|
I appreciate it. I'm sourcing the displays from AliBaba, and haven't had issues. They're by Dalian GoodDisplay, and this is the 4.2" model. Available at Amazon and other Western realtors, including with a board/backing-circuitry branded as WaveShare. The conductivity sensor is connected to ground, and a DAC. (With intermediate circuitry including a PWM-driven polarity-switching circuit, and a multiplexer + gain_resistor.) I had already been disabling the DAC when not taking a conductivity reading, but evidently, the ⏚ connection is passing current through the pH or ORP probe terminals, since those readings go entirely bogus when ⏚ is in the water via the conductivity probe. (The other end I manage by setting the highest-resistance gain resistor, and disabling the DAC) I also added switches on the pH and ORP connections that go to the ADC and opamp output (Their other connection is opamp input, which I think/hope is high impedance). There is minor interference between pH and ORP probes, and unknown interference of the pH /ORP on conductivity readings.
|
# ? Aug 22, 2021 17:53 |
|
|
# ? May 21, 2024 15:24 |
|
This is a bit of blind leading blind because I certainly don't claim to be an expert, but I'm curious if the leads from the probe are picking up some kind of interference and inducing a current in the other sensors through the water (or whatever solution it is). Basically I have no idea what the solution is, but I would propose trying to eliminate all the variables you can and adding them back in slowly. So like, sensor with no other sensors test for a problem. Dip another sensor into the solution but leave the lead disconnected and see if there's a problem. Maybe even have another connector on your sensor module (maybe you already do) and try putting the sensor in there but without any cable connected and see if it causes the problem. It's tedious as hell, but I would guess those sensors are detecting not a whole lot of current to get readings to begin with and could be interfered with in tons of different ways. If I've learned anything by messing with crude radio circuits lately, it's that very minor things that I'd never have expected can cause big changes, e.g. am I sitting near the thing or not. I know you're not dealing with radios, but I think a sensor that is detecting very small electrical changes to deduce a measurement could have similar issues (if that is the operating principle of your sensor, I really don't know, just a hunch). So barring having someone with a lot of real world experience take a look at it, you might have to get crazy with your troubleshooting steps. Edit: Oh and I do think you're correct than an op amp should have a very high input impedance. But I've still seen people throw resistors on there that I wouldn't have thought necessary to reduce coupling to the other input or other stages and whatnot. I'm not nearly knowledgeable enough to tell you when that is and isn't necessary though. Forseti fucked around with this message at 18:10 on Aug 22, 2021 |
# ? Aug 22, 2021 18:06 |
|
Making a new post because I don't want to bother with imgur to attach this image. Oh, I remembered where I saw that resistor that I didn't quite understand the purpose of: http://www.labguysworld.com/VideoCookBook_002.htm quote:If you recall your op amp theory, you will see the amplifier is configured for a voltage gain of two. (R3 = R4) The purpos of R2 is to balance the currents in the plus and minus inputs to reduce introducing DC shifts in the video. Again, I couldn't tell you why that needed to be done unfortunately, just that I've definitely seen instances where people worry about things like that with op amps. I would have thought the 75 ohm to ground to match the input impedance would be enough, but thinking about it now, I guess if the op amp input impedance is Real Big, the signal could hit that "wall" and reflect back without a resistor there to dampen it and persuade it to go through the 75 ohm resistor to ground?
|
# ? Aug 22, 2021 18:53 |
|
The reason that's there is that the inputs of that op-amp use bipolar transistors and there are small currents going into the bases of the input pair. At steady state the current will be the same on both inputs, so matching the equivalent resistance on both sides (the 2 1k in parallel vs the 1 ~500-ohm) cancels out the input voltage shift caused by that base current.
|
# ? Aug 22, 2021 19:14 |
|
Stack Machine posted:The reason that's there is that the inputs of that op-amp use bipolar transistors and there are small currents going into the bases of the input pair. At steady state the current will be the same on both inputs, so matching the equivalent resistance on both sides (the 2 1k in parallel vs the 1 ~500-ohm) cancels out the input voltage shift caused by that base current. Oooooh, thank you! Very helpful and makes sense. Does the Real Big input impedance apply only to MOSFET based op amps or is the impedance of a BJT op amp still much greater than the input impedance to the base of a typical common emitter configured single transistor amp? I think that the op amp's configuration still results in a higher input impedance regardless of the transistor type they used, but all things equal a MOSFET based one would still have the higher impedance?
