|
insta posted:Add Q13 and Q8 to that mix. They both went up too. And surely that's not all of the damage either.
|
#
?
Dec 10, 2010 16:23
|
|
|
|
| # ? Jun 10, 2024 23:39 |
|
|
My thought is that maybe its made for American outlets and someone plugged it in to the wrong kind. This AND/OR they tried to play "Working Man" by RUSH (you did say it was a bass amp right?) edit: OR maybe each one of those power MOSFETS has a lead that is tied directly to ground, but someone didn't use a 3 pronged chord to plug it in.
|
|
#
?
Dec 10, 2010 19:56
|
|
|
If it's DC coupled internally even a blown input stage could destroy everything all the way to the outputs. If the transistors are bipolar you can measure if they work pretty easily with a diode tester and the base-emitter junction. Though I wouldn't do this without a good sized pile of spare transistors...
|
|
#
?
Dec 11, 2010 00:08
|
|
|
SnoPuppy posted:If you want the DC operating point, just treat capacitors as opens and inductors as shorts. Wow thank's, using the lapace is kind of like magic. Well anyway I've found out that the circuit I had made up was simply a boost converter. Now I've found out about the boost converter the results I've got so far don't seem to match up with what I've seen on the internet. I think it might have something to do with the the diode but can't really place it. Everything I've seem online analyses it from an energy or steady state kind of way which I'm not really convinced with. Is there an analysis of the boost circuit using standard KVL and the like? What I got was Vo=Vi(1-cos(wD'T)+wDTsin(wD'T))+Vpcos(wD'T).(plus some current term i'll ignore for now) Where DT is the time the switch is closed and D'T is the time the switch is open. Vp is the voltage on the capacitor from the previous cycle. I was simply going to treat the diode as preventing the current from going negative, or stopping Vo from decreasing. Circuit diagram can be found here. http://en.wikipedia.org/wiki/Boost_converter Edit: In the continuous mode the current doesn't drop to zero and is quite important, so ignoring that term like I did is why things were not matching up.. So Vo=Vi(1-cos(wD'T)+wDTsin(wD'T))+Vpcos(wD'T)+IpLwsin(wD'T). Assuming CdVo/dt=Ip, Vo=Vp, I get the same 1/(1-D) relation shown pretty much everywhere. FSMC fucked around with this message at 18:26 on Dec 19, 2010 |
|
#
?
Dec 11, 2010 15:04
|
|
|
This is all off the top of my head, so I might have some stuff a little wrong here. I could dig out my old notes a little later, maybe. KVL won't work because it doesn't take the magnetic field of the inductor into account, I think. The buck and boost converters rely on the fact that V(t) = L (di)/(dt), so you can pretty much figure out what's going on using that. There are three states in each period: Q1 on, D1 off Q1 off, D1 on Q1 off, D1 off (this is very brief) With each state, you can find the voltage drop across the inductor. The input voltage is always the same, and the voltage wrt common at the other side of the inductor goes from Vce to (Vout - Vdiode)
|
|
#
?
Dec 11, 2010 22:49
|
|
|
It's a little more complicated but you can replace the diode with a complementary-switched transistor if the voltage drop bugs you. I'm making a power supply using a PIC with complementary PWM outputs so it's no big deal for me, anyways.
|
|
#
?
Dec 11, 2010 22:55
|
|
|
Unparagoned posted:Wow thank's, using the lapace is kind of like magic. ante posted:KVL won't work because it doesn't take the magnetic field of the inductor into account, I think. ANIME AKBAR fucked around with this message at 06:28 on Dec 12, 2010 |
|
#
?
Dec 12, 2010 06:26
|
|
|
ANIME AKBAR posted:Boost converters don't really operate based on resonance, don't let the fact that there's an LC in there fool you. The inductor is purely for energy storage. You generally assume that the switching period is much shorter than the resonance period. However, for the purpose of controlling it, it can be modeled as an LCR circuit (though the model is only valid at frequencies lower than the switching frequencies). In that regime, it will have resonant behavior. I split it into two main circuits the L and LC circuits. The way I considered it was to model the LC circuit and then find the conditions such that current doesn't become negative, for now I'm ignoring the diode voltage drop. So I'm finding the point such that the current becomes negative and making sure the time constants are chosen such that current doesn't become negative. So I have Vo=Vi(1-cos(wD'T)+wDTsin(wD'T))+Vpcos(wD'T), which has resonant behaviour but I'm finding out what the conditions are and the limits so there is no resonant behavior. I would have thought that would work? I was also looking at was what's the optimum switching time constants, and what's the long term behavior. When t goes to infinity I got the voltage going to V=Vi(1+DT/(tan(wD'T/2)). Surely there must be a decent analysis of the boost converter online somewhere? FSMC fucked around with this message at 13:21 on Dec 12, 2010 |
|
#
?
