Boy With Stick posted:With your circuit arrangement right now you are outputting from your amplifier a sine wave on top of a DC bias - the average value of the waveform, in this case 1/2 the microcontroller's Vcc. In addition to not conveying useful audio information, this DC voltage can also adversely affect your amplifier and speaker performance. In most typical cases a speaker with DC voltage applied to its terminals will more easily distort (due to the driver statically deflecting from its center) and worst-case self-destruct due to the coil overheating. I was sort of starting down this path, but I couldn't figure out what the difference is between this and a class A amp. If the signal was biased on either end of the operating spectrum and clipping, I could see a problem (and maybe he is doing this?) but if he stays below that level and doesn't draw too much power from the amp (and this might be happening too) but otherwise it would work. Seems like it would be easier to fix these fundamental problems (gain and speaker load impedance) rather than just rebiasing the signal. Or, I don't understand what you are saying
|
|
# ? Dec 11, 2012 10:50 |
|
|
# ? May 9, 2024 09:48 |
|
I got it to work, removed the opamp stuff and just went with this circuit that some other MSP430 guys in irc showed me. Which isn't far off from the LM386 setup I already had in place. I set the cutoff on that to 4Khz. I had a few other issues that were contributing to all the crazy distortion i was getting, including a software error and apparently poor SPI connection to my sd card. Now that that's sorted out, the sound it really very nice. Works amazingly well considering it's only single stage filter now. I might have been able to get that 2 stage opamp filter to work with it, but this is working well enough that I'm not gonna mess with it. I swapped out the speaker and LM386 is actually driving a 6 Ohm speaker fine right now.
|
# ? Dec 11, 2012 11:13 |
|
Got my TekPower TP4000ZC meter in from Amazon. This was the $30 deal. Very nice! Seems pretty accurate and I love the RS232 function on it. Hooked it to a USB-Serial adapter, download the VC 840 and QtDMM software and it shows up and works like a champ. Very pleased with this purchase. Edit: HATE TROLL TIM fucked around with this message at 21:48 on Dec 11, 2012 |
# ? Dec 11, 2012 21:34 |
|
Delta-Wye posted:I was sort of starting down this path, but I couldn't figure out what the difference is between this and a class A amp. If the signal was biased on either end of the operating spectrum and clipping, I could see a problem (and maybe he is doing this?) but if he stays below that level and doesn't draw too much power from the amp (and this might be happening too) but otherwise it would work. Seems like it would be easier to fix these fundamental problems (gain and speaker load impedance) rather than just rebiasing the signal. Anytime you have a amplifier with a single supply voltage you're going to have this issue. In order to generate both positive and negative AC voltages from such a circuit, the final output stage will need either a DC-blocking cap or transformer-coupling. Within the amplifier stages the signal should be biased to 1/2 the rail in order to achieve maximum dynamic range for the AC signal. Peepsalot posted:I got it to work, removed the opamp stuff and just went with this circuit that some other MSP430 guys in irc showed me. You didn't remove the op-amp stuff, you just opted for smarter op-amp stuff. The LM386 is an interesting case as it automatically biases the output voltage to 1/2 the rail. This makes it so you don't have to worry as much about the bias present in the input signal. (Still have to worry about it the output of course, hence the DC-blocking cap in yours as well as all the example circuits in the datasheet.) To make this circuit work with a conventional op-amp you'd have to bias the inverting input to 1/2 the rail, usually done with a simple resistor divider from Vcc to ground. Glad to hear that the circuit ended up working okay even with the low impedance speaker. The sensitivity must be high enough that you can drive it with low voltages and still produce acceptable loudness.
|
# ? Dec 12, 2012 02:36 |
|
LM386's are great little cheater chips. I ended up throwing one in a bass preamp and it sounded really good.
|
# ? Dec 12, 2012 05:59 |
|
Is a "decoupling" cap just used to filter noise and ripple like a filter, or is there some other meaning/use for the term?
|
# ? Dec 12, 2012 06:31 |
|
Martytoof posted:Is a "decoupling" cap just used to filter noise and ripple like a filter, or is there some other meaning/use for the term? Pretty much. It means a capacitor that is used on a power supply/voltage reference for the purpose of providing a low impedance power distribution network (i.e. noise filtering). If it's part of a filter, I'd just call it a filter capacitor. Possibly shunt or series if I wanted to get more specific.
|
# ? Dec 12, 2012 06:43 |
|
Martytoof posted:Is a "decoupling" cap just used to filter noise and ripple like a filter, or is there some other meaning/use for the term? Decoupling caps filter power rails by forming an RC filter along with the resistance of the copper trace, placed as near as possible to the IC (or whatever you are trying to supply clean power to) They are different than bulk filtering caps--those are generally large electrolytic caps used to smooth out power rail voltage ripple. They wouldn't work well for filtering high freq noise.
