_Death

>>1405792
Haha dumb belgs. Just bring a wire with you and short the fence into the ground.

>>1405824
Actually you could just bring bulk cutters with insulated handles and cut the wire

Can't believe 2000 people were so stupid to not understand how easy is to fuck up the current flow.

Just getting started which multimeter should i buy anons?

>>1405827
>His electric fence does not detect cutting and shorting attempts
Those guys were already taking the rather high risk of getting shot, so it shouldn't really come as a surprise that they weren't the smartest or the most risk-averse guys in existence. Also, the dangers of electricity weren't as well known a century ago than they're now.

I'm researching using a 3 phase dvd spindle motor as a rotary encoder. Is it possible to have switchable cogging so that the motor freewheels, and acts as a detent?

>>1405845
That ANENG that EEVblog showed off once looks pretty decent for the price.

>>1405685
>32-bit-fisting-glove

>>1405855
>I'm researching using a 3 phase dvd spindle motor as a rotary encoder. Is it possible to have switchable cogging so that the motor freewheels, and acts as a detent?

If you are saying that you want the motor to act as a generator providing pulses like an encoder, then yes, it can do that.

If you are asking if the motor will lock into position if you apply dc power to one or more phases, then yes, it will do that.

The overall scheme looks like it will be very hard to implement if you want it to work from a dead start to a dead powered stop, but it might possibly work.

>>1405901
>If you are asking if the motor will lock into position if you apply dc power to one or more phases, then yes, it will do that.

and it might actually have enough detent torque when unpowered to do what you want. In general, BDCMs are designed to minimize cogging or detent torque, but the one you have might be ok without having to apply DC to a winding.

>>1405903
Thanks. all i really want is enough torque to stay in position. the spindle motor i bought is too large for the project though.
the smaller motors are all steppers. can they be used as an encoder?
ali express com item/New-Wholesale-5pcs-lot-3-5V-Micro-Stepping-Motor-HOT-2-Phase-4-Wire-20-Ohm/32840467038.html

>>1405907

They are very similar to BDCMs in that they have permanent magnet rotors so they will generate waveforms if you spin the rotor, and they will lock into position when unpowered. PC fan motors that I've looked at also have pretty decent detent torque that snaps the rotor to a position when they are turned off.

The main difference, which you probably already know, is that an encoder provides a strong signal at any speed, but a motor acting as a generator will provide a signal proportional to speed.

And, if you amplify that signal to make it useful, you will have to deal with the mess that is generated when the shaft stops, and bounces back and forth at the final detent position, which will give you a decaying sinusoid that reverses polarity when the rotor reverses as it settles into position, which means that you will have a decent tachometer at speed, but at very low speeds you won't have a very good encoder.

>>1405917
I've seen a few youtube videos on hard disk motors, some need a quick flick to register, yet others work quite well at low speeds. I think the steppers seem to be more accurate at low speed. I saw one post referring to converting a 3 phase motor into a selsyn by feeding one phase a khz square wave. would you recommend the selsyn approach or would the programming be too difficult to manage?

>>1405924
>I saw one post referring to converting a 3 phase motor into a selsyn by feeding one phase a khz square wave. would you recommend the selsyn approach or would the programming be too difficult to manage?

You're out of my league now. Maybe someone else can answer this.

>>1405924
>selsyn
Oh I saw that once in an AvE vid.

Now assuming you'll be using the encoders at low speed you probably can't rely on the voltage swinging as you rotate them since this voltage is proportional to the speed as the other anon said. This means you'll either need to run DC through the encoder in such a way that it doesn't torque the rotor (has to be via very low current unless each winding is only on one side of the rotor) or trying some sort of AC method. Now I'm not sure how they made a synchro out of a BLDC/stepper because they don't have a rotor winding, so I'm guessing they put the power into one of the three windings and got feedback out the other two. A similar method would be to put low-power or deliberately asynchronous AC into the three windings such that the rotor feels no torque and measure the current though the three coils, which should change slightly depending on the orientation of the permanent magnet. The DC method with single-sided coils would be similar, but easier since you'd just be measuring DC amplitude and it would be easier to filter out noise from the rotor itself moving.

But all-in-all it looks like the synchro method would be far more reliable, provided you can get it to not torque the rotor. Mini steppers will have a boatload of stator windings and I'm not sure how they connect, but I think that will give you more freedom with the above methods than less.

>>1404484
Depends on the quality of the components but from my point of view 35 bucks is an okay price for the kit. You will spend considerably more buying them at a piece.

>>1403081

Any good sources on learning RF electronics? Seems to be a requirement for all most electronics jobs but my uni doesn't specifically have it.

