|
Forseti posted:Reminds me of the construction of this dude's discrete transistor pong. My reaction to this was "Holy poo poo this guy is German as gently caress". This is the most awesome yet batshit nuts thing I have seen. That wiring job is just perfect though. Kudos to him though. It really highlights how great IC's are though. You could replace 80% of that with something half the size of a postage stamp.
|
# ? May 8, 2020 16:26 |
|
|
# ? Jun 13, 2024 04:38 |
|
that’s getting close to “early computers using wire wrap circuitry” territory
|
# ? May 8, 2020 17:20 |
|
My favorite wire wrap thing is that NASA created a goddamn robot to do it consistently for Apollo: https://www.youtube.com/watch?v=ndvmFlg1WmE&t=1526s (That whole video is excellent if you got half an hour to kill)
|
# ? May 8, 2020 18:00 |
|
There's all kinds of cool poo poo for Apollo. It's more programming than electronics, but this video about how the computer actually works and how advanced it really was is really great if you have an hour. https://www.youtube.com/watch?v=xx7Lfh5SKUQ
|
# ? May 8, 2020 18:27 |
|
Ambrose Burnside posted:in other news, i’ve become hopelessly engrossed in making Pretty Component Subassemblies instead of just, building the fuckin circuit. weird things happen as i try to redirect my mental rudder from art metalworking to electronics Flywiring has a some legendary ee supporters:
|
# ? May 8, 2020 18:27 |
|
Wow that's (almost) literal spaghetti code
|
# ? May 8, 2020 19:21 |
|
Cojawfee posted:There's all kinds of cool poo poo for Apollo. It's more programming than electronics, but this video about how the computer actually works and how advanced it really was is really great if you have an hour. I love the audience reaction when he gets to the superbank
|
# ? May 8, 2020 19:29 |
|
Long shot question: I'm trying to identify this IC on a circuit board for a small electronic fire safe. I know it's probably custom, but does anyone know of a 14-DIP microcontroller with these Vdd and Vss locations? I've found some EM78 series microcontrollers with a similar pinout but they're OTP ROM so I'm guessing they can't support the safe's feature to set an arbitrary access code
|
# ? May 8, 2020 21:49 |
|
Code inbound re pH: Here's what I'm thinking: - Linear model for 2-pt calibration, using only the pts to determine line characteristics, with slope having a linear, hard-coded dependence on temp (Still trying to figure out the best way to do the temp correction; might have to fudge it based on data) - Lagrange polynomial (quadratic) for 3-pt calibration, with the same hard-coded linear dependence on temp. Notice no mention of the nernst eq or constants, outside the temp correction. Going to try this with the 3 buffer solns I got, but could theoretically be used with any 2 or three known pH solns, with sufficient spacing. I think the 2-buffer linear approach will be fine, but if you have more data pts, why not, since buffer solns tend to come in sets of 3.
|
# ? May 8, 2020 21:52 |
|
What's your goal? You could create a new controller for the safe with a microcontroller dev board as an alternative to figuring this out.
|
# ? May 8, 2020 21:52 |
|
handle posted:Long shot question: I'm trying to identify this IC on a circuit board for a small electronic fire safe. I know it's probably custom, but does anyone know of a 14-DIP microcontroller with these Vdd and Vss locations? I've found some EM78 series microcontrollers with a similar pinout but they're OTP ROM so I'm guessing they can't support the safe's feature to set an arbitrary access code Honestly everything else on that board looks really standard, including all the connectors, so it would probably be much easier to just make something new than try to force an equivalent micro into place
|
# ? May 8, 2020 23:18 |
|
If it's six volts, it's probably some custom IC.
|
# ? May 8, 2020 23:34 |
|
Not temp compensated, but otherwise working pH code, supporting 2-pt and 3-pt calibration. Working great using buffers, even with the cheap probes. It jumps +-.2 pH, bouncing around the real value, so I think I need smoothing code.The only assumption is a linear mapping of voltage to pH. In the 3-pt, we compensate for a minor non-linearity.Rust code:
Essentially, you measure voltage at 2, or 3 known values, eg from buffer solns discussed above. You then put these in the `cal` values. It uses this to form either a linear, or quadratic fn mapping V to pH. I made Rust and Python modules. Rust is for any microcontroller, Python is for linux SBCs or boards using CircuitPython. Py and Rust modules published. Example code in Py calling the module: Python code:
Dominoes fucked around with this message at 06:18 on May 9, 2020 |
# ? May 9, 2020 00:12 |
|
So I have this Epson EH-TW 5500 projector that I got for cheap. Pretty sure it got hit by lightning sometime last year (after I bought it). It turns on fine but it doesn't take any video input at all. I ripped it open and found this: which, from the L300 markings, google suggests might be an inductor? Would it be at all feasible to find a similar one and just pop it out and replace it or am I better off trying to track down the entire PCB? (no I'm not going near this thing myself with a soldering iron as I'm clearly clueless, but I do know a couple of dudes who could probably do it for me)
|
# ? May 9, 2020 04:36 |
|
Try it. See what happens. That's the fun, right?