|
# ? Dec 12, 2012 07:42 |
|
Martytoof posted:Is a "decoupling" cap just used to filter noise and ripple like a filter, or is there some other meaning/use for the term?
|
# ? Dec 12, 2012 10:20 |
|
tonberrytoby posted:I would first think of a DC decoupling cap, like C1 and C3 a few posts back. But when I checked the dictionary it seems like this is a translation error on my part. That would be a dc blocking cap.
|
# ? Dec 12, 2012 13:53 |
|
Boy With Stick posted:Anytime you have a amplifier with a single supply voltage you're going to have this issue. In order to generate both positive and negative AC voltages from such a circuit, the final output stage will need either a DC-blocking cap or transformer-coupling. Within the amplifier stages the signal should be biased to 1/2 the rail in order to achieve maximum dynamic range for the AC signal. I am still a noob in electronics related endeavors, but building a CMoy amp with a "virtual ground" was a useful little project to teach myself about this. It is cheap, easy, and you end up with something you can use in later projects.
|
# ? Dec 12, 2012 14:00 |
|
Slanderer posted:That would be a dc blocking cap. Depending on local jargon, it can be a coupling capacitor, decoupling capacitor, or occasionally a dc blocking capacitor.
|
# ? Dec 12, 2012 14:09 |
|
A decoupling cap prevents noise from coupling to the power rails, potentially causing erroneous operation, a coupling cap couples an AC signal without coupling the DC component.
|
# ? Dec 12, 2012 14:50 |
|
Older folks seem to use the term 'bypass cap' almost to the exclusion of 'decoupling cap', incidentally
|
# ? Dec 12, 2012 15:06 |
|
Martytoof posted:Is a "decoupling" cap just used to filter noise and ripple like a filter, or is there some other meaning/use for the term? Even the most carefully designed power distribution (the traces, wires, planes etc that make up power and ground) will have some non-trivial amount of impedance. This is comprised of two components: 1. resistive due to copper loss, contacts and the like and 2. the more problematic inductive component due the loop area of the current path. Many devices such as op-amps, motor drivers, voltage regulators and logic are often operated in a fashion wherein they can demand significant peak or high-frequency currents from their power rails. For example when responding to a load transient, switching many gates simultaneously, starting a motor, etc. Peak current/high-frequency -> large dI/dt. This dI/dt will tend create a transient voltage drop across the impedance of the distribution path which may cause problems for devices on the rail. A big peak for a motor driver might for example cause a large voltage drop which causes a microcontroller to reset. A decoupling cap placed from power to ground at each device provides a localized source of energy for such transients. Instead of flowing through the entire distribution, the dI/dt now flows out of the nearby capacitor, preventing the large voltage drop that would have occurred if no cap was present. More generally if a device's current draw contains high frequency content it will tend to come out the cap instead of from the source. Essentially these capacitors "decouple" individual devices from disturbing one another due to transient-current voltage drop or noise that a device might inject onto the rail. Boy With Stick fucked around with this message at 19:17 on Dec 12, 2012 |
# ? Dec 12, 2012 15:07 |
|
So I checked around a bit. The words used here in Germany literally translate to: Decoupling cap or separation cap or blocking cap for the serial DC blocker. Throttling cap or dejamming cap or bias cap for the parallel HF blocker. And a lot of people start using those literal translation once we move out of our core areas of activity.
|
# ? Dec 12, 2012 15:31 |
|
Looking into getting a copy of Stroud's Engineering Mathematics for home and work use; is there an analogous text for analog EE?
|
# ? Dec 12, 2012 19:21 |
|
Otto Skorzeny posted:Looking into getting a copy of Stroud's Engineering Mathematics for home and work use; is there an analogous text for analog EE? But I think the best general practical introduction to electronics is Horowitz and Hilt's Art of Electronics. It is a bit outdated but still very interesting.
|
# ? Dec 12, 2012 19:28 |
|
It looks like there's a third edition in the works; I think I'll wait for that, even though it may be a year and a half off. My purpose here is thus: the CSE program at my school was nominally about an even mix of CS and EE, but the choice of electives that I took tilted toward CS type stuff and the digital side of EE, and now I'd like to bring myself to the point where I'm able to handle both sides of a design competently. I'm not in a crazy hurry though, and I learn a good bit through osmosis from coworkers.
|
# ? Dec 12, 2012 19:46 |
|
This has probably been mentioned before, but the icircuit app is pretty nifty for just seeing how electronic stuff behaves. http://icircuitapp.com/ but nicely done and can be helpful for visualizing what is going on. Shows you where the current is flowing, etc.