>>1405529
>>1405584
> t. I built my own computer and think I'm an engineer

>>1405970
Does it have a physics program? RF is mostly electromagnetics and waves

>>1405970
>>1405989
the physics classes are generally just theorical and do not go into actual engineering/construction.

i would recomend this "aplication note" by jim williams as a guide in good practices for RF, read the rest of the papers he did for analog technologies they are classics in the field, RF its not really that difficult:

http://www.analog.com/media/en/technical-documentation/application-notes/an47fa.pdf

>>1405997

not the guys you are replying to, but thanks for that pdf link. tons of good information there.

>>1405970
VHF intro course prof recommended David M. Pozar Microwave Engineering and Reinhold Ludwig RF circuit design, found both on google as PDFs, I mostly skimmed through them as I relied on other material but they seem solid

>>1405641
Not the anon you're replying to, but how do I go about getting a PCB made? What's the best/free-est software for a beginner to use and how long does it usually take to produce once I've designed it?

>>1403081
Hey guys, I could really use your help. I'm building a home-made walkie talkie and I need your opinion on this circuit. Is there anything wrong with it? Could it use any improvements? Sorry but I'm new to this.

So I bought a cheap audio switcher from ebay. It has 2 inputs and one output. Problem is, when there are two active inputs, I get annoying static on the active one. I'm guessing it's because of the common ground? What can I do to remedy this?

>>1406098

the stickie (top post) list the software, whereas youtube has 1 million how-to videos. which is good, coz seeing is better than reading.

>>1406106

forget the tuned LC circuit crap. this is 2018, you want digital frequency synthesis.
you know you can get a ton of FRS walkies already made for about $3 each at the thrift store.
make sure you get one with replaceable cells so you dont have to spring $25 for a new custom battery.
nice thing is they're all compatible, so a Uniden unit will talk to a Motorola or a Baofeng.

>>1406132
replace the switch with a
"2 Position 3PDT Toggle Switch" and wire in both the common grounds to the third switch pole.

>>1406132

first thing is to check your theory, so it's no longer a theory but a fact. disconnect the unused input and see what happens.

if it works ok, then you might consider replacing the 2-way switch with a 3-way.

otherwise, since this should never happen, check your sources. maybe plug them into different wall outlets.

>>1406136
I know there are cheap walkies but I want to build one just to learn more about electronics and for the fun of it, not for the actual use of the walkie talkie. As for the digital frequency synthesis, I've never heard of it before; Can you elaborate?

>>1406137
>>1406142
Yeah, unplugging one source gets rid of the static. As for replacing the dpdt switch with a 3pdt one, it would probably be easier to build the whole thing from scratch or to add a spdt switch for the ground. Couldn't I just add some resistors for the grounds?

>Couldn't I just add some resistors for the grounds?
that'll screw up the impedance. it looks like you'll have to replace one of the input jacks with an insulated type to prevent it earthing through the case, if you install the switch.
or take the innards out and put them in a plastic case.

>>1406157
The switch looks like that.

>>1406151
we have this question every week
I forgot solution 3: don't plug speaker/headphone outputs into line inputs, ever

How can I convert an .epw footprint file into something KiCad will recognise?
I'm trying to import/convert the footprints for these two components:
https://uk.rs-online.com/web/p/products/8154871/
https://uk.rs-online.com/web/p/products/8155127/

>>1406236
looks like you don't, as their part wizard doesn't seem to support KiCAD (yet?)
if you can find it in Ultra Librarian, that does export to KiCAD
if all else fails you could just draw a quick and dirty footprint in the footprint editor, which takes maybe a couple of minutes once you get used to it
you may need to draw a symbol too, which you could just copy from one of the existing Form C relays

>>1406250
Awesome, they do have them - thanks anon

>>1406146
Think they use PLLs.

>>1406146
as anon said, basically PLLs, with fractional dividers in their loops. two commonly available ICs representative of the genre:
AD9850 (outputs sine waves via LUT/DAC, can use directly as LO) http://www.analog.com/media/en/technical-documentation/data-sheets/AD9850.pdf
Si5351 (outputs square waves, not greatest LO but fine for switching detectors and other special applications) https://www.silabs.com/documents/public/data-sheets/Si5351-B.pdf
also consider visiting the ham radio general for extra tips
finally, in some jurisdictions, it's my understanding you can't transmit using home-built equipment in certain bands, and I think the FRS/GMRS portion of VHF is one of them
>0.22pF
kek

>>1406265
also, I've had better than zero luck just searching the web for footprints, though I'll usually make the symbol myself if it's not some 440-ball monstrosity. the footprint editor thankfully has an array generator so that large, complex BGAs and other regular package footprints can be created easily/quickly

Anyone here an EE? Can someone please explain to me my SATA power interface has so many lanes (not sure if that's the proper term)? Why is there 3x +12,+3.3,+5V?
Thanks.

