|
longview posted:Have you tried setting different start/stop times for the different channels? Switching all those channels in at the same time could cause some weirdness. I had not tried that, but the fact that I need to be able to adjust through the full range of PWM settings for each of 48 LEDs across 3 chips makes that seem like an unlikely solution for anything beyond a small subset of the problematic PWM values. Stabby McDamage posted:Anything that's not a square wave at the requested frequency. I'd run the duty cycle up and down slowly while looking at the waveform. It should look like this: I managed to track down the oscilloscope (which was then abruptly moved along with the entire electronics bench to another room to satisfy a new floor plan requirement the head of the lab gave) and measuring voltage changes revealed absolutely nothing discernible at low frequencies. I was able to detect mains frequency at the point where the DC power supply met the controller board, but nothing after that other than the high frequency PWM signal. ante posted:Yeah, pretty much like the above. Tap into the signals and test your assumptions about what you think is supposed to be on the output. So, this approach actually lead to finding the solution, more or less. I decided I should try to track down an oscilloscope compatible photometer head (my old lab had several of them) and it turns out a lab with more than half a dozen visual psychophysics researchers has nothing suitable for measuring light output frequencies beyond a few Hz, which was a frustrating surprise. My group's electronic technician had no useful suggestions other than people to ask, and I ended up talking to the top electronics guy for the entire lab only to find out that the closest thing he had was an ancient hand-built deadbug circuit phototransistor set-up which was completely swamped and non-responsive with the light intensity I had, so that was fun. A bit more probing later and then some "let's see what sticks" testing and it turns out that the magnitude of the low-frequency LED flicker is related to the length of the cable connecting it to the controller. (I'm running power, ground, and I2C over micro USB because I'm a terrible person and the cables are easily replaceable.) The 3m cables I had produce more flicker than a 2m cable I brought from home which produced a slight amount while a well-shielded ~1yd phone cable produced none. I think the answer to my problem is actually just getting short cables with ferrites and putting a set of holes for an electrolytic filter cap into the next version of the board. The flickering frequency would probably be detectable if I was willing to cut a contact on the board and measure current instead of voltage (since voltage was steady enough outside of the PWM frequency to not show up on the scope, but total draw increased on a constant-voltage PS when the cable length was shortened). However, I'm going to consider the problem fixable as-is and move ahead so we can get a protocol off to the IRB. For anyone else who's also in the situation of building their own lab equipment with ~0 formal training in electronics, don't get discouraged. Science isn't about "why", it's about "how". 
|
#
?
Feb 2, 2017 05:35
|
|
|
|
| # ? May 9, 2024 13:27 |
|
|
I had asked a while back if it was possible to wire a bunch of batteries in full series (to run a load) but also have a charger (i.e. from a solar panel) that charges up each individual battery in parallel. The answer at the time was no, because I'd be shorting that batteries. What about if I were to get a whole bunch of these "Blocking Diodes" though? http://www.ebay.com/itm/Wind-Turbine-100-AMP-High-1600-Volt-Blocking-Diode-Solar-Panel-100RIA160bu-ONE-/301960468376 Is there any way that I could safely wire everything up so that electricity can flow only out of the batteries to the load in series, and only into the batteries from the charger in parallel? Could I have both charging and discharging occurring simultaneously with a properly designed circuit using these diodes?
|
|
#
?
Feb 2, 2017 07:31
|
|
|
Zero VGS posted:I had asked a while back if it was possible to wire a bunch of batteries in full series (to run a load) but also have a charger (i.e. from a solar panel) that charges up each individual battery in parallel. If you only have one charger, and you want simultaneous charging of all cells, then no. Think of it this way: your charger only produces one potential difference. EACH cell produces a potential difference, but since they are in series, they all have a voltage relative to one another. When you charge, you're tying a common reference point together, equating the voltage at the cell with the voltage at the charger, so you can only do one at a time, because otherwise, the nodes coming out of your charger would need to be all of those voltages at once. You need: A) an equal number of cells and chargers B) a full stack charger and cell balancer C) the ability to disconnect cells from each other for charging (switching to parallel) Or D) the ability to switch between cells and charge sequentially D is the only one that matches your requirements, but it's a lot more complicated than the circuit you currently have, and not something I'd suggest for a beginner, especially since battery charging is involved, so there are safety issues. KnifeWrench fucked around with this message at 15:59 on Feb 2, 2017 |
|
#
?