|
# ? May 9, 2020 04:42 |
|
Can you read what it says on it or find a similar photo with one that isn't melted? Measure it, go to digikey, and try to find one that has the right size, and looks about right, then dig in the datasheets and try to find out what the code means so you can get the right replacement. What chip is that near? Can you see something on the other side connected to the vias next to the inductor? There is probably a switching power supply IC and you could check out the datasheet for that (assuming you can find out what it is) and estimate what inductor would be required based on the application advice, as a way to work from the other direction. You might need to know that anyway, the IC might be (is probably) damaged too. taqueso fucked around with this message at 04:51 on May 9, 2020 |
# ? May 9, 2020 04:47 |
|
Dominoes posted:Try it. See what happens. That's the fun, right? Sure but like, where do I get one? All I have to go on is "L300" and that's not turning up much. I don't know what to look for beyond that. taqueso posted:Can you read what it says on it or find a similar photo with one that isn't melted? Measure it, go to digikey, and try to find one that has the right size, and looks about right, then dig in the datasheets and try to find out what the code means so you can get the right replacement. I can try to get as better picture of it, I didn't even realize there was anything written on it until after I'd taken the picture. The chip next to it says Silicon Image plus a bunch of stuff I would have to google again, I assume it handles the various input ports. I can look at it a bit closer tomorrow, I've been up til 4 am now staring at this thing and decided to just post it before I hit the sack in case this was the sort of thing where you can just buy a whole bag of em for 5 bucks Nastyman fucked around with this message at 04:59 on May 9, 2020 |
# ? May 9, 2020 04:50 |
|
well you've got "2E" something at least https://www.digikey.com/products/en...e=1&pageSize=25 size will be a big discriminator See the big chip at the bottom of the pic? That's probably supposed to be powered by this. You could look up that chip and see what it's power requirements are for further way to cull the list. Such as, expected voltage leads to an inductor that is rated for at least that voltage and probably not more than 1-ish order of magnitude more than that. PS L300 by itself is meaningless, you need either a schematic or bill of materials that will tell us more. Or maybe some random internet posting about how they fixed their identical projector and they mention it. Service manuals often have bill of materials sections or at least part lists for commonly replaced parts taqueso fucked around with this message at 05:00 on May 9, 2020 |
# ? May 9, 2020 04:51 |
taqueso posted:What's your goal? You could create a new controller for the safe with a microcontroller dev board as an alternative to figuring this out. Or even just graft a new micro on a daughterboard or wired adapter into the slot, it's all through-hole anyway
|
|
# ? May 9, 2020 04:55 |
|
taqueso posted:well you've got "2E" something at least Thanks, that should at least get me a bit further. I'll give everything a fresh look tomorrow and see what comes up.
|
# ? May 9, 2020 05:01 |
|
Nastyman posted:Thanks, that should at least get me a bit further. I'll give everything a fresh look tomorrow and see what comes up. Just to emphasize though, if something was strong enough to kill an inductor (which is basically just a fancy coil of wire so they tend to be reasonably durable) it's almost definitely enough to kill everything that inductor was connected to, so you've probably got a bit of an uphill battle if you're going to try and repair this thing...
|
# ? May 9, 2020 06:13 |
|
And that big downstream chip is a prime candidate and it may be be hard/impossible to find without taking apart other projectors.
|
# ? May 9, 2020 06:22 |
|
Actually here's a quick "go/no-go" test maybe: I bet if it's a beefy inductor that got extremely fried, it's a choke for filtering, right on the main input of the device. You don't reeeeeally need those usually (if you don't want to pass FCC testing), so just glob a whole bunch of solder to jump over it and try booting the thing right up And maybe use a GFCI plug to do it in
|
# ? May 9, 2020 06:45 |
|
Cojawfee posted:If it's six volts, it's probably some custom IC. taqueso posted:What's your goal? You could create a new controller for the safe with a microcontroller dev board as an alternative to figuring this out. Shame Boy posted:Honestly everything else on that board looks really standard, including all the connectors, so it would probably be much easier to just make something new than try to force an equivalent micro into place The safe doesn't retain lockouts after power cycling, so my other idea is to hook into the multi-color LED and use some kind of adder to briefly interrupt power after the third incorrect try. Is there a name for like.. building parasitic circuits? Or guidelines?