I bought a bunch of shit on aliexpress to make my own manual robot arm.
Assuming I can get wire, batteries, dowels and cable ties from home depot, am I forgetting anything important?
Input would be nice as well. Total price was $80

AVR Programmer
3 x Atmega328P
2 x set of assorted pressure sensors 0 - 20Kg
2 x 12V geared 60rpm dc motor (brushed)
2 x 12V geared 120rpm dc motor (brushed)
and assorted diodes, resistors, capacitors, transistors and mosfets

>>1406372
>am I forgetting anything important?

feedback. how will you use pressure sensors to control position, speed, etc.

>>1406369
more surface area more better; more wire more amps

>>1406393
>more surface area more better

that can't be the whole story, because there would be more area if they had one huge pad instead of three separated ones.

>>1406369
Instead of running a single larger gauge wire they use several smaller ones due to size constraints, look at a moden CPU pinout sometime and notice how most are for power
>>1405970
Forgot to mention, the same VHF intro prof had a bachelor's thesis done exploring the use of breadboards in RF prototyping, IIRC above ~150MHz you can't build a good discrete amplifier on a breadboard

>>1406393
>>1406397
interesting. thanks.
>>1406396
Yeah, that's what threw me off as well. If you look at SATA communication interface, it has 1/3 of the pins. It's weird that a SATA power connector has so many sections.
>>1406372
>make my own manual robot
are you following any online guide? can you link it?

>>1406391
well, it's going to be manual so it's not going to be very smart, I assume without rotational feedback it will move slow under load and fast with no load?
What's the cheapest parts I'd need to make an encoder?
I was going to use the pressure sensors between some springs on the hinges just to make sure it doesn't accidentally break itself by turning the arm in on itself. pic related

>>1406400
no guide, just wanted to see how good of an arm I could make for under $100 so I'm kinda just muddling through.

>>1406412
>What's the cheapest parts I'd need to make an encoder?

potentiometer

>>1406396
Some of the pins are for soft-start. And also making all the pins the same width and pitch kicks ass for manufacturing.

>>1406417
wouldn't a binary rotation encoder be easier to read with a chip?
if not then how could I hook up a potentiometer? would I attach it to the motor or to the arm hinge?
most of the ones I am seeing just look like knobs.

>>1406427
>Some of the pins are for soft-start.

If you look at the original image all the grounds are long so they engage first (soft start???). And how exactly do you think the contacts are manufactured, so that separating each one into three makes it "kick ass"? You could still have one huge pad for each type, no matter how many wires are ultimately soldered to it.

>>1406429
>binary rotation encoder

I was replying to the request for "cheapest".

As for your other questions, attaching and using a potentiometer is not very much different from a rotary encoder. they both rotate, and one provides an analog voltage while the other provides digital

>>1406431
I think the chip I bought has analog pins available so I could actually do that.
I'm having a hard time imagining how to stick one those knobs into the arm though.

bought some potentiometers and some photoresistors to try the binary encoder if I can't get the potentiometer to work well, either way feedback should be good now yeah? Am I missing anything else big?

>>1406429
>wouldn't a binary rotation encoder be easier to read with a chip?

Not if it has an ADC. If it does, it's just a matter of taking a sample and using the value pretty much as-is.

If you use a rotary encoder, you need either a separate chip to keep track of movement between checks (most encoders only output a quadrature signal; they don't give you their position directly.), or you need to constantly be checking for transitions on the outputs of the encoder. The latter can eat up a lot of CPU time if you're using a small controller.

>>1406436
Use phototransistors or photodiodes, not LDRs. You can buy shaft-mount encoders.

>>1406441
But microcontrollers only have one or two ADCs that they multiplex. You can probably get away with it for something this slow but it's something to take into account.

>>1406442
thanks for the tip, I didn't know they had anything but LDRs.
>>1406441
I didn't consider that but I ordered 3 chips so I could just dedicate one or two of them to the checks.

Thanks guys I'm learning a lot.

>>1406447
>I didn't consider that but I ordered 3 chips so I could just dedicate one or two of them to the checks.

That kind of ends up being how you have to do it. The problem isn't just that it eats up CPU time, the problem is that your _entire_ main loop has to run within the time between transitions at the maximum rotational velocity you're expecting, or you have to code it such that you ditch the main program in order to check on the encoder every so often. Ends up not really being feasible if you want the controller to do much of anything other than running the servo loop.

Might be some chip-specific workarounds (or built-in considerations; the Xmega chips have built-in quadrature decoders, for example), but for the $1 AVRs or PICs, easier to just have a dedicated chip counting edges.

>>1406455
>easier to just have a dedicated chip counting edges.

AVR interrupts can count edges without much overhead. and if he goes with a potentiometer he can start an ADC conversion and let it trigger an interrupt when it's done.

