|
ante posted:I built an amp for my car, but the left speaker only gives off a buzz. If I touch its filter cap, it starts working again. Do you have an oscilloscope, or can you otherwise figure out what frequency this ringing is happening at? That can tell you a lot about where your instability is. Also, I'm assuming you're using a NFB configuration. Do you have low pass filters in your gain loop? Is everything properly grounded? Are you using a virtual ground, and if so how? How are you handling ground from your input (ie tying it to your amp ground, or using it to drive your virtual ground)?
|
#
?
Jan 31, 2010 07:04
|
|
|
|
| # ? Jun 11, 2024 07:53 |
|
|
clredwolf posted:Yes. Invest in some power resistors before you start testing on/blowing up SLAs though. quote:That'll work. Most SLA chargers I've seen don't actually use the temp. sensor. A switching supply will use less energy, however I'd suggest looking at getting that as a self-contained module and NOT a DIY, switching supplies are not for the feint of heart. quote:Battery? Transformer off the A/C line? I mean you're not gonna get a thunderbolt to the head for having two plugs for this type of thing (just buy another power strip). quote:Three words: isolated sense resistor. Look at using a sense resistor, and an isolated amplifier like an iso 124 (expensive, but worth it for not shocking yourself). Some really rugged opamps can do it too, but you'll need to still isolate it somehow. I've got some 32gauge magnet wire and a couple of 1.000 ohm precision resistors. Is winding my own CT a good solution? Do CTs work on DC? quote:Note sure here, there's a lot of temperature sensors out there though. Sometimes you can work with funky configurations until something falls out that works well (whetstone and kelvin bridge configurations can be extremely accurate). quote:Overvolting the batteries, reverse polarizing the batteries (it WILL happen, how will you fix it?....hint: it can be a pretty easy fix), shorts...any time you work with things over 5-12 volts you have to be mindful of what you are doing. Respect the machine, for it can be dangerous. I'm all about safety. I've got some DIN-rail 120V AC/DC breakers available for pretty cheep. I'm a commercial electrician as my day job. 48VDC at 5A is nothing compared to 480VAC @ 2000A, but that doesn't mean it's safe to play with. You can weld fairly easily with one of these small batteries. The last concern is cost: A 48V smart SLA charger costs about $100-200. I'd like to get this done for around $20 or so. I've already got the PIC and a few precision resistors. I'll scrounge for power resistors for testing, and scrounge for things like filter caps and the like. I've got a k-type thermocouple available, but it doesn't look like it's able to directly interface with a PIC. I think I can also dig up some opamps from some old projects, maybe even a couple of SPI ADC/DACs. babyeatingpsychopath fucked around with this message at 00:01 on Feb 1, 2010 |
|
#
?
Jan 31, 2010 19:51
|
|
|
quote:Any ideas on how to get a 48V signal into a 3.3 or 5V PIC besides a voltage divider? There's nothing wrong with using a resistor divider with low-tolerance resistors of a known value. That's how the High-Low Voltage Detect module in some high end PICs works. If you know what the reference voltage is and you know how the input voltage is being scaled by the divider you can easily figure out what the input voltage is even if it's above the reference voltage. Just make sure to select resistor values that will ensure that the voltage at the ADC pin never goes over the PIC's VDD during normal operation, and maybe even a little overcompensation might be advisable. Edit: Depending on the resistor values you use you may want to use an op amp as a buffer so that the ADC input impedance doesn't go exceed the recommended maximum which can gently caress with the accuracy. quote:What's the best (1 degree accuracy) temperature sensing solution for under $8? An LM35 will do just fine. Accurate to 1 degree Celsius, and definitely under $8. It will interface directly with an ADC with no external parts. The only problem is that it is a 5 volt minimum part so you'll probably want to use a 5 volt PIC to avoid adding a second voltage regulator to the design. quote:I think I can also dig up some opamps from some old projects, maybe even a couple of SPI ADC/DACs. You may want to buy some cheap rail-to-rail op amps rather than use old ones with huge output swings, depending on what you want to use them for. For instance, if you do need to use a buffer on the resistor divider you don't want to risk misreading the input voltage because the crappy op amp can only output like VDD-3 volts. Also there's probably no need to use an external ADC because most PICs have one built in except for the really old ones or some of the 6 and 8 pin ones. BattleMaster fucked around with this message at 20:22 on Jan 31, 2010 |
|
#
?