Feb 2, 2017 15:39
|
|
|
ante posted:It'll be sending some sort of digital signal via IR, which then controls a bunch of stuff on the TV, probably. It's not like breaking out a single wire to a button. Hardwire the remote to power and tape it to the CRT? Arduino with an IR LED plugged into it sending the signal on button press? Arduino hooked into the signal path of input IR, bypassing the light stuff entirely. My best idea was somehow capturing that button's IR code and building a small circuit that could reproduce it on the IR receiver's data line when a button was pressed. I admit that I'm not very strong when it comes to circuits. Also, this is a 9" TV. There's no place I could mount that remote that would have its signal reach the IR sensor and not block the screen. Or could I somehow wire the IR sender in the remote to the receiver instead of relying on the IR light? kid sinister fucked around with this message at 17:40 on Feb 2, 2017 |
|
#
?
Feb 2, 2017 17:37
|
|
|
kid sinister posted:My best idea was somehow capturing that button's IR code and building a small circuit that could reproduce it on the IR receiver's data line when a button was pressed. I admit that I'm not very strong when it comes to circuits. This is actually what I was thinking, you could probably replace the IR receiver with something like a MOSFET and just feed it the "IR" signals directly using an Arduino - they're low enough frequency that you should be able to generate them manually without too much trouble. This all depends on how the circuit in the TV is built, however.
|
|
#
?
Feb 2, 2017 17:51
|
|
|
kid sinister posted:
Got access to any scrap optical fiber? 
|
|
#
?
Feb 2, 2017 19:20
|
|
|
kid sinister posted:My best idea was somehow capturing that button's IR code and building a small circuit that could reproduce it on the IR receiver's data line when a button was pressed. I admit that I'm not very strong when it comes to circuits. I like hacking cool solutions as much as the next guy, but why not just slap velcro tape to the remote and screen to keep them together?
|
|
#
?
Feb 2, 2017 22:06
|
|
|
Stabby McDamage posted:I like hacking cool solutions as much as the next guy, but why not just slap velcro tape to the remote and screen to keep them together? I tried that. The velcro pad didn't make it through the summer in my truck.
|
|
#
?
Feb 3, 2017 00:57
|
|
|
kid sinister posted:I tried that. The velcro pad didn't make it through the summer in my truck. Ah. 3M Command Strips with dual-lock, then? Or ziptie it to a string and screw the string to the chassis? Basically, because I don't see an easy way to lock it to the mode you want, I feel like you'd be building some kind of remote-faker, and if you're going to do that, you may as well just make it easier to keep track of the real remote.
|
|
#
?
Feb 3, 2017 21:41
|
|
|
Effective-Disorder posted:https://en.wikipedia.org/wiki/Universal_motor Fat Turkey posted:Edit: Literally the moment after I post my question post relating to DHT11, I solved it. I'm an idiot! peepsalot posted:That's the commutator, it's made of copper, not magnets. It's a brushed motor, those graphite blocks on springs are the brushes. Stepper motors are brushless and way different from that. Despite my coming into this thread consistently entirely ignorant and largely incorrect with my posts, I really appreciate being set straight because I've pretty much gleaned like, 1 or 2 semesters' worth of electrical engineer and theory classes off of you-all, and I won't have a chance to even take a basic EET 101 course until next fall - so I fully intend to have bull-head my way through all that poo poo. dit: I accidentally hit "post" instead of "preview" so if part so this post are entirely incomprehensible - it's because I was trying to save it and edit it tomorrow morning. My bad. If nothning else, I'm beginning to be able to eyeball a PCB and read what is on it, and then be able to immediately plot out a schematic and begin working on the math. coyo7e fucked around with this message at 07:24 on Feb 5, 2017 |
|
#
?