Extra chips are nice but not necessarily needed. I don't think he's building a high performance device if his original plan was to run it manually, whatever that means; one axis ok, but two or more operated manually would be interesting.

>>1406463
manually as in pushing a button to make the arm extend and another to rotate.
as opposed to telling it a position to adjust to.

like a crane game instead of a cnc machine.

>>1406475
>manually as in pushing a button to make the arm extend and another to rotate.

In that case I don't see any need for encoders, so your original idea looks good. You might not need the pressure sensors either, if your mechanical parts are strong enough to stall the motors at the end of travel.

Hi friends
Looking to get into ILDA animation on a budget, I've got a lot, a lot of willingness but not heaps of money.
I'm wanting to do some proper shows, so I would like at least 30k pps and 500mw, so I was looking at purchasing some housing like
https://www.ebay.com.au/itm/400mW-RGB-DMX-8CH-Laser-Light-DJ-Party-Show-Projector-Stage-Lighting-DM-RGB400/273000087662?epid=2011224055&hash=item3f9012c06e:g:VKgAAOSw-09aRLb9
and retrofitting it with
https://www.aliexpress.com/item/RGB-900mw-1W-Laser-Module-G100F-B500F-RL300-Combined-Laser-Module/32622632544.html?spm=2114.search0104.0.0.23283976pJzr67 and https://www.aliexpress.com/item/LH-LASER-30Kpps-Galvo-Scanning-system-for-Laser-light-30K-galvo-scanner-Laser-light-scanner-30Kpps/32725744714.html?spm=2114.search0104.3.70.41b06797qfplaP&ws_ab_test=searchweb0_0,searchweb201602_1_10152_10065_10151_10344_10068_10130_10342_10547_10343_10340_5722611_10341_10548_10696_5722911_5722811_10084_5722711_10083_10618_10307_10131_10132_10133_5723111_10059_100031_10103_10624_10623_10622_10621_10620_5722511,searchweb201603_32,ppcSwitch_5&algo_expid=41ed59f5-7c3b-44d1-be9f-e6122ba3e513-10&algo_pvid=41ed59f5-7c3b-44d1-be9f-e6122ba3e513&priceBeautifyAB=0
(sorry for phat links)

Anyone have any experience with anything like this? Thoughts and concerns?
Also any hints on whether to go analog or ttl?

How's my board layout looking? This is my first attempt, any suggestions/tips on how to improve it?

>>1406556
Here's the schematic for the circuit I'm creating, it basically uses 3 buttons to control whether the motors both move in one direction, both move in the other direction, or both move in opposite directions

>>1406556
I'm no expert but those over/under vias don't look too conventional. If your software can do this try showing it with the actual trace widths, should give a better idea of what it looks like.

>>1406560
Oh, and ground planes.

>>1406556
It looks autorouted.
Since the currents are not tiny and there's more than enough space, use much thicker wire and bigger vias. Minimize the number of vias; you can get rid of 3 easily and the rest with some effort. There seems to be a pointless wire loop too. You're also supposed to avoid < 90 degree corners like in bottom left corner on red layer, but this doesn't really matter on a one-off low speed board.

>>1406560
I'm using KiCad/Pcbnew, there's a "show vias in fill mode" which makes them look like this, if that's what you mean?

>>1406567
Yeah I used the default 0.25mm trace/0.60mm via width and I was just thinking that bigger would probably be better; I'm inferring from >>1405140 that it doesn't cost any more to have wider traces. Thanks for the advice - I mostly manually routed it, but I did take some of the suggestions from the software. As a result there are fewer traces on the board than in the schematic, but I have to admit it does make it harder to understand so I might ignore it completely and try to mimic the schematic more closely.

Can anyone recommend any good places to get the board printed? I'm in the U.K. if that makes any difference.

>>1406574
>it doesn't cost any more to have wider traces.
True. On the other hand, holes and particularly tiny holes can increase the price. (Not in small runs though, unless you want really small holes.)

>>1405140
>they do the flood fill themselves
What is this shit factory so that I can avoid it?

>>1406590
Ok here's a second attempt without any vias, using 1mm track width - should I up it further? I don't know how much current the motors use, but the power supply I'm using is 500mA and they work fine.

>>1406618

Asked this in the RC General as well >>1406629
So which one would you recommend?