|
# ? May 9, 2020 08:34 |
|
Hardware hacking
|
# ? May 9, 2020 08:39 |
|
Shame Boy posted:Just to emphasize though, if something was strong enough to kill an inductor (which is basically just a fancy coil of wire so they tend to be reasonably durable) it's almost definitely enough to kill everything that inductor was connected to, so you've probably got a bit of an uphill battle if you're going to try and repair this thing... That + the HDMI out port on my amp It's sounding more and more like replacing the board is the easiest and quickest solution, which is really what I wanted to know. I might try some of the solutions posted here but I had my suspicions from the get-go that this wasn't going to be an easy fix.
|
# ? May 9, 2020 13:24 |
|
Nastyman posted:So I have this Epson EH-TW 5500 projector that I got for cheap. Pretty sure it got hit by lightning sometime last year (after I bought it). It turns on fine but it doesn't take any video input at all. I ripped it open and found this: Sorry bud it's a total write off. Replace the entire PCB and even then if there's other PCBs they might be fried too.
|
# ? May 9, 2020 13:41 |
|
handle posted:The safe doesn't retain lockouts after power cycling, so my other idea is to hook into the multi-color LED and use some kind of adder to briefly interrupt power after the third incorrect try. Is there a name for like.. building parasitic circuits? Or guidelines?
|
# ? May 10, 2020 02:15 |
|
Subjective question: Is it OK to run multiple ADCs vice one with many inputs? I have to read 9 signals (4x differential pairs, one single-ended input). I have a 4-channel ADC that works well, has standards-adhering drivers in the lang I'm using, and I've verified as working for my use, with a lower number of inputs. It has an easy addressing system for deconflicting multiple ones, and is reasonably cheap. Should I try to find a 9+ channel AC, or is it OK to use 3 of these?
|
# ? May 10, 2020 03:40 |
|
Sure you can use more than one.
|
# ? May 10, 2020 04:01 |
|
Awesome; going to roll with that. It looks like finding a single one with more than 8 inputs limits options significantly.
|
# ? May 10, 2020 04:37 |
|
If you don't need to sample them simultaneously, you can also stick an analog mux IC in front instead of relying on the mux in your ADC.
|
# ? May 10, 2020 07:35 |
|
So instead of pseudocode like code:
code:
I guess this comes down to part cost / complexity etc. It's kinda neat how the ADC lets you select the address by connecting an address pin to a specific one of the other pins. Not sure if this is standard, but it's convenient and explicit.
|
# ? May 10, 2020 15:55 |
|
Dominoes posted:So instead of pseudocode like Yes, pretty much. Depends on whether you have a library for your chosen part or not. If somebody's already written the code it might indeed be as simple as code:
https://www.sparkfun.com/products/9056 It looks like there's some example code in the comments but the description explains how you operate it even without that. The 4 pins marked S0-S3 allow you to give the device a parallel 4-bit binary number (so 0 to 15) which selects the input you want. So to select input 11, you'd have to put a binary 11 (b1011) on the pins before doing the read: code:
And yes, most little I2C devices that support multiple addresses do it with a couple of address-select pins. They usually also work in binary, as above. Maybe your device defaults to address 0x70 and there are three pins, and each one can be activated (set to 1) by connecting it to ground. With three pins/bits you have 8 possible selections. Leave all the pins disconnected and you get the base address, or connect the pins to form a binary number and add it to the base address to get the new address: code:
|
# ? May 10, 2020 17:17 |
|
A few months ago I asked about CPLD families, so I thought I'd give an update now that I've built my boards and am working on the software side of it. I chose the MAX V CPLD series in 64 EQFP (0.4mm pitch). I think it's pushing the limits of what OSHParks 4 layer process can do, but I had no real issues getting the CPLD on the board without shorts or damaged pads. I used a plated through hole for the center ground pad, that way I could just solder the pad directly (a trick I first saw in a weird low volume network switch I got on AliExpress). Quartus Lite was pretty simple to get working in schematic mode, and worked well with the USB Blaster Digi-Key sells. Amazingly I got the pinout right for the JTAG port. The schematic mode is nice (if a bit limited/clunky to work in) and I was able to do all the essentials very easily. For a hardware guy like me it's very convenient to be able to use literal 7400 part models in the design (and it's definitely helpful when porting a discrete logic design). Still working on the final code, but I'll definitely be using CPLDs more in the future. Only downside I can see is needing a 1.8V core supply, but this design already needed that for the DSP core so it didn't cost me anything to use it.
|
# ? May 11, 2020 12:09 |
|
longview posted:I chose the MAX V CPLD series in 64 EQFP (0.4mm pitch). I think it's pushing the limits of what OSHParks 4 layer process can do, but I had no real issues getting the CPLD on the board without shorts or damaged pads. This is a great trick, I highly recommend it if you're doing something low volume, particularly if you're relying on that central ground pad to get heat into the ground plane.