Feb 5, 2017 07:20
|
|
|
I want to strip down a dead computer power supply, I want to reclaim the fans, fan grills, screws, etc. I've noticed the thing makes a high pitched squealing sound when I plug it in, this last about 20 seconds and then goes away, and then the squealing repeats when I flip the switch off or unplug the power supply. I have heard this is a stupid idea and answers I've found on google says the caps will hold a charge for seconds, to hours/days or forever. One source did say "sometimes you will hear a ticking sound as the <something> resistors are draining the caps". I'm guessing I should be relatively safe if I let the thing sit for 24hrs and wear a pair of gloves. I think the standard procedure for discharging a capacitor is to short the terminals together? I honestly want to just clip the fan leads and throw the board in the trash (I don't see a "no trash can" sticker on it anywhere, although I suspect a haz mat place might be better disposal). It's only a little 400W ATX power supply, it cant kill me too much, right??
|
|
#
?
Feb 5, 2017 23:28
|
|
|
Crotch Fruit posted:It's only a little 400W ATX power supply, it cant kill me too much, right?? Leaving it for twenty‐four hours is plenty. Unlike big vacuum tubes, which can store kilovolts for months, ordinary electrolytic capacitors will discharge rather quickly even if no specific precautions were taken by the circuit designer (e.g. bleeder resistors). Besides, I expect the highest any of the big caps are charged to is 12 V, which is touch-safe. Leave it unplugged overnight and short the rails before opening it up if you want to be extra sure everything is discharged. Then open it up and take note of the ratings on the big caps to confirm that they aren’t even capable of having mains voltage across them.
|
|
#
?
Feb 5, 2017 23:43
|
|
|
Platystemon posted:Leaving it for twenty‐four hours is plenty.
|
|
#
?
Feb 5, 2017 23:49
|
|
|
Crotch Fruit posted:By shorting the rails do you mean just touch the 12v to ground, 5v to ground and so on at the molex connectors or something more elaborate? That’s what I mean.
|
|
#
?
Feb 5, 2017 23:52
|
|
|
Every PC power supply I've taken apart charges some big electrolytic caps up to rectified mains voltage. They've also all had bleeder resistors across them but I don't know if I'd want to necessarily assume that.
|
|
#
?
Feb 6, 2017 00:23
|
|
|
I thought energy standards put an end to that in the past decade or so.
|
|
#
?
Feb 6, 2017 00:29
|
|
|
coyo7e posted:If you are interested - I'd still like to know what your question was and what you answer as, and why it stumped you... I'm still really green at this and listening to people work through their own fuckups has been one of my biggest boons. Heh, it's a very silly one, but I'll share, as it is a lesson against hunting for a complicated answer to your problem, without thoroughly checking the basics first. I had the DHT11 fully wired up to a Nano and a PC to begin testing. I knew the DHT was 'on' and doing its thing but the Nano wasn't getting any data from the DATA pin for the readings. I tried different ways of defining the pin, and trying different pins in case I had a faulty one. I tried lots and lots of things and finally posted here. And then it struck me to double check the DATA connection really was connected to Pin 2. I had had to force the Nano into my breadboard because it was a bad fit, and while 3.3v and GND from the Nano were able to power the DHT, and I was sure data was coming out of DATA (the voltage kept changing), I had never checked the breadboard was connecting to pin 2 in my botched fitting. So I removed the breadboard and did it all with connecting wires. And it worked. D'oh! I was going to make the next step in my quest to add the ESP8266 WiFi module to this setup to get the DHT to send the info to the net, but the more I read, I don't think the Nano is needed, the ESP has enough power to do it itself. Might be harder to trouble shoot but that's the next step. I think I'll sign up to ThingSpeak to get a dedicated site to upload my data onto, saves me the trouble of coding up my own website and server, even if it means my precious precious domestic temperature is out there for all to see! Hope that helps in some way!
|
|
#
?
Feb 6, 2017 14:43
|
|
|
Using the esp8266 without the nano will actually be easier to debug, because it frees up the serial connection to the computer. Programming is harder though as, unless you have a very fancy esp board, you'll probably need to use a programmer rather than just a USB cable. For those new to the game, the arduino uno and nano only have one serial (aka UART) connection, and that's what you use to get the esp and arduino to communicate. It's also what you use when you program the arduino, and what you use when you use the serial monitor. Which makes debugging with serial.print() a pain and programming fiddly.
|
|
#
?
Feb 7, 2017 00:02
|
|
|
Fat Turkey posted:a lesson against hunting for a complicated answer to your problem, without thoroughly checking the basics first This was my EE undergrad, in a nutshell. This is a really hard lesson to internalize.
|
|
#
?