>>1406372
figuring out how you fucked up is part of the fun

>>1406430
it kicks ass because you can use the same stamping dies, mold inserts, EDM electrodes, etc. for every position
>one huge pad
which requires special consideration in paste application and reflow. in real volume manufacturing, consistency is king

>>1406492
thank you for your tracking info

>>1406556
3/10
first, traces are way too thin for a high-current load. there are trace current calculators out there that will tell you the max current given a width, copper thickness, and allowable temperature rise (keep it to about 10°C). in fact, one is included with KiCAD
also remember to use through-hole component pads as vias when needed, and feel free to route thru them
U3 (U? it's a connector, call it J or CN) pins 3 and 5 could be routed outside the pins (Pic related, blue). now you can route other traces topside, without crossovers (light blue)
move D1 west of the M1 connector, now you can route it all topside, without crossovers (sea green)
wrap the reversing switch around M2 pin 2 and wrap the common pin of "U2" through the big space (light yellow)
boom, now it's a single-sided layout, which you can run single-sided to save cost, or duplicate onto both sides of the board for added current capacity
another note: if you can't meet some criterion any other way, and you can do so, it's often worth rearranging your input connector pinout to accommodate your layout

>>1406567
>he believes the sharp corners meme
wew

>>1406620
better, but that acute angle near "U3" pin 3 and "U1" pin 4 is bad practice. it can trap etchant on particularly low-tech processes. for the first, just route thru the pad, or route +12V between the coil terminals as in Pic related. for the second, use a T junction

>>1406674
>wews at sharp corners meme
>recommends to remove the sharp corner
wew

>>1406556

you can wire that shit point-to-point in half the time it takes to merely draw the PCB. never mind actually fabricating it or ordering it.

>>1406717
It's one of those things everyone should do at least once

>>1406717
I just threw up on my laptop

>>1406719

whaa, you don't like blob-o-lanche technique?

which brake pad buy???

Not completely new, but new enough

How difficult would it be to emulate the Gecho Loop Synth, I know enough to create mess and it looks like fun

>just spent 6 days tracking down a shorted smd capacitor
How to get over the shame?

>>1403148
On my rigol In trigger>menu>set-up>sensitivity set it to something much higher like 1.00div (default was 0.38div). Will probably say something different on yours.

Auto button often resets this in my case, its trying to trigger on the noise atm.

>>1406719
Nothing wrong with one-offs or prototyping like this. Breadboards are not very reliable and PCB's cost time and money and you only get one shot unless you place lots of spare pads everywhere.

Has anybody found any interesting deals from aliexpress's mid-year sale?
Wouldn't mind 50% off a 300€ scope...

This is my latest attempt, do the bits circled in blue matter, or do the traces and the filled areas all go under the same mask when the board is being manufactured?

>>1406717
Seeing as the cheapest I've found for a 90x60mm board so far is ~£50, I'm starting to reconsider that option; I'm starting to wish I had picked up that "make your own PCBs the ghetto way" kit at Maplin in the closing-down-forever sale. That said if anyone can recommend any reputable 3rd world manufacturers I'm all ears.

>>1406955
i ordered 5 100x77mm boards for ~$10 last month.

those bits will have the solder etched away. you should fill them in, if only to satisfy everyone's autism.

>>1406961
forgot to mention, the manufacturer was elecrow. the boards are shit quality. small pads love to lift, and their HASL is incredibly paste-averse, but i've had no issue working around that.

>>1406961
Well the reason I added trace across the left edge of the triangle was because the notches down the side of were triggering my autism.

I'm a bit confused, what exactly is the difference between the traces and the filled areas? I thought both just copper.

>>1406962
Thanks for the recommendation, I found another place that does them around the same price so I'll shop around a bit more (I'm never going to use the other 4 pieces, but the other place does 5 as a minimum too - I feel a bit bad if they just go to waste though).

>>1406971
sorry, i thought at a glance that the darker red was the bottom layer in your software. there is no difference between copper in traces and fills.

>>1406976
Ah ok cool, thanks again anon.

>>1406955
8/10, you don't have to run traces under your zone fills, but sometimes it's the easiest way to get connected to them. if this is a low voltage device, 10 mil spacing should be plenty, you don't have to have equal spacing between traces. note it will take a bit longer to solder the trace-reinforced pads on Net-(M1-Pad2) and you could maybe conceivably overcook a component if you're using unleaded solder
>board houses
did you check pcbshopper?
I just ordered some pcbway boards (2-sided) and they're alright. have hand-soldered SOT-23 and 0603 to them with no problem. they've got a flat rate deal for most boards under 10x10cm. you should be able to get 10 of those boards for like 25 burgerbux at most, with express delivery

>>1406975
>never gonna use
not for their intended purpose, perhaps, but if you get into SMD you will probably find those boards are useful as spacers for solder paste application

What would be the cheapest way to measure and record time intervals between signals with a 1-3ns resolution? I know even small computers have a clock frequency in the range of multiple GHz, but I've no idea how relevant this is in measuring external signals. On the other scale, I know logic analysers and oscilloscopes will do the trick, but they're not too flexible and certainly not cheap, especially since I just want a single digital channel. The aim is to measure the speed of light as radio is emitted and received by a beacon, or inversely to measure the distance from the beacon to the reflector.