|
# ? May 11, 2020 13:36 |
|
I'm an electrics idiot who has a question to throw at the thread, if I may? For want of better things to do during COVID-19 lockdown, I'm sorting out various jobs on bits of the house and my garden machinery, one of which is my old Makita generator (recently featured in the Small Engine Repair Thread in A/I). Now that's running properly again, but in the process I managed to break the Oil Warning Lamp which flashes if the automatic low-oil-level shutdown system is triggered. The lamp still works, but the body has snapped where it passes through the control panel, which is all that holds the lamp in place, and it has defied my attempts to glue/epoxy it back together. So, out of a combination of boredom and perfectionism I'm looking to source a replacement. My question is - what sort of lamp is it? (With the messy remains of my glueing attempts) (It has a diode in the body below the actual lamp - I add this only because I've seen pics of mini neon lamps with a diode on one of the 'tails' (?). There are no voltage or other markings on it) It looks like an LED to me, but this is an early/mid-1980s bit of machinery. I can find electronics suppliers who will sell me a lamp that looks exactly the same (like this: https://www.radioshack.com/products/radioshack-120-volt-neon-red-square-lamp-2-pack) but they're all miniature neon lamps and I'm pretty sure that's not it because I really doubt this is working off mains voltage. It's a 240/120V generator, but the oil shutdown system is totally self-contained and runs off the engine's ignition. It has a piezo-electric sensor sticking into the engine oil, and a control box screwed to the side of the engine. When the oil level drops and exposes the probe, the piezo-electric pulses are no longer damped out and they trigger the control box to ground the engine magneto and flash the warning light until the engine stops. The control box and the lamp get power from the magneto's primary coil. I've put a multimeter across the two terminals at the back of the lamp and when the shut-down is triggered and the lamp flashes you get between 3 and 4 volts for around six seconds while the engine coasts to a stop. I'm not fussed about getting an exact replacement - any indicator lamp that will fit in a 10mm hole will do - but my question is what lamp do you think I need? 12-volt LEDs are easy to get. A 6V one? Or is it really a neon unit? This seems to something that's impossible to get a straight answer to by browsing the internet, so I'd rather trust goon experts here.
|
# ? May 11, 2020 21:08 |
Is that a diode or a resistor?
|
|
# ? May 11, 2020 21:52 |
|
I'm not an engineer and I need a safety/sanity check. I have some stuff on a breadboard, I will power it from USB device A, it has a 3.3v LDO. I will have USB device B with it's own LDO(s). Device A provides USB connectivity and device B will provide clocks until the 74xx arrives. There's also USB device C, a saleae logic analyzer clone with, I assume, it's own LDO. All of it will be connected to my laptop, which I very much don't want to fry/short-circuit, or lose USB ports. I got a $2 USB hub to hang everything of off. I assume I can't count on the USB GND on different laptop ports to be continous but I probably can count on the USB GND of the USB hub downstream ports to be connected. Concrete questions: - Breadboard power rails are from device A, am I safe as long as no LDO +3.3v from different devices touch? (Incidentally, what happens if you put 2 LDOs in parallel?) - There's no issue in feeding 3.3v outputs (the clocks) to the breadboard devices as long as they share a GND, so the 3.3v are similar enough within the tolerances of the LDOs, right? - The logic analyzer has a "GND" connector, is it 'output' or 'input'? Should I connect it to the device A GND pin or just ignore as it's just the same GND from the USB hub anyway? - Will I fry my laptop if I apply a multimeter in beep mode to different USB ports' GND to check for continuity? - Is there anything else I should consider? Cumulative power use should be under 500mA, probably, unless I go Christmas tree mad with the LEDs. e: googled the parallel LDOs question *nervous twitch* karoshi fucked around with this message at 22:05 on May 11, 2020 |
# ? May 11, 2020 21:53 |
|
|
# ? Jun 13, 2024 04:38 |
|
BalloonFish posted:what sort of lamp is it? That looks to me like a bog-standard 5mm LED in a custom threaded housing. Looks like a resistor inside soldered to one of the legs, not a diode. You might be able to pry it apart and read the color code; I'd guess it's about 100-300 ohms if you're getting 4v on the circuit to illuminate it. It is almost certainly not a neon lamp because those take >100vAC to illuminate. A "12v" LED is just an LED with an attached resistor sized to allow it to run properly on a 12v source. If you know that the circuit driving that lamp outputs 4v, you can replace it with any <4v LED (for instance a red one that drops about 1.8v) and an appropriately sized resistor (4v-1.8v = 2.2v; 2.2v/0.020A nominal current = 110 ohms, 100 would be fine).
|
# ? May 11, 2020 22:04 |