Feb 7, 2017 00:35
|
|
|
Another dumb question from me: If have a battery charger meant for 24v li-ion batteries, and want to charge a li-ion battery that is 312v in series, can I use a buck-boost converter or something to that effect to step the charger's voltage up? My two concerns would be: a) Would the charger "see" the 312v battery as a 24v battery on it's own side of the converter? Would it have an accurate idea of voltage level for purposes of determining state of charge? b) Would this be really inefficient? Are these any off-the-shelf parts, even expensive ones, that would solve this and keep the efficiency loss as low as possible?
|
|
#
?
Feb 7, 2017 01:30
|
|
|
Splode posted:Using the esp8266 without the nano will actually be easier to debug, because it frees up the serial connection to the computer. Programming is harder though as, unless you have a very fancy esp board, you'll probably need to use a programmer rather than just a USB cable. Fancy ESP board is just a $4 NodeMCU off AliExpress, and they work great. Recommended usually over the bare $2 models because they got power regulation, serial converters, broken out pins, everything
|
|
#
?
Feb 7, 2017 01:41
|
|
|
For better for worse, I've started off with the basic ESP8266, not connected to NodeMCU but I'm happy to flash it if need be. Played about with it tonight and got a bit stuck but learning 1) Accorsimg to some YouTube videos for the ESP8266, neither the USB port nor the Arduino can supply it with sufficient power even at the correct 3.3V. I've ordered some 800mA 3.3V regulators but this project doesn't sound like a current drain, and if it is, it's not going to last on 2 AA batteries very long... 2) The DHT11 seems to vary tempt and Humidty by 2 degrees/2% depending on whether I supply ti with 5V or 3.3V. That's annoying. Testing it with another cheap thermometer shows the 3.3V to be correct, guess I'm going to have to get this voltage sorted! 3) I had bought a "RS232 TO USB SERIAL CONSOLE ADAPTER" to speak to it via USB, but it feels like it took so long to get here I can't quite recall how to work it all up. Seems to run off the 3.3V, but maybe not when it actually gets sending. I'm sure I watched some youtube video on how to get this to work, and I'm sure they to speak to it through the Arduino IDE despite being plugged in with this adaptor. I think I need a sleep and a night off from it and attacking again tomorrow morning. But for something that should be so simple, the fact I can't find a tutorial for ESP+DHT only, I'm wondering whether it can't actually be done. Fat Turkey fucked around with this message at 02:55 on Feb 7, 2017 |
|
#
?
Feb 7, 2017 02:36
|
|
|
The NodeMCU is the ESP8266 dev board. You should get it, especially if you're having issues. I mean, it also comes with LUA firmware on that, but you can wipe it over USB just like you would a bare module + serial adaptor (but with the built-in adaptor).
|
|
#
?
Feb 7, 2017 03:13
|
|
|
Today I realised that the two RFID read and write modules I'd been using for the past few weeks are supposed to be powered by 3.3V rather than 5. Oops! I guess those chips are pretty tough. I do wish that shops advertising stuff as arduino compatible would at least make it clear that it's only good for 3.3V. Now I have to shoehorn a bunch of level shifters in to my testing device.
|
|
#
?
Feb 7, 2017 07:22
|
|
|
Splode posted:Today I realised that the two RFID read and write modules I'd been using for the past few weeks are supposed to be powered by 3.3V rather than 5. Oops! I guess those chips are pretty tough. I do wish that shops advertising stuff as arduino compatible would at least make it clear that it's only good for 3.3V. Now I have to shoehorn a bunch of level shifters in to my testing device. Every single listing for every electronic item now has "arduino" stuck in the title somewhere because it's how you get hobbyists to buy them so I don't trust it at all anymore. Also a while back I put 5V through an ESP8266's data lines without realizing it for some time and was just wondering why this crappy chip got so darn burning-hot when it wasn't even doing anything, drat Chinese junk 
|
|
#
?
Feb 7, 2017 08:00
|
|
|
The switch from the 5V standard to 3V3 has been sneaky and destructive. The worst part is I know as soon as its finished displacing 5V it'll start getting beaten by 1.8V...
|
|
#
?