I'm going to read this book to brush up on my fundamentals, and because it apparently takes you through how to implement a MIPS processor in verilog.
If you have other books you like in the same vein, speak now or forever hold your peace.

>>1407046
Wow, money is that guy's middle name!

>>1407048
Textbook authors aren't even trying to hide it anymore.

Why are there no Microchip-branded Atmega328p on Aliexpress? Could it be... that they're all clones?

>>1407057
Don't microchip still make them with Atmel logos on them anyways? They're probably selling old stock so you should see Atmel logos on them regardless of when they were made. No clue how you'd actually tell whether it's a knockoff or not.

>>1407046
Pic related is not so much a walkthrough as a survey of architectural components and motifs, but it's good if you know Verilog and digital logic design already. it helped me design a 16-bit pipelined RISC CPU that I have yet to put into some silicon (but that shiny Artix-7 board on my desk awaits...)
anyway I hear the complementary volume by the same authors, Computer Organization & Design, is a bit more hands-on and ISA-specific (ARM, MIPS, RISC-V editions)

>>1407057
>he doesn't know about the GD32F103C8T6

>>1407060
tangentially related observation: weird af to see "adesto" where one would ordinarily expect to see "ATMEL" on a flash chip. all the rest of the marking on the chip is identical to Atmel-style marking, font, size, spacing, etc. they must have spun them off as part of the Microchip merger? idunno

>>1407045
a fast FPGA could do it, look into the Lattice Mach-XO2 line which has on-board PLLs for clock multiplication (just like those GHz CPUs), and deserializers built into some of the IO blocks. DigiKey has the LCMXO2-1200HC-6SG32I in a not-too-difficult-to-use 32QFN for bout fi senty

>>1407066
And I suppose it would be relatively trivial to use one as a stopwatch to time the few ns to µs between two pulses?

Not too sure what I'm looking for on the vast datasheet, though
>DDRX4 Serial Input Data Speed = 756-524Mb/s
looks somewhat promising.

Trying to fix a couple of CRT TVs, one of which I can't find a circuit diagram for.

Model is RCA F19426.

There's a blown 470uF 25V cap and I'm wondering if I can sub one of these 330uF, 50V caps I have on hand instead of ordering and waiting for a new 470 but I don't know what that part does. Does anyone know a good resource for finding TV diagrams?

The other model I'm working on is a Zenith C27A24T for which I found full diagrams easily through Google.

>>1407085
I also have a few 100uF and 22uF's I could put in parallel, I guess.

>>1407082
not really trivial, there's a learning curve to climb before you can write your ranger, but it's a way to do it such that you won't have to fight with non-deterministic caches and slow I/Os and so on. it's the first way that comes to my mind when high-speed signals need to be dealt with. how many are you building and how much money/time is it worth to do this?
as a rough sketch of the design, you'd set up the PLL to multiply some input clock to a usable frequency for the counter and gearbox. you'd write a state machine that is in a particular state whenever the counter should be running (according to your input pulses or the gearbox output). that state output would enable the counter of however many bits. you'd write some way of getting the counter values (and first/last gearbox values, for finest accuracy) out of the chip, maybe a UART-like thing, or a 7-segment driver, or just raw binary if you prefer

>>1407085
elektrotanya is a good one, they appear to have that very manual. iirc it's a minor hassle to get docs from them
https://elektrotanya.com/rca_bd20tf10_f19426_f19436_f20tf10_chassis_tx_td809.rar/download.html

>>1407045
The proper search term would be "time to digital converter" or "TDC". You can make one quite easily from an integrator, some switches and an ADC or you can get a ready-made IC. TI's TDC7200 is one such IC and it has a resolution of 55ps (which isn't even particularly impressive).

But yeah, if you're happy with 3ns resolution, timers made from FPGAs should be fast enough.

>>1407099
wow. I did not know such things existed. very nice, and only bout tree fiddy
you, anon, are awes/ohm/

>>1407092
>elektrotanya
Excellent, thank you very much. A helpful site named after best girl.

Can I get a ton of cheap usb powered fans, put into a usbhub which is connected to an usb battery? I don't know enough how it works.

Want to do something quick and cheap for the car-cage to my dogs.

>>1407116
USB specs say 500mA so I doubt you'll run many fans on that, the cigarette lighter should provide higher current if you can find compatible fans

>>1407116
>>1407120
USB2 is 500mA but USB3 can go much higher than that.

>>1407120
>if you can find compatible fans

shouldnt be hard as all PC case fans are 12V. dude could probably take a couple of fans out of his PC given that they now put too many fans in them.

cigarette lighter is a much better idea than USB coz it provides a lot of juice and never runs out.

just gotta be careful to shield the fans coz a strong one can break a fingernail clean off if you stick you fingers near them.

>>1407120
>>1407123
>>1407205
Yeah ok ciglighter seems like a better idea.
Does it stay on when the car is off though? Not in accessory mode.