|
# ? Aug 22, 2021 19:24 |
|
Forseti posted:Oooooh, thank you! Very helpful and makes sense. The input current is small for BJTs (bias current over input beta), tiny for JFETs(diode reverse leakage) and practically zero for MOSFETs (MOS gate leakage). To add to the subtlety, the small signal resistance being very high doesn't necessarily mean the base current is very small since a near-vertical line at 5μA on the V-I plot still has a large slope! A lot of bipolar op-amp designs have also tried a lot of different techiques over the years for input base current cancellation, so it's not like it's universally true that there's significant base current, but matching the resistance like that can knock tens of dB off of an already small offset and resistors are very cheap.
|
# ? Aug 22, 2021 19:49 |
|
Forseti posted:This is a bit of blind leading blind because I certainly don't claim to be an expert, but I'm curious if the leads from the probe are picking up some kind of interference and inducing a current in the other sensors through the water (or whatever solution it is). The brings up the point of even if I make the sensors not interfere with each other, is the system robust against enough outside interference, like from submerged pumps?
|
# ? Aug 22, 2021 19:49 |
|
Stack Machine posted:The input current is small for BJTs (bias current over input beta), tiny for JFETs(diode reverse leakage) and practically zero for MOSFETs (MOS gate leakage). To add to the subtlety, the small signal resistance being very high doesn't necessarily mean the base current is very small since a near-vertical line at 5μA on the V-I plot still has a large slope! A lot of bipolar op-amp designs have also tried a lot of different techiques over the years for input base current cancellation, so it's not like it's universally true that there's significant base current, but matching the resistance like that can knock tens of dB off of an already small offset and resistors are very cheap. Great info, appreciate it! I'm definitely still in the phase of my learning where I need to hear or realize the same thing over and over in various contexts because I'm still frequently having great realizations where I'm like "Oh, that's because of X! Wait a minute... I knew that already from a different application, why didn't I realize that's what was going on here?" It's very helpful to be reminded of these things over and over again in different variations or contexts Dominoes posted:Could be. Of note, some of that experimentation is what led to the analog-switch-I-ordered approach. Ie, I confirmed that the main source of the conductivity readings breaking the others is its ground terminal being in the water. The radio analogy may be interesting, since the readings are indeed very sensitive to mV-order voltage changes. There are some of these tests I should conduct, like making sure the 20M resistor + switched off DAC doesn't interfere with readings; likewise with the amp-input lead from the pH/ORP probes. I guess abstractly I was just trying to say that it may be helpful to eliminate all the variables you can and reintroduce them slowly, even if you're pretty sure they're not affecting it because we can become myopic when we're very familiar with the problem. I remember at my first job I was working on the desktop application for reading data from our product and had spent a lot of time and effort getting something I was confident to send over to my coworker for testing. He broke it pretty much immediately, and when I investigated the bug he found, that code hadn't been changed in like six months. I was baffled that in all the time I'd worked on it and tested it, I'd never done what he did to break it because it wasn't like it was something unreasonable or anything, his workflow was just slightly different from mine.
|
# ? Aug 22, 2021 20:01 |
|
Dominoes posted:The brings up the point of even if I make the sensors not interfere with each other, is the system robust against enough outside interference, like from submerged pumps? It's a way to transform an O(1) problem into an NP-hard problem. If you can solve it, I hear there's a prize or something. Good luck getting your algorithm to run in Rust on an STM microcontroller; there are supply chain problems right now.
|
# ? Aug 22, 2021 20:11 |
|
babyeatingpsychopath posted:Sounds like you just need robust machine learning to pull the signal from the noise even in the face of arbitrary interference. Was this post written using machine learning
|
# ? Aug 22, 2021 20:31 |
|
I'm bootstrapping; no need for VC funds. Btw, for anyone considering Rust: It's a lovely language, and is perfect for embedded. However, much of the OSS libraries, and generally what you find online, is over-abstracted garbage.