|
# ? Dec 12, 2012 19:50 |
|
Otto Skorzeny posted:It looks like there's a third edition in the works; I think I'll wait for that, even though it may be a year and a half off. Keep in mind that the third edition has been a year and a half away for about five years now. Writing a book of that scope is no easy task, but they've slipped multiple promised release dates and I'm not holding my breath for 3rd edition coming out any time soon. That said, it's a really great practical design guide that explains a lot of subjects in a low-math, nearly conversational style. My 2nd edition copy is an early go-to source for whenever I need to read up on something that I either haven't done before or haven't done for a long time. The 2nd ed. has held up well for the most part - only the sections on digital electronics are really getting out of date. Basically a cheap used or eReader copy might be worth picking up for the time being.
|
# ? Dec 13, 2012 00:23 |
|
taqueso posted:I've been putting off buying AoE because the new edition is supposed to come out this year. That is what I posted more than a year and a half ago. As PDP-1 says, don't hold your breath.
|
# ? Dec 13, 2012 00:52 |
|
priznat posted:This has probably been mentioned before, but the icircuit app is pretty nifty for just seeing how electronic stuff behaves. Looks great, but can I get it on something that doesn't have a tiny screen and that I don't have to poke at with my big greasy fingers? Actually, I've been looking for a good, cheap circuit simulator for the computer for a while too. I tried some free ones out but I didn't really like any of them. I used to have this one back in the 90's that I learned a ton from, but I can't remember the name of and I'm pretty sure the company went out of business. It had a "rocket" component and a "car" component for some reason, which was enough to keep my 8 year old self interested in learning about poo poo. It was a full-fledged circuit simulator though, not just a kid's program, and I remember it had a pretty straightforward UI and I liked it.
|
# ? Dec 13, 2012 01:27 |
|
Parallel Paraplegic posted:Looks great, but can I get it on something that doesn't have a tiny screen and that I don't have to poke at with my big greasy fingers? If you qualify you could also look for an educational license for whatever the cool LF kids are using these days. Those can be surprisingly cheap or even free.
|
# ? Dec 13, 2012 01:39 |
|
Parallel Paraplegic posted:Looks great, but can I get it on something that doesn't have a tiny screen and that I don't have to poke at with my big greasy fingers? It's available on the Mac App Store so that certainly fits the bill. I've got it on both my iPhone and Mac; it's pretty great for honestly. It's nice to be able to sit down and fiddle with something on my laptop at night then transfer it to my iPad to work with in the garage.
|
# ? Dec 13, 2012 03:17 |
|
timb posted:It's available on the Mac App Store so that certainly fits the bill. I've got it on both my iPhone and Mac; it's pretty great for honestly. It's nice to be able to sit down and fiddle with something on my laptop at night then transfer it to my iPad to work with in the garage. I didn't see that, but I don't have a mac. Thanks anyway EDIT: I'm trying out LTSpice (yeah it's still free) and that seems to be fine, thanks tonberrytoby. Shame Boy fucked around with this message at 03:44 on Dec 13, 2012 |
# ? Dec 13, 2012 03:39 |
|
Parallel Paraplegic posted:I didn't see that, but I don't have a mac. Thanks anyway I'll put in another vote for LTSpice - don't think that just because it's free, it's a lesser tool. LTSpice is easily the best spice simulator I've used, both in terms of performance and capability. And, while I know you said you don't have a mac, I can vouch that it seems to work perfectly under wine if there are any other mac/linux users who need a good simulator.
|
# ? Dec 13, 2012 04:57 |
|
SnoPuppy posted:I'll put in another vote for LTSpice - don't think that just because it's free, it's a lesser tool. I went ahead and bought that app for my android phone to play with when I'm bored at work, it's designed well enough to use on my phone comfortably so I can get some fun out of it. But yeah, I'll be using LTSpice on my PC (and I'm actually running it under WINE and it works fine).
|
# ? Dec 13, 2012 05:01 |
|
So I'm looking into getting a CPLD or FPGA in the near future and Lattice Semiconductor is looking pretty appealing with their inexpensive (~$30-40) development kits. Has anyone else tried them out? I'd like to know how they compare to Xilinx or Altera. I'm not planning to implement soft CPUs or anything, I just want to be able to do stuff like drive displays, so a huge FPGA is probably overkill.
|
# ? Dec 13, 2012 10:37 |
|
Silver Alicorn posted:So I'm looking into getting a CPLD or FPGA in the near future and Lattice Semiconductor is looking pretty appealing with their inexpensive (~$30-40) development kits. Has anyone else tried them out? I'd like to know how they compare to Xilinx or Altera. Lattice should be fine for what you're doing, they sometimes also have steep discounts on their IP cores / hand them out for free.