I got a lot of old pc fans, I can diy it all together, just want to make sure they can run at 100% for at least a few hours.

>>1407215
Ok. So I should simply be able to get some cheap pc fans, hook them all together with the usual y-splitters and directly into a ciglighter adapter, 12v.

Right?

It's that the selection available for what I want is real shitty and "expensive" for little benefit.

So if that works I have a nice design for it to change between the cars that I can post in diy thread.

>>1407092
I'd be using the data in some calculations, for which I'd probably run it into a microcontroller set up to provide a data feed into my computer, or otherwise some sort of simple database.

>>1407099
55ps? That's absolutely incredible. But I can't figure out how to search for them on digikey or the usual suspects. The TDC-GP22 is an ubiquitous ~$4 IC that has a TDC but also has a bunch of "ultrasonic flow sensing" stuff in it. It's $11 to get it on a breakout board from ali, which isn't bad all things considered, and there's a $25 one with an extra IC specially for laser range finding.

But using such a resolution with such a low frequency wave (ultrasonic) is a little baffling. How much would the time resolution of the TDC be effected by using a lower frequency wave? I assumed I'd have to use 2.4GHz or so but maybe I can get away with something that propagates easier and can be emitted easier. I suppose it depends on the signal:noise ratio of the demodulator. If the wave has a period of 10ns and you're trying to get a 100ps resolution then you're just looking at the low-amplitude very front of the wave, not it's peak. Of course accounting for a fixed time delay is pretty trivial also.

>>1407242
I have serious doubts about the ability of PC fans to keep your doggos cool while you keep them imprisoned in your car for extended periods.

>>1407251
It just needs to get the circulation going, it becomes a deadzone with the cars AC.

But yes, it won't cool them down 10-15c, not even 5c I bet.

Are there perhaps larger fans being able to move huge amounts of air at 12v?
I'm putting together 9x 120mm fans á 100cfm^3 each, so like 700-800cfm^3 I guess.

>>1407251
To elaborate a bit further, fan's don't magically cool air. There's no point in just recirculating already hot air, what you need to do is to create a draft. That means you need somewhere for cool air to get into the car, and somewhere else for the hot air to get out, and it's probably best to mount the fans at one or the other locations to push/pull air through the car.

>>1407253
You can get larger case fans (200-230mm) but they tend to spin slower so I'm not sure you'd be moving significantly more air.

I don't know if the cigarette lighter provides power when the car is off, I would have thought not but I could be wrong.

>>1407254
If cars had a vent in the rear it'd be much better circulation in cars.

I thought of doing a push pull confuguration but it won't work that much better I reckon. Definitely if you can have the rear open when you're parked.

But I will make it easy to switch between our two cars, placing it on top to push down the car AC air that comes in from an angle from the bottom.

>>1407257
Yeah for pc fans larger usually means same amount of air moved but more quiet.
I was thinking deskfan/towerfan tier.

I wanted usb for the thought of using it with the car off. But it will have to do for now.

>buy chinkshit 315mhz pt2272 key fob and receiver
>cut antenna to 23.79mm (actually 9-3/8") and coil it
>solder the code pads to be the same on the rx and tx
>test
>works from 1 (one) foot away
based chinkshit

>>1407247
digi-key and others categorize their search results. I just punched TDC7200 into the digikey part search box, clicked on the "ICs - sensors and detectors" category, and there were four buying options for the same part (full reel, cut tape, partial reel, and tube)
>period
any necessary ALC would occur over a several-pulses period, outside the TDC. once you've gotten a good zero reference, it's unlikely that the properties of the medium would change so quickly

>>1407257
it could be model-dependent. I have yet to see a car (in burgerland) that doesn't provide power to the accessory plug in the ACC position, and plenty that provide power when off

>>1403081
>Components/equipment sources?
Can you add RS to this list, they do free next-day delivery in the U.K. on all orders if the part is in stock. Mouser and Digikey charge £12(!) on orders under £30.
https://uk.rs-online.com/web/

>>1407261
Oh, and I should just be able to use the resistor adapters that lowers voltage, for lower rpm, to control the rpm.

This is so good!

>>1403376
So that's why some amplifier circuits take 2 9v batteries

Interesting

>>1407457

forget resistor adapters and y-splitters; these things just add crazy costs to the project. you can do the same thing with a wire stripper and electrical tape from the dollar store. cut off all connectors (keep them if you wanna re-use the fans later) and splice the wires and any resistors.

>>1404570
>full-analog
>CD4017
anon, Bob Widlar would like a word...

>>1404912
>Vias should be covered with solder resist.
That is untrue. There are plenty of reasons to leave vias untented.

>>1407418
Well I can find a particular model like the 7200 or GP22 if I know its name, but I can't search for a list of different TDC models.