Jan 31, 2010 19:56
|
|
|
babyeatingpsychopath posted:The software side is easy; I'm looking for specific hardware advice. These are 20Ah batteries, and my charging guide says no more than 30% capacity for current charge. That means 6A max charge, but I'm thinking probably 3A or so to prolong life. quote:I meant how do I get string voltage back to the PIC? I stop constant-current charge when voltage is at X and move to constant voltage mode. Is a precision voltage divider sufficent, or is there a better way? I've gotta get something in the 50V range to work with a PIC's ADC. quote:I've got some 32gauge magnet wire and a couple of 1.000 ohm precision resistors. Is winding my own CT a good solution? Do CTs work on DC?
|
|
#
?
Jan 31, 2010 20:16
|
|
|
yo anime, I got a question. I'm making a circuit using this bad boy (or will spec something very similar) http://www.avagotech.com/docs/AV02-0886EN I'll be using a basic topology like the one in 12b. I would choose the resistor values based on the desired LED current, but how would you choose the capacitor values? does it effectively form an RC filter? I know people like to chuck on these kind of bypass-like caps for unity gain followers, how are they chosen?
|
|
#
?
Feb 1, 2010 06:22
|
|
|
What's the trick to using an ADC and DAC with audio signals when the circuit only has a single positive supply? I've got a PIC with ADC, an SPI DAC, and an LM386 audio amp. I'd like to have the audio signal digitized by the ADC and recreated using the DAC but with the way I have it set up it only outputs varying high-frequency whines.
|
|
#
?
Feb 1, 2010 23:18
|
|
|
BattleMaster posted:What's the trick to using an ADC and DAC with audio signals when the circuit only has a single positive supply? I've got a PIC with ADC, an SPI DAC, and an LM386 audio amp. I'd like to have the audio signal digitized by the ADC and recreated using the DAC but with the way I have it set up it only outputs varying high-frequency whines. Virtual ground, generally. Although a lack of virtual ground shouldn't cause a whine, but rather really bad distortion. You could narrow down the location of the problem by outputting a known audio signal (say, a 440Hz sine wave) through the DAC. If it works, the problem's on the ADC side, if it's still a high-pitched whine, problem's on the DAC side.
|
|
#
?
Feb 2, 2010 01:15
|
|
|
catbread.jpg posted:yo anime, I got a question. quote:I know people like to chuck on these kind of bypass-like caps for unity gain followers, how are they chosen? BattleMaster posted:What's the trick to using an ADC and DAC with audio signals when the circuit only has a single positive supply? I've got a PIC with ADC, an SPI DAC, and an LM386 audio amp. I'd like to have the audio signal digitized by the ADC and recreated using the DAC but with the way I have it set up it only outputs varying high-frequency whines. ANIME AKBAR fucked around with this message at 03:57 on Feb 2, 2010 |
|
#
?
Feb 2, 2010 03:53
|
|
|
Mill Town posted:Virtual ground, generally. Although a lack of virtual ground shouldn't cause a whine, but rather really bad distortion. I looked into that and it's really handy to know. I can think of a few project ideas where it would save me a lot of trouble. Unfortunately the ADC and DAC I am using can't accept negative voltages as the lower reference so that's out. ANIME AKBAR posted:AC couple the input to half the supply voltage, then AC couple it back to ground on the output. However that helps a lot and I've got it working though it's fairly noisy - probably because of the lovely breadboard, or maybe I need to tweak the sampling time on the ADC. Edit: Or it could be that the DAC wasn't meant for audio reporduction because it's silent when the program isn't running. Maybe I should use that dsPIC I have laying around with a built-in audio DAC. Thanks guys. BattleMaster fucked around with this message at 06:32 on Feb 2, 2010 |
|
#
?