Feb 7, 2017 08:10
|
|
|
Zero VGS posted:Another dumb question from me: Almost definitely not. Charging lithium-ion cells in series is tricky even when you're not dealing with that many of them, this seems like a great way to explode yourself. Also where the hell are you getting a 312V battery? If it's a reasonable capacity to charge it in any sort of reasonable time you're going to need like a 20A charger or something.
|
|
#
?
Feb 7, 2017 08:43
|
|
|
312v li-ion sounds like a hybrid car or EV battery pack, which IMO is something you truly do not want to be messing around with if you're still fuzzy on stuff like voltage converters.
|
|
#
?
Feb 7, 2017 08:52
|
|
|
Zero VGS you're going to die. Not to discourage you, but please make sure you carry out your experiments with a fire extinguisher in arms reach. Lithium batteries are insanely dangerous, film everything just in case. Youtube Superstardom is just a housefire away. Splode fucked around with this message at 10:21 on Feb 7, 2017 |
|
#
?
Feb 7, 2017 09:15
|
|
|
Sagebrush posted:312v li-ion sounds like a hybrid car or EV battery pack, which IMO is something you truly do not want to be messing around with if you're still fuzzy on stuff like voltage converters. Believe it or not I've done both electric bike and electric car conversions before... But usually I'm just plugging batteries to an EV charger and controller, very straightforward so the knowledge I have from going to a vocational school for electromechanical was sufficient. My actual vocation for a decade has been IT though, so yeah I'm rusty. This time I have 14 Tesla batteries (24v, 75kw system total) and I'm trying to make an EV that is also a solar RV (think u-haul type box van with 2000w of pv panels on roof). So it needs to charge from both a public charging station and the solar panels. The motor controller runs in full series for 312vdc, and the main charge port is high voltage too, but the appliances in the RV would do better with the batteries in full parallel for 24vdc, as would the solar charger (solar charge controllers are only built to handle 12v - 48v pack ranges). No one on the planet seems to have done anything like that before, so I need to figure out whether I... a) run in full parallel for 24v and only step up the volts on the way to the motor controller, not sure how efficient that would be and the boost converter would need to be like like 2000 amps (!), but otherwise a much safer system overall at the cost of some range reduction due to efficiency loss from the huge voltage boost b) run in full series and step down to 24v for accessories (stepping up the solar charge controller output somehow, to charge the series pack) c) design some elaborate circuit of relays and switches that can safely toggle the entire battery pack between full series (when driving or charging from plug) or full parallel (when parked/camping and charging from pv array) So yeah this is considerably more complicated than my previous projects. It is not lost on me how easily a 312vdc pack can vaporize me. Thankfully, the Tesla packs have cell-level fuses and there's YouTube footage with people shorting them directly without a meltdown. But of course I'm still giving the pack the utmost respect and isolating them in a sealed compartment with liquid cooling and using additional fusing and contactors, linesman gloves and full face polycarb shield, maybe someone on standby for CPR 
|
|
#
?
Feb 7, 2017 15:27
|
|
|
Zero VGS posted:
Is there a reason you can't configure your solar panels in series to reach the higher voltage, and then use a proper EV charge controller designed for the battery voltage? You'll definitely want to avoid stepping up or down anywhere in your charge path.
|
|
#
?
Feb 7, 2017 15:47
|
|
|
KnifeWrench posted:Is there a reason you can't configure your solar panels in series to reach the higher voltage, and then use a proper EV charge controller designed for the battery voltage? You'll definitely want to avoid stepping up or down anywhere in your charge path. The more solar panels in series, the more susceptible to any kind of shading affecting the output of the entire pack. To reach a voltage that high I'd need the entire array in series. It is also much more efficient overall to use an MPPT charge controller as it can optimize the performance of the solar array, which a normal EV charge controller won't do at all. There's some solar race car MPPT charge controllers that go up to around 180vdc but that's as high as they seem to get.
|
|
#
?
Feb 7, 2017 16:12
|
|
|
Zero VGS posted:The more solar panels in series, the more susceptible to any kind of shading affecting the output of the entire pack. To reach a voltage that high I'd need the entire array in series. It is also much more efficient overall to use an MPPT charge controller as it can optimize the performance of the solar array, which a normal EV charge controller won't do at all. That makes sense. It's seeming more and more like you have a significant design roadblock that will need to be solved by an engineer who knows what they're doing. I don't know that the answer you need can be conveyed in a forum post. 
|
|
#
?