Going off Digikey's "Interface - Sensor and Detector Interfaces" category I could only find two brands' TDCs, TI and ams. But it looks like TI's TDC7200 and maybe TDC1000 are the way to go just on a price basis.

Can I fuck up the antenna signal for my whole complex if I accidentally shorted the signal and shield when installing a coaxial outlet?

>>1407736

not likely. presumably there are splitters which will isolate the fault to a confined area. also, RF is weird, so shorting a cable in a branch somewhere is not catastrophic. presumably it sees the shorted cable as a some kind of load.

>>1407085
What series of capacitor is it? Might be low ESR type.

Will I fuck something up if I connect two open-drain output pins to external pulldown resistors?

I'm not using those pins at all. I'm not fucking up the internal circuitry by doing this, right?

So, I`m planning to code my Uno to turn this 50w led on and off under certain conditions. (When my pickit arrives I`ll use a pic). But I don`t know what transistor to use (or what other thing) in order to do the switching. Anybody has a recomendation? It needs to handle 5-6 A at 12v.

I can't remember the reason I put the diode in parallel with the motor, but I know it was important.

Does anyone else know?

>>1407955
http://www.redrok.com/MOSFET_IRL540N_100V_36A_44mO_Vth2.0_TO-220.pdf

>>1407966
motors are inductors. inductors try to stabilize current in the same way that capacitors stabilize voltage. when you shut off your transistor, the inductor tries to keep the current flowing by creating a voltage spike. the voltage will increase until the current flows, eventually exceeding the voltage rating on your fet and killing it. that diode opens up a new path for the current when the switched terminal on the motor reaches ~0.7v above the supply voltage. so on turnoff you'll have a brief period of current flowing through that diode until the inductor expends its stored energy.

>>1407972
I see, So I need to find a diode that can withstand the current spike?

>>1407975
yes. but to be sure you're not thinking of it wrong, it's not really a spike as in a brief very large amount of current. it's just that the same current that ran through the transistor now runs through the diode. so a diode rated for a pulse current equal to the continuous current in your fet is fine.

1) fet is on
2) fet turns off (<1 microsecond), meanwhile voltage increases on drain until it reaches ~36.7v sending the current through the diode
3) current in the diode decreases linearly until the motor is out of stored energy

>>1407853
https://www.digikey.com/product-detail/en/rubycon/25YXG470MEFC10X16/1189-1735-ND/3563695

>>1407972
>motors are inductors. inductors try to stabilize current

not quite. the motor becomes a generator when you cut off power to it: a moving coil in a magnetic field makes a generator, as everyone learned in high school physics. in this case, a DC generator where the polarity is always the opposite of the previously applied voltage.

even a weak diode (1N4001) should be able to handle that generated current easily.

I have a problem visualizing/understanding the importance in volts in a circuit.

If we go back to the hydraulic model, volts are essentially how large the water tank is.

Why then does it have any bearing what so ever on how much energy goes through a circuit when it's just the container, and not the rate of flow between the two?

Unless it's the sheer difference in "pressure" between the container and ground that forces more power through, but I thought that would be amperage

>>1407063
>it's good if you know Verilog and digital logic design already
I do. I'll definitely check this book out, thank you

>>1408157
voltage is pressure, current (not amperage) is flow.

>>1408166
so voltage is essentially if I were to cut a hole at the bottom of a water tower, it would be under enough pressure to probably cut shit even though the rate of flow would still be low

>>1408168
yes. The hole being small means that the resistance to flow is high, which is why the flow of water is low.

While the water pipe analogy can help at the beginning, you'd do well to leave it behind asap, as it doesn't completely hold up for certain concepts. The true physics behind electricity isn't super complicated anyway. Just think about electric fields, charges, and potentials.

>>1408142
The diode would still be required even if the switched motor doesn't rotate at all. Can you imagine why that is?

Hey guys I’m tring to build a washing machine controller with a H bridge and an arduino uno r3, I want it to go high frequency does anyone know where I could get a free program?

>>1408166
>not amperage
voltage * amperage = tension * current

>>1408190
what about ohmage?

>>1408191
Ohmage lol

NEW BREAD
>>1408236
>>1408236
>>1408236

>>1403148
Your scope is doing its job, If you want the signal to look different change the cct.

>>1403530
A women emotions over the course of a month. Was this a trick question?

>>1404073
It been a long time since i was in a nice car with fantasy tech like that. 80's caddi if i can recall correctly (i was a passenger). Projector from the dash reflecting off a special surface impregnated into the windshield. You need to know some light physic OP + lectronics + materials.

>>1408142
What if i run 4x 1n4001 diodes in parallel, would that still work?

>>1408647
Diodes don't share current evenly in parallel like resistors do. Usually one will end up sinking most of the current and the others will basically do nothing.