Feb 2, 2010 06:05
|
|
|
BattleMaster posted:However that helps a lot and I've got it working though it's fairly noisy - probably because of the lovely breadboard, or maybe I need to tweak the sampling time on the ADC. Edit: Or it could be that the DAC wasn't meant for audio reporduction because it's silent when the program isn't running. Maybe I should use that dsPIC I have laying around with a built-in audio DAC. make sure you preamplify the input enough so that it uses up a decent amount of the input range of the ADC. Otherwise your effective resolution will be low and you'll get more quantization noise.
|
|
#
?
Feb 2, 2010 06:48
|
|
|
ANIME AKBAR posted:make sure you preamplify the input enough so that it uses up a decent amount of the input range of the ADC. Otherwise your effective resolution will be low and you'll get more quantization noise. Oh I never thought about that, good idea. I'd like to use the minimum number of analog components, though, so maybe I'll use reference voltages closer to 1/2 VDD.
|
|
#
?
Feb 2, 2010 09:53
|
|
|
BattleMaster posted:Oh I never thought about that, good idea. I'd like to use the minimum number of analog components, though, so maybe I'll use reference voltages closer to 1/2 VDD. most AD converters are meant to receive a reference of precisely half the supply voltage, so there's nothing exotic about it.
|
|
#
?
Feb 2, 2010 14:31
|
|
|
The PIC ADC has an upper and lower reference so if I make them say 2/3 and 1/3 respectively it should have a similar effect on the result to amplifying it without actually amplifying the signal.
BattleMaster fucked around with this message at 17:28 on Feb 2, 2010 |
|
#
?
Feb 2, 2010 17:25
|
|
|
I'm looking at making a custom vaporizer - a heating element (ceramic?), temperature control and all that jazz. So far, because of the temperatures (~180C) required, it appears that I'll need to use a thermocouple if I want to measure and display temperature. If I just want to control the temperature (via a rheostat?), how does that work? How are they calibrated, manually? I'll end up posting a project thread about this, but since it's mostly electronic for now I figured I'd check here.
|
|
#
?
Feb 2, 2010 17:52
|
|
|
I so I have this: The gap between the 6v and the shift register will have a block of wood with metal contacts on it that when pushed down will conect the circuit together (I'm making a tic-tac-toe game thingy). Will I need any current limiting resistors and/or de-coupling caps?
|
|
#
?
Feb 2, 2010 18:53
|
|
|
Input pins on digital ICs have their own internal current-limiting resistors, but you'll need a resistor pulling each pin to ground when there's no contact or else weird things will happen. As for decoupling caps, if you're asking if you need them at all, yes. if you're asking if you need more than the usual one or two per IC, no.
|
|
#
?
Feb 2, 2010 18:59
|
|
|
Anybody have any experience with conductive pens? I have an smd jumper I need to trip, and I would rather not have to solder it. Using some conductive ink seems like the best option.
|
|
#
?
Feb 2, 2010 21:30
|
|
|
Just solder it. Get a big ol blob on the tip of the iron, let it sit for a sec to burn off the flux (this is the one time flux is not your friend), then touch it to the two pads - Instant jumper.
|
|
#
?
Feb 2, 2010 21:33
|
|
|
ignore this
ante fucked around with this message at 21:58 on Feb 2, 2010 |
|
#
?
Feb 2, 2010 21:52
|
|
|
Twozzok posted:I so I have this: At lowish frequencies, you're not going to get enough noise or whatever to really need decoupling caps. But yeah, pins can't be floating. Here's one way to do it  That resistor is to limit current to ground. ICs are high impedance, as the dude above me said, so they don't need limiting. Solaron posted:I'm looking at making a custom vaporizer - a heating element (ceramic?), temperature control and all that jazz. So far, because of the temperatures (~180C) required, it appears that I'll need to use a thermocouple if I want to measure and display temperature. If I just want to control the temperature (via a rheostat?), how does that work? How are they calibrated, manually? Should be ok right from the factory. If you want to really half-rear end it (like I would), you can probably get a $20 ceramic heater and pull it apart. Don't burn your house down or anything
|
|
#
?