Feb 7, 2017 17:22
|
|
|
KnifeWrench posted:That makes sense. It's seeming more and more like you have a significant design roadblock that will need to be solved by an engineer who knows what they're doing. I don't know that the answer you need can be conveyed in a forum post. No worries, I figured I'd throw it out there in case there was a super-engineer lurking about who'd point me in the right direction. I am budgeting for like, paying actual engineers further down the line This is the solar MPPT I mentioned that goes all the way up to 160vdc for the battery, can be over 99% efficient too, so they claim: https://www.aerl.com.au/solar-racing-car-mppt.html I just found some more reading material on it, looks like some college kids are trying to reverse-engineer and open-source it: https://sites.google.com/a/onid.oregonstate.edu/ece44x201324/technology-review/1-1-introduction They also post a comparison chart I had never seen before: https://sites.google.com/a/onid.oregonstate.edu/ece44x201324/technology-review/1-2-technology-tables Interestingly, on their MPPT comparison chart, there is a Texas Instruments solar MPPT development kit that is rated for 200-300vdc in from the panels, and 400vdc out, at 500 watts max: http://www.ti.com/tool/tmdshvmpptkit If that TI MPPT is possible then it sounds like I can modify or build an MPPT to get the voltages I need. I'll study up on all the schematics. Maybe changing out the transformer on a working MPPT is all that will need to be done?
|
|
#
?
Feb 7, 2017 19:16
|
|
|
Wait by "motor controller" you mean the motor of the van? Yeah you're going to need a lot more help than something awful dot com fun comedy forums posters can probably give you.
|
|
#
?
Feb 7, 2017 19:18
|
|
|
Nah this forum is full of experts in every field. For the record, ZeroVGS, option B sounds the most likely (full series, step-down the voltage for peripheral devices using 24V, figure out how to get a high voltage solar charger). If the EV stations can safely charge the whole battery array at the series voltage, that's great, as now you know it can be done. The battery array presumably has circuitry to balance the voltage in each cell if that's the case. Stepping the voltage down from 312DC to 24DC is very easy, you can buy something for that assuming that the 24V circuit doesn't draw too much current. The solar panel thing is the hardest, but if you stack small solar arrays in series you can theoretically generate enough voltage. In summary look further into how EV charging works now. If they're charging all the batteries at the high voltage this might be doable!
|
|
#
?
Feb 7, 2017 21:12
|
|
|
Splode posted:Zero VGS you're going to die. I’d ignore the fire extinguisher and take advantage of a good pair of running shoes.
|
|
#
?
Feb 7, 2017 21:13
|
|
|
I'm curious if there's any cycling enthusiasts here. A video showed up on my FB purporting to show "mechanical doping," or "cheating with a motor." So, how hard would it be to detect a motor? Assume the bike's too expensive to chop up, have to do everything non-invasively. There are companies like Faraday Bikes that have a motor stuffed into the pedals, but they don't have to worry about detection and are free to use metal, magnets, etc.
|
|
#
?
Feb 7, 2017 23:14
|
|
|
JawnV6 posted:I'm curious if there's any cycling enthusiasts here. A video showed up on my FB purporting to show "mechanical doping," or "cheating with a motor." If you're not allowed to chop up the frame it's probably impossible to conceal. While you can hide most of your electronics in a water bottle, it's going to be difficult to pass the cables coming out of it and going to a mysterious box on the rear axle as brake lines or something. Doing it invasively should be possible, though concealing a motor of a useful size might be tricky even if you're allowed to chop stuff up.
|
|
#
?
Feb 7, 2017 23:23
|
|
|
|
| # ? May 9, 2024 13:27 |
|
|
Chop up for detection, presumably the bike creator is free to stuff tubes then weld them together. Or whatever you do to join carbon fiber. e: Like, can I detect a permanent magnet with a butter knife? Would you be able to fit an induction motor that wouldn't have any permanent magnetic field?
|
|
#
?
Feb 7, 2017 23:25
|
|