|
# ? Dec 13, 2012 16:17 |
|
I use MachXO2s at work. The Lattice tools are fine, but not quite as polished as the Xilinx/Altera tools. Your experience with one manufacturer will transfer over to the others, so I wouldn't worry about "picking sides" too much. After all, it is all RTL design. Buying programming dongles can get expensive if you switch around, but I think the dev kits have a built-in programmer. Lattice support has been good the couple times I have contacted them. Response times around 1 day.
|
# ? Dec 13, 2012 18:05 |
|
The papilio board looks pretty interesting. It has a Spartan 3E and is sort of presented as an arduino for the FPGA world. I don't own one, but I was considering buying one. They give some examples of using an AVR core to run arduino programs (not sure why you'd want to though..), but I think it can be used with the Xilinx ISE and VHDL/Verilog.
|
# ? Dec 13, 2012 22:17 |
|
It's quite a bit more than the Lattice board but the Zedboard from Digilent is pretty cool. Has a Xilinx Zynq device (ARM A9 + fpga fabric on one device) along with RAM, SD Card, HDMI/VGA out, ethernet, audio etc. http://www.zedboard.org Has a FMC connector for further expansion too. Price is $320 for academics and $400 for everyone else, so not cheap.
|
# ? Dec 13, 2012 22:21 |
|
Ordered an (ostensibly new) export copy of The Art of Electronics 2nd Edition for $18. Maybe by the time I wear out the lovely binding, the 3rd Edition will have materialized from the luminiferous æther and be available in hardcover
|
# ? Dec 13, 2012 22:27 |
|
Otto Skorzeny posted:Ordered an (ostensibly new) export copy of The Art of Electronics 2nd Edition for $18. Maybe by the time I wear out the lovely binding, the 3rd Edition will have materialized from the luminiferous æther and be available in hardcover Ahahahahahah your avatar. My custom title for you is no more. I spent more than that on a used copy of the book, but the analog bits are definitely worth it. Good book to have on your shelf. priznat posted:It's quite a bit more than the Lattice board but the Zedboard from Digilent is pretty cool. Has a Xilinx Zynq device (ARM A9 + fpga fabric on one device) along with RAM, SD Card, HDMI/VGA out, ethernet, audio etc. Zynq is pretty cool, though we decided to go with the Cyclone SoCs for a project here at work. The only Zynqs that have transceivers run Kintex fabric vs the Artix of the two lower end Zynqs, so the price goes through the rough. The Cyclone and Zynq share essentially the same ARM + peripheral architecture though, straight down the number and configuration of AXI ports and xbar switches.
|
# ? Dec 13, 2012 22:32 |
|
movax posted:Zynq is pretty cool, though we decided to go with the Cyclone SoCs for a project here at work. The only Zynqs that have transceivers run Kintex fabric vs the Artix of the two lower end Zynqs, so the price goes through the rough. The Cyclone and Zynq share essentially the same ARM + peripheral architecture though, straight down the number and configuration of AXI ports and xbar switches. Yeah some guys here have a theory the two just share the identical ARM die sandwiched on top of the switch fabric. We're planning on slapping a Zynq down to act as a board controller/web server for our emulation platform (Virtex 7 2000T ) so we don't need the transceivers, thankfully. Doing a PCIe card right now with a Kintex-7 on it and it is a really sweet price/size/transceivers balance. It's funny because every couple projects we get Altera in to have a look at their stuff and then end up going back with Xilinx anyway. Change is hard! With the Vivado tools at least Xilinx is making an effort to make their tools somewhat less lovely (with limited results so far, however).
|
# ? Dec 13, 2012 22:40 |
|
movax posted:Ahahahahahah your avatar. My custom title for you is no more. The song was stuck in my head from a hundred hours playing that blasted game On a different note, anyone tried to drive nixies from rectified mains? I'm unsure whether I need to have my isolation transformer boost it up a bit as well, as 120v * 1.41 = 169v and that would be the top end of my ripple and the sparking voltage is ~165v (although nominally I only need ~150v or so to sustain).
|
# ? Dec 13, 2012 22:54 |
Otto Skorzeny posted:The song was stuck in my head from a hundred hours playing that blasted game Otto Skorzeny posted:On a different note, anyone tried to drive nixies from rectified mains? I'm unsure whether I need to have my isolation transformer boost it up a bit as well, as 120v * 1.41 = 169v and that would be the top end of my ripple and the sparking voltage is ~165v (although nominally I only need ~150v or so to sustain). It should work fine but your tubes may be different than mine. I had mine wired up to mains ran through an isolation transformer and they lit up fine.
|
|
# ? Dec 13, 2012 23:21 |
|
Delta-Wye posted:Curious, what game? Fallout New Vegas https://www.youtube.com/watch?v=pPArO-OI_3U
|
# ? Dec 14, 2012 00:18 |
|
|
# ? May 9, 2024 09:48 |
|
https://www.youtube.com/watch?v=b7KG48-RUeA
|
# ? Dec 14, 2012 03:58 |