Feb 2, 2010 21:57
|
|
|
ante posted:At lowish frequencies, you're not going to get enough noise or whatever to really need decoupling caps. He probably doesn't need nearly that strong of a pull-up. I would recommend something larger, like 47K ohms or so. Also, I always do pulldowns but I can't think of why. Is there any difference between these two? In hindsight, this seems safer with a pull-up as your button should stay at ground potential.
|
|
#
?
Feb 2, 2010 22:53
|
|
|
I don't know what the advantages are to either pulling up or pulling down in that case. I use pull ups if I need an active low button press or pull downs if I need an active high one.
|
|
#
?
Feb 2, 2010 23:04
|
|
|
It really depends on your application as to what is "best". For TTL inputs, pull-downs will need to be lower resistance and will pull more current since these inputs usually default to high. On the other hand, pull-ups can possibly add noise from the power supply attached to them. Either way, 99% of the time these considerations won't matter and like BattleMaster said, you should just pull the signal the opposite way of active.
|
|
#
?
Feb 3, 2010 02:29
|
|
|
Hillridge posted:Just solder it. I ended up using a conductive pen. I forgot I even posted ITT. It actually worked very well. It is a very thin liquid, so you have to be careful that it doesn't run, but I just drew a line between the pads and that did it. It can be cleaned up with alcohol or acetone if needed.
|
|
#
?
Feb 3, 2010 07:29
|
|
|
Is anyone interested in Nixie tubes? As I mentioned earlier in the thread, a coworker has 4 Burroughs B-7971 Alphanumeric Nixie Tubes with sockets that he wants to sell for $100. I'm looking for a new truck right now, so I don't think I'm going to buy them, but they are in great shape and all the segments work. He said he bought them new (this guy is old) and never used them. The sockets were cut out of something, but work fine. Here's the data sheet: http://www.akh.se/tubes/htm/b7971.htm These things are huge and would look really cool as a clock or a 4-letter word device. I may put up an ad on awfulmart, but I figured anyone interested would probably be in this thread.
|
|
#
?
Feb 3, 2010 16:41
|
|
|
I found out that the horrible noise on my ADC-DAC project was not caused by anything on my end but is actually just white noise from the crappy onboard audio in my PC. When I connect it to other things like my DS or a CD player it sounds great. Well, great for 24KHz 10-bit audio. Pretty decent considering the DAC doesn't even list audio as one of its intended uses.
|
|
#
?
Feb 3, 2010 20:49
|
|
|
Hillridge posted:Is anyone interested in Nixie tubes? As I mentioned earlier in the thread, a coworker has 4 Burroughs B-7971 Alphanumeric Nixie Tubes with sockets that he wants to sell for $100. I'm looking for a new truck right now, so I don't think I'm going to buy them, but they are in great shape and all the segments work. He said he bought them new (this guy is old) and never used them. The sockets were cut out of something, but work fine. If he doesn't sell them in a month or so, I'm interested. I just bought a bunch of IN-18 and IN-14 nixies and as you might imagine, that's pretty much blown through the electronics bugdet for this month and about 5 more months to come.
|
|
#
?
Feb 3, 2010 21:14
|
|
|
hi I'm currently working on a DC-DC converter for a project and currently I have a circuit working to step 48V down to 12V. The problem is the chip I have is only rated for 500mA and I'd like to get around 10A. I was wondering the best way to set up this circuit to drive some sort of FET or something to get that full 10 A out of it. Heres what I have so far:
|
|
#
?
Feb 3, 2010 22:28
|
|
|
dxt posted:hi I'm currently working on a DC-DC converter for a project and currently I have a circuit working to step 48V down to 12V. The problem is the chip I have is only rated for 500mA and I'd like to get around 10A. I was wondering the best way to set up this circuit to drive some sort of FET or something to get that full 10 A out of it. Heres what I have so far: There aren't too many chips out there with integrated FETs that can handle those voltages that high (let alone currents that large! yoikes). Consider using a chip with external FETs, those can withstand much higher currents since the power-dissipating components (FETs and maybe the inductor) are external and can be cooled by convection or heat sinks. Here's one sample chip: http://www.national.com/pf/LM/LM5116.html edit: or the comedy option - build 20 of them in parallel, connect them using ballast resistors (resistors on the order of 100mOhm to prevent directly shorting them together), and tie all the feedback pins to the post-ballast connection point Poopernickel fucked around with this message at 02:09 on Feb 4, 2010 |
|
#
?