|
# ? Aug 22, 2021 20:46 |
|
Speaking of high input impedance op amps, I got hooked on the LMC660/LMC662 because it was hyped on the EEVblog metrology forum. The 2 femtoamp input bias current at room temperature does not seem to be a lie and the unit cost is low. The 3 mV offset voltage and 1.4 MHz bandwidth aren't astounding but isn't a big deal for current measurement applications, though the offset voltage risks drowning out measurement of tiny voltages.
|
# ? Aug 22, 2021 22:09 |
|
Forseti posted:I'm starting a little project to get a little better acquainted with RF circuits this weekend. Months ago I made a very crude attempt to see if I could transmit the already constructed signal coming out of my NES' RF modulator adapter across the room and receive it on my TV's antenna. I basically attempted to jam it directly into a general purpose NPN (a MMBT3904) configured as a basic common emitter amplifier with resistors to set the bias and such. As you might expect, it didn't really do much although I could see the noise change when it was turned on and also pick it up a bit with an antenna hooked to my oscilloscope. Yeah, you should try to design impedance matching circuits for your transistor device--I think it would be instructive. I suspect that 6 MHz of bandwidth at 60 MHz is not too broadband to make the matching circuits too complicated. I think at 60 MHz the transmission line effect of the circuit board traces and RF connectors will not dominate the result. Measure your transistor with the bias circuit on your nano-VNA, and use graphical techniques with shunt or series inductors/capacitors on the input match and the output match of the device+bias circuit on the Smith Chart to match the device for good small-signal gain. In principle, you could iteratively match the device experimentally, using the match information you get from the VNA. If you have some kind of circuit simulation software where you can plug in an .s2p Touchstone file measurement from your VNA which would represent the small-signal performance of the device+bias circuit, you can do a lot of the tuning in the circuit simulation software, and get an intuition of what kinds of inductor values and capacitor values you need, and what kind of matching circuit would likely work the best, before you try stuff out on your test board. silence_kit fucked around with this message at 14:46 on Aug 23, 2021 |
# ? Aug 23, 2021 14:07 |
|
silence_kit posted:Yeah, you should try to design impedance matching circuits for your transistor device--I think it would be instructive. I suspect that 6 MHz of bandwidth at 60 MHz is not too broadband to make the matching circuits too complicated. I think at 60 MHz the transmission line effect of the circuit board traces and RF connectors will not dominate the result. Thanks for the input! I did actually spend the weekend studying up more on impedance matching, so it would be nice to give it a try. Especially since I have the equipment to actually see what I'm doing a bit better. I didn't know you could measure params with the NanoVNA for use in simulations, that is an awesome feature! Open source tools that get wide adoption always amaze me with things people come up with. Like I understand it's a bit limited with a low sample rate and using harmonics (at least in the variant I have), but there's so much info on how to use it and people are still cranking out new tools to make it even better. I haven't had it for long and I've never used a real network analyzer, but I'm very impressed so far. I did end up installing Wine and LTSpice this weekend (which was really straightforward, I just never had gotten around to it before). Mainly because all my components are surface mount and it would be nice to get things pretty close before going through the trouble of putting them on a board. I'll look around and see what people are using for desktop software to integrate with the NanoVNA, ideally one that is available in Linux natively, although Wine works pretty drat well these days. And just to clarify my own understanding, a purely resistive matching network is still an impedance matching circuit, right? I know it also attenuates the signal, but it would still prevent reflections from a mismatch I think? So I guess in other words, the input device would still see the proper match? Are these just referred to as attenuators and not impedance matching networks in common use?
|
# ? Aug 23, 2021 15:54 |
|
Forseti posted:I didn't know you could measure params with the NanoVNA for use in simulations, that is an awesome feature! Oh I don’t know if the nano-VNA has this feature, but many VNAs have this feature. It sounds like you are a computer guy, so if the nano-VNA didn’t have that feature, and you REALLY REALLY wanted to do this, you probably could find some way to digitize the information on the display. Forseti posted:And just to clarify my own understanding, a purely resistive matching network is still an impedance matching circuit, right? I know it also attenuates the signal, but it would still prevent reflections from a mismatch I think? So I guess in other words, the input device would still see the proper match? Are these just referred to as attenuators and not impedance matching networks in common use? Yes, you can use attenuators to improve the match of a device. Since the reflected signal will first go through the attenuator, reflect off of the device, and come back, you get 2dB of return loss improvement for every dB of attenuation you add. Also, resistors are sometimes added to purely reactive matching circuits to help make the matching problem easier, at the cost of some signal loss, and/or to shape the response of the device with frequency without using intentional mismatch loss.
|
# ? Aug 23, 2021 17:48 |
|
Can I get some "is this realistically possible" advice for my silly project? I got this basic layout planned to go into a case: The usb 3.0 cables currently in use are really stiff and requiring me to use 90 degree adapters and some forceful cramming to get it to fit into the space available. So my genius plan is to desolder the usb 3 connectors from the pi and attach my own wires. The white one is an SD card reader and will solder directly to the pi. The black one leads to the exterior of the case and will be a female plug to connect a usb hard drive. My concern is the shielding. How sensitive are usb 3 speeds to getting everything perfect? Will running what's basically an extension cable inside the case to a socket in which another usb cable plugs in to work okay? What kind of tolerances do I have on the shielded pairs being unshielded? I've done some searching around on this and the best I've got is "as close as possible" which is not super precise. I found a usb-c port on digikey that has terminals I think my clumsy rear end hands will be able to solder, plus it has screw holes for mounting. The price is pretty hilarious though, so if someone knows of a better option I'd love to hear it. It doesn't HAVE to be usb-c, but that's what my ssd drives all use so I figure it's more usable long term if I go that route. But as I type this I'm wondering if maybe I should just go with usb A. https://www.digikey.com/en/products/detail/MUSBRM5C130/MUSBRM5C130-ND/5877552?itemSeq=374774032 (the pi power port is stuck with that 90 degree adapter, the contacts on the circuit board are way too small for me to attempt soldering them)
|
# ? Aug 24, 2021 22:22 |
Hi I wanted to show off another cool find. To begin with, I was just looking for really cool enclosures to make project kits out of. Problem was, I keep coming to the realization that the box and its contents are too valuable to be separated. This appears to be another such case. This cool gadget comes in a humongous alloy case, same case style as my last cool project but far bigger. And I really loved the art once you open this seamless, watertight box. Then I went to look up what the heck I had here and discovered this is apparently the transformer box for a mfing whack invention, a static-electricity based flocking machine, built for large scale operations such as the inside of a small boat or an airplane, I dunno. The process uses adhesive and highly charged fibers are shot like death rays onto every surface the gun points at. I have no intention of plugging this baby in but if I understand correctly here, I have probably stumbled on a very high voltage transformer inside this box, among other things. I wonder what extracurricular use something like this might have, if any?
|
|
# ? Aug 25, 2021 14:49 |
|
I have an amusing image in my head of someone dual-wielding flocking guns and laying absolute fuzzy waste to their enemies.
|
# ? Aug 25, 2021 15:03 |
|
petit choux posted:Hi I wanted to show off another cool find. To begin with, I was just looking for really cool enclosures to make project kits out of. Problem was, I keep coming to the realization that the box and its contents are too valuable to be separated. This appears to be another such case. This cool gadget comes in a humongous alloy case, same case style as my last cool project but far bigger. And I really loved the art once you open this seamless, watertight box. It looks like flocking guns use voltages in the 10s of kilovolts so that power supply would theoretically be useful for electrohydrodynamic pumps (pumps with no moving parts) and things based on that sort of principle like "ion lifters"/"ionocraft" which use ionization-induced fluid motion to generate thrust like rockets. I wired up a CRT transformer to make arcs with and it's pretty interesting how you can get surprisingly-strong air motion even if you don't want it to happen on purpose. I found out that the small amount of ionization in the flame of a candle can allow arcs to happen when the electrodes are just a bit too far apart for it to happen on its own but you have to play with the candle position and angle or the ion wind can nearly blow it out. Of course don't play with high voltages unless you know what you're doing. I did EHD stuff for my undergraduate engineering capstone so I had training and a background, and I still think that doing it at home isn't really a great idea.