Feb 4, 2010 02:06
|
|
|
Poopernickel posted:There aren't too many chips out there with integrated FETs that can handle those voltages that high (let alone currents that large! yoikes). hmm I was hoping to modify this circuit with some external FETs to increase the current somehow. I'll look into other chips though. Thanks.
|
|
#
?
Feb 4, 2010 02:32
|
|
|
My additional temperature sensors, RF receivers and transmitters, photocells, second arduino, RTC breakout board, etc arrived from SFE. I love SFE! (Don't love that $90 worth of stuff could fit inside my apartment's PO box sized mailbox, though.) Programming the time onto it and getting it to run in 12 hour mode was a bitch, but it's done. Sadly I work tomorrow, otherwise I would be getting this RTC sketch meshed with my existing temp to RGB LCD sketch. I also need to figure out how to fade / dim my RGB char LCD with the use of one if not both of my newly gained photo resistors. Progress!
|
|
#
?
Feb 5, 2010 05:51
|
|
|
A while ago I built a temperature sensor based around a PIC18F2553 that connects to a PC through USB as an emulated serial port. I made a board for it but it lacked a debugger connector so I had a breadboard set up that I could quickly pop the PIC into to program with my ICD3. One time I wasn't paying enough attention and I socketed the PIC one set of pins over so that a ground pin was connected to VCC and an I/O pin was connected to ground. I was trying to figure out why the debugger wasn't seeing the PIC when I started to smell something. I touched the PIC and almost burned myself because it was so hot, so I quickly unplugged it and waited for it to cool off. The breadboard was melted where the microcontroller was socketed but it still seemed to work perfectly when I connected it properly on another breadboard. I programmed it, popped it into the project board, and forgot about it. Well today I was working on a new project with that same PIC and wasn't having any luck getting the SPI module to work. After hours of frustration I wrote a quick program to drive all of the SPI pins high so I could check to see if they still worked. The SPI data out pin was only reading 0.15 volts. Then I realized that that was one of the pins incorrectly hooked up during the mishap months ago. Looks like that PIC is going to have to do without hardware SPI 
|
|
#
?
Feb 5, 2010 06:51
|
|
|
BattleMaster posted:A while ago I built a temperature sensor based around a PIC18F2553 that connects to a PC through USB as an emulated serial port. I made a board for it but it lacked a debugger connector so I had a breadboard set up that I could quickly pop the PIC into to program with my ICD3. Heh, I did same sort of thing. Was playing with I2C temp sensor, and I read the datasheet wrong (top veiw!=bottom view) and plugged it in backwards vcc to gnd and +5 to gnd. After getting a bunch of bad CRCs I poked the sensor with my thumb. I got a BB shaped burn instantly. I still have a little divot in my thumb where I touched it. Because of the fact that it was a temp sensor, it was designed for high temps. Tmax was something like 257F, and that was only until the accuracy dropped. The temp sensor was totally unharmed, ans same goes for the pin on the ATMega. I got lucky. I only found out the temp sensor was fine after spending $15 for a new one + shipping though . I was almost sure I hosed it, but decided to try it after I got the new one working. It works perfectly and the accuracy is right on par.
isagoon fucked around with this message at 08:05 on Feb 5, 2010 |
|
#
?
Feb 5, 2010 08:02
|
|
|
The burn test is a valid diagnostic tool. We also have a FLIR camera at work, which makes it instantly obvious if something is shorted or pulling way too much current.
|
|
#
?
Feb 5, 2010 12:54
|
|
|
Do you mind sharing what your job roughly is, Hillridge?
|
|
#
?