|
# ? Aug 25, 2021 15:28 |
|
xzzy posted:Can I get some "is this realistically possible" advice for my silly project? It doesn't help your data lines but you can feed the Pi power via the GPIO pins instead of the USB port if you wanted to get rid of one 90 degree adapter.
|
# ? Aug 25, 2021 15:40 |
|
xzzy posted:I found a usb-c port on digikey that has terminals I think my clumsy rear end hands will be able to solder, plus it has screw holes for mounting. The price is pretty hilarious though, so if someone knows of a better option I'd love to hear it. It doesn't HAVE to be usb-c, but that's what my ssd drives all use so I figure it's more usable long term if I go that route. But as I type this I'm wondering if maybe I should just go with usb A. Rather than a bare connector you can get sockets with small breakout pcbs pre-soldered or short cables that you could strip the wires on the pi side
|
# ? Aug 25, 2021 17:09 |
|
Man I loving hate hack-a-day's userbase. I just read a comment blaming the US opiate epidemic on the availability of cheap thermal cameras from China that allow the narcoguerilla heroin traffickers to evade the border guards.
Sagebrush fucked around with this message at 23:27 on Aug 25, 2021 |
# ? Aug 25, 2021 23:22 |
|
The internet's a scary place. Even on SA, a number of the subforums are a disaster.
|
# ? Aug 25, 2021 23:39 |
|
Sagebrush posted:Man I loving hate hack-a-day's userbase. I just read a comment blaming the US opiate epidemic on the availability of cheap thermal cameras from China that allow the narcoguerilla heroin traffickers to evade the border guards. I feel like that's the kind of worldview you can only get if your brain's been steeped in Elon Musk and LessWrong until it looks like a century egg. Once you strip out all the subjective imaginary ~feelings~ and ~politics~ it's apparent that all the world's problems are objectively caused by "Bad Guys" and "Cool Technology I Read About Once". There is literally nothing else in the universe worth considering and you're mind killed for thinking otherwise.
|
# ? Aug 26, 2021 00:02 |
BattleMaster posted:It looks like flocking guns use voltages in the 10s of kilovolts so that power supply would theoretically be useful for electrohydrodynamic pumps (pumps with no moving parts) and things based on that sort of principle like "ion lifters"/"ionocraft" which use ionization-induced fluid motion to generate thrust like rockets. Yeah, this was quite a lucky find, I don't know what to do with it but it looks like it has some potential. Imagining it in a murder mystery production, somebody gets flocked to death. Yeah, Imma exercise extreme caution with anything like this, don't plan on doing anything at all with it RN. Woo hoo!
|
|
# ? Aug 26, 2021 00:42 |
|
petit choux posted:looks like it has some potential. heh
|
# ? Aug 26, 2021 00:46 |
|
petit choux posted:Imagining it in a murder mystery production, somebody gets flocked to death. I think that was a plot point in the 2001 movie Cats and Dogs
|
# ? Aug 26, 2021 00:46 |
|
BattleMaster posted:Of course don't play with high voltages unless you know what you're doing. I did EHD stuff for my undergraduate engineering capstone so I had training and a background, and I still think that doing it at home isn't really a great idea. i really don’t want to go anywhere near high voltage circuits. this must be offset against my desire to build and operate a birkeland-eyde spark gap reactor. the Poison Generating Machine will inevitably win the struggle against my better judgement, but that day has not yet come. in the meantime i will enjoy my youthful ”normal lung capacity” and “limited neurological damage”
|
# ? Aug 26, 2021 01:21 |
|
Ambrose Burnside posted:i really don’t want to go anywhere near high voltage circuits. this must be offset against my desire to build and operate a birkeland-eyde spark gap reactor. the Poison Generating Machine will inevitably win the struggle against my better judgement, but that day has not yet come. in the meantime i will enjoy my youthful ”normal lung capacity” and “limited neurological damage” Good idea. I work with some high power systems (>10MW, >10kV) and we have to go through training courses every two years and refresher courses every 6 months, plus maintain proper gear and PPE. The training courses always include a section of gross videos and pictures meant to beat into your head that electrical accidents happen quickly and there will be no time to react or takebacks on the damage done. It always scares the poo poo out of me to think of someone trying to even work on their house mains with one of those crappy Harbor Freight meters and no PPE. USA electrical safety standards are not great (they accept 2nd degree burns and partial loss of fingers) but at least better than nothing. They list a 50V max differential voltage or 50V off of earth ground as being the limits of 'safe' stuff to work around with no PPE. Above that you get into 'shock' territory which is what most people are familiar with as electrocution where your muscles seize up and your heart can stop, even hours after the event. Next up is 'arc' where you ignite a burning plasma ball, this can happen at >250V under normal atmospheric pressure and temperature and you have to start wearing fire retardant suits and full face shields. Then there is 'blast' where the arc is so large that the pressure wave throws you physically across the room, no amount of PPE will help you there. For those situations we have a thing we call 'the robot' that we set up in front of a high power switch and activate remotely (it's boring, it just turns a worm screw that opens/closes the contactors). On a good day I'm at my desk messing with 3.3V control logic but when poo poo really breaks down we gotta put on our 40 cal/cm^2 bomb suits like this one (not shown: internal cooling packs and filtered air pumps to keep you from inhaling copper plasma) and hope you don't end up in a gross training video.
|
# ? Aug 26, 2021 03:45 |
|
I could use some guidance on building 12 volt battery packs out of 18650 Li-ion batteries. I am aware that I need to wire 3 in series and uses a BMS board of somekind...but I literally just learned about this and would like to know the best practices. If somebody could point me towards a tutorial of some kind that would be great. My use case is for 12 volt audio amplifiers. I am stepping up my boombox game and need to make use of 12v amps, but don't want to weight down the project with a lead ion 12 volt battery. edit: also, we really need to start a bereavement flowers fund. Some of you guys are getting wild.
|
# ? Aug 26, 2021 03:51 |
|
Also, since I'm rambling on about high power systems, we have a special type of fuse on the high power stuff that includes an integrated explosive charge. The idea that if the fuse blows for any reason it likely has a highly conductive plasma ball remaining between the electrodes, the system detects that and sets off the explosive to disperse the plasma and interrupt the fault current. The enclosure for these fuses is designed to fall apart and vent the pressure without flinging metal bits all over the place. The explosive fuses have gone off one time since I've been at this job and hoo boy they are not cheap to replace. You can't exactly ship explosives via Fed Ex so a specialty shipper has to bring in the replacements.
|
# ? Aug 26, 2021 04:04 |
|
PDP-1 posted:Also, since I'm rambling on about high power systems, we have a special type of fuse on the high power stuff that includes an integrated explosive charge. The idea that if the fuse blows for any reason it likely has a highly conductive plasma ball remaining between the electrodes, the system detects that and sets off the explosive to disperse the plasma and interrupt the fault current. The enclosure for these fuses is designed to fall apart and vent the pressure without flinging metal bits all over the place. I know BigClive took apart a fuse that just straight up had a regular ol' blank cartridge for a gun in it to ensure the two ends of the connection were blown as far apart as possible. Can't for the life of me find the video, though I remember him using a Dremel to get at the part with the explodey bit in it which was very anxiety-inducing. e: Here we go https://www.youtube.com/watch?v=7hvg2Wey92E It wasn't a fuse, it was a lightning arrestor, but same idea. e2: Oh I forgot, he doesn't find the actual blank until an update video: https://www.youtube.com/watch?v=3G6NXge8DAE Shame Boy fucked around with this message at 04:26 on Aug 26, 2021 |
# ? Aug 26, 2021 04:15 |
|
Marsupial Ape posted:I could use some guidance on building 12 volt battery packs out of 18650 Li-ion batteries. I am aware that I need to wire 3 in series and uses a BMS board of somekind...but I literally just learned about this and would like to know the best practices. If somebody could point me towards a tutorial of some kind that would be great. I don't know if they have answers to those specific questions, but I have found the aptly named Battery University to be a great beginner's guide as well as refresher reference. Learn enough about the way the individual cells work, and you'll better understand what it is you want the BMS to do for you.