Feb 6, 2010 01:31
|
|
|
Update! My second arduino arrived in addition to the additional temperature sensors, protoshields, photocells, RTC, RF receivers and transmitters, etc. I've got internal and external temps working off one arduino currently, I want some 3 and 4 pin spring loaded terminals because I want to make the RTC modular and making the external (off the protoshield) wiring easily removable without being destructive. as you can see i'm using a flat ribbon cable and a 40 conductor IDC connector with 12 pins surgically removed. I'm still looking for a narrower possible option with a greater flexibility on how i can house this thing in an enclosure. Ideas?
|
|
#
?
Feb 6, 2010 02:57
|
|
|
catbread.jpg posted:Do you mind sharing what your job roughly is, Hillridge? I do hardware design, typically embedded systems based around a micro-controller we designed that is cheap and has a ton of features. I get to do a pretty good variety of projects. My group is small and kind of unclassified, but our chip gets used throughout the company so I'm involved in a lot of stuff. I'm primarily working on an LCD touch screen based interface with a few other bells and whistles right now. I do a little design verification related firmware (i.e. testing out the hardware), but I don't usually write anything that goes into released products. The bigger groups have dedicated techs for things like building boards and testing. Since we're small, I get to do all my own rework, which has made me pretty good at soldering little surface mount stuff. My boss used to do pcb assembly, so he is ridiculous at it. I've seen him successfully put on a 256 ball BGA with hot air. My company also has people who are dedicated PCB designers. Is it unique in this aspect? I'll do a schematic, hand it off, and the PCB guy places and routes everything. Then we review and revise as needed. It's a fun job and the people I work with are great. If any of you are in the greater (northern) Boston area, are high school/early college age, and are interested in this stuff, PM me. We MAY have a part time internship position open in a few weeks or so. No promises though  .
|
|
#
?
Feb 6, 2010 04:23
|
|
|
Hillridge posted:The burn test is a valid diagnostic tool. having an infrared camera has been really useful a couple of times. a while back I was testing a large board I had just assembled, and something shorted out on the underside due to a loose piece of wire, and something died and started drawing a lot of current. it could have taken hours to go over each of the 25+ ICs to see which was dead, but we just whipped out the camera and it immediately showed the culprit (inverting charge pump). Another example was last week when we were testing out our prototype solid state tesla coil with a dummy load. We noticed we were pumping in a lot of power but not getting a lot of current in the load, and nothing was really getting hot. We couldn't figure out where the energy was going, so we used the camera and it turned it out we were magnetically coupling a ton of power into the steel desk the coil was lying on. It was a good 5 degrees C above ambient.
|
|
#
?
Feb 6, 2010 04:46
|
|
|
ANIME AKBAR posted:Another example was last week when we were testing out our prototype solid state tesla coil with a dummy load. We noticed we were pumping in a lot of power but not getting a lot of current in the load, and nothing was really getting hot. We couldn't figure out where the energy was going, so we used the camera and it turned it out we were magnetically coupling a ton of power into the steel desk the coil was lying on. It was a good 5 degrees C above ambient. That is loving metal  I need to get out of IT and start geting paid to do this stuff 
|
|
#
?
Feb 6, 2010 07:28
|
|
|
|
| # ? Jun 11, 2024 07:53 |
|
|
I want to make an advanced tens machine. (http://www.amazon.com/LGMedSupply-Platinum-Electronic-Stimulator-Physical/dp/B001PT9QW8). For now, just assume I want to make a basic tens machine. From what I've read would the best option be to get a PIC or AVR kind of microcontroller and program the functionality into that. I've never worked with microcontrollers or any serious electronics before, but the space limits of the microcontrollers apear very small. Basically I will have inputs: 1. Pre-set Programs. Outputs: 1. 4 outputs, with variable intensities and durations. So my thoughts are based on. Program knob setting--->PIC--->electronics to amplify singal to (tens strength). So mainly I want advise on what kind of microcontroller to start learning, also ones where the complexity of what it can handle can be increased. For example if I want 16 output signals, and want 20 complex "settings", will there be a more advanced version which can handle that so any code and work done can easily be transfered. Also considering I'm a newbie something with decent tutorials and tools is kind of required. Most of what I've seen, is people recomending PIC or AVR. I'm quite familiar with c, c++.
|
|
#
?
Feb 6, 2010 16:21
|
|