|
# ? Aug 26, 2021 04:47 |
|
KnifeWrench posted:I don't know if they have answers to those specific questions, but I have found the aptly named Battery University to be a great beginner's guide as well as refresher reference. I'll take a look, thank you. I know what I want: 12 volts! I think I have it figured out, I just have to determine which Chinese BMS board is the least bad. On a some what related/unrelated note, I do actually try to look up this stuff on YouTube before I ask real simple questions here...but it's almost all Hindi speaking guys. A handful of them are very good and I can suss out what they are doing, but 99% are just bad. Like, how have you not burned your house down, bad. Anybody else run into this? I think the DIY bluetooth speaker scene is just super popular with Hindi speaking electrical engineering students.
|
# ? Aug 26, 2021 06:12 |
|
Look up ebike battery building? Battery packs are battery packs, the one you'll build is the same, just smaller.
|
# ? Aug 26, 2021 06:28 |
|
PDP-1 posted:Good idea. the saving grace with spark gap/open-arc processes, from my weird-rear end perspective, is that an arc in plain air, any ol arc, can be all you need to party. with nitrogen-fixing arc reactors, you increase productivity by boosting the amperage or getting creative with magnetic / microwave plasma shaping, jacob’s ladder arrangements etc- most DIY birkeland-eyde reactors draw a couple of mains-voltage amps to get anything done on a timescale of hours vs days- but any spark will do, even really low-current arcs, it’ll just do its thing at its own sedate pace anyways, this is all to say that i’ve been having powerful and stupid thoughts about turning the front wheel of an exercise bike into a manually-pedalled wimshurst wheel electrostatic generator use electrode spacing and maybe supplementary circuits to encourage many frequent small sparks instead of infrequent but long/showy arcs when the wheels are turned, and use this arrangement to generate actual tangible work from my working out; that is, to very gradually turn a beaker of water into nitric acid, by fixing atmospheric nitrogen into nitric oxide n no2 all lightning-style and then bubbling it through water to steadily convert it to acid. could even have the pedalling action pump the necessary continual airflow through the arc into the reactor via a simple peristalsis-tubing setup, have it be entirely human-powered Ambrose Burnside fucked around with this message at 08:01 on Aug 26, 2021 |
# ? Aug 26, 2021 07:11 |
|
ante posted:Look up ebike battery building? Battery packs are battery packs, the one you'll build is the same, just smaller. That's a good idea, thanks. Ambrose Burnside posted:
Ok, so what we do at the county water utility is kick in a dollar out of every paycheck towards the flowers fund. I think we can work out something like that, here.
|
# ? Aug 26, 2021 07:39 |
|
Anybody good with op-amps? I'm trying to get a transimpedance amplifier circuit working with a photodiode package I have. I'm currently running the following circuit: Which I've now realised is buggered and not working: The photodiode array I'm using has a common cathode and so I have to apply my bias voltage to the cathode, whereas a lot of example circuits I've seen have it the other way around. Also unfortunately I've had some PCB's made up and ideally I could find a hacky solution that would let me use the existing PCB's for now with a few bodge wires or cut traces vs. having to re-make them. Can anybody see a good way to fix the above circuit the the minimum of re-wiring? I've tried this layout but I also had issues, correct me if I'm wrong but in the following schematic it's also a requirement that Vbias >= Vout_max otherwise current will flow the wrong way through the photodiode?
|
# ? Aug 26, 2021 10:29 |
Blackhawk posted:Anybody good with op-amps? I'm trying to get a transimpedance amplifier circuit working with a photodiode package I have. I'm currently running the following circuit: This one works because the op amp model is ideal and doesn't have a finite supply voltage to limit it. But it shows that the output voltage will be negative, which is why it needs a negative supply. quote:There's no one right way to wire it up. Depends on whether you want the output voltage to have a bias on it, whether you want it to increase or decrease vs light exposure, what response time you want, etc. Also the photodiode datasheet should recommend a reverse bias voltage (or zero bias).
|
|
# ? Aug 26, 2021 13:08 |
|
|
# ? May 21, 2024 15:24 |
|
mewse posted:Rather than a bare connector you can get sockets with small breakout pcbs pre-soldered or short cables that you could strip the wires on the pi side This was super helpful, thanks. I really gotta teach myself to scan aliexpress when looking for weird stuff.
|
# ? Aug 26, 2021 15:52 |