Dec 12, 2010 13:18
|
|
|
Unparagoned posted:Surely there must be a decent analysis of the boost converter online somewhere? Using a Spice tool (Like LtSpice I posted earlier) will be much, much easier. The problem is that analytically solving a buck or boost converter would require iterating over several differential equations with discontinuous input functions. And that's hard. So that means you have to analytically solve the system when the switch is closed and then solve the system when the switch is open. If we assume that your switch follows a square wave, next you have to compute the each part of the square wave (i.e. the "on" or "off") using the results of the previous part as your initial values. And you would have to do this for each cycle until you get bored and decide you're at steady state. Now, if all you want are efficiency numbers you can make some decent approximations by using the Rds-on of your FETs, the Miller capacitance (input capacitance of the FET), and the Rseries of the inductor. Just sum your losses and you're done. But like I said, using a spice tool will be way less headache and probably way more accurate.
|
|
#
?
Dec 12, 2010 18:19
|
|
|
This document from Microchip goes into various switched-mode power supply topologies (including boost and buck) with large numbers of equations, graphs, schematics, and current flow diagrams. It isn't PIC-centric until a bit at the end where they talk about using dsPICs as power supply controllers.
|
|
#
?
Dec 12, 2010 18:41
|
|
|
Unparagoned posted:I split it into two main circuits the L and LC circuits. The way I considered it was to model the LC circuit and then find the conditions such that current doesn't become negative, for now I'm ignoring the diode voltage drop. So I'm finding the point such that the current becomes negative and making sure the time constants are chosen such that current doesn't become negative. So I have Vo=Vi(1-cos(wD'T)+wDTsin(wD'T))+Vpcos(wD'T), which has resonant behaviour but I'm finding out what the conditions are and the limits so there is no resonant behavior. I would have thought that would work? I was also looking at was what's the optimum switching time constants, and what's the long term behavior. When t goes to infinity I got the voltage going to V=Vi(1+DT/(tan(wD'T/2)). Here, I made a series of posts on how boost converters operate way back in this thread. Starts around here: http://forums.somethingawful.com/showthread.php?threadid=2734977&userid=109669&perpage=40&pagenumber=2#post342675135
|
|
#
?
Dec 12, 2010 19:27
|
|
|
SnoPuppy posted:The problem is that analytically solving a buck or boost converter would require iterating over several differential equations with discontinuous input functions. And that's hard. That is exactly what I did and was able to generalise for pretty much any n(Ignoring 1 term). I found that the system tends towards a steady state, is this correct? I'm going to read up on LTspice I couldn't get it do anything when I last tried it. ANIME AKBAR posted:Here, I made a series of posts on how boost converters operate way back in this thread. Starts around here: That's pretty interesting. Seems like I would need to have a feedback system. BattleMaster posted:This document from Microchip goes into various switched-mode power supply topologies (including boost and buck) with large numbers of equations, graphs, schematics, and current flow diagrams. It isn't PIC-centric until a bit at the end where they talk about using dsPICs as power supply controllers. edit: It seems to do stuff I just don't get. For example on page 14, equation 47 is the simple L equation for an inductor charging up. (I=Vt/L). But when the capacitor is then connected in equation 49 they use pretty much the same kind of equation. Is that correct or are there some unsaid assumptions such that the capacitors voltage doesn't change after being charged, etc.... The current I derived does simplify to that equation if I assume wt is small and use first Taylor expansion, and assume T^2 is small enough to ignore. Thanks everyone for the help. Next week I'm going to take a couple of weeks off work to spend more time on this. One of the reasons I'm kind of laboring on about this is that it's going to be one of the more simpler circuits I can think of and wanted to see if I can analyse and work out what it does using usual methods. (I didn't even know it was a common boost circuit until recently). My reasoning is that if I can't work out what this does then I don't have any chance figuring out what more complicated circuits I design would do. FSMC fucked around with this message at 03:06 on Dec 13, 2010 |
|
#
?
Dec 13, 2010 02:15
|
|
|
Got fed up with an LCD monitor getting progressively weirder acting. Opened it up and the power circuit board looks fine but I noticed two 470mF capacitors that had just slight bulges compared to the rest which were perfectly flat. Is this likely to be my issue? I ordered from digikey tonight hopefully Ill have a few new caps here in a few days. The caps don't appear to have split or leaked or anything, they're just a little bulgy.
|
|
#
?
Dec 13, 2010 08:38
|
|
|
AnomalousBoners posted:Got fed up with an LCD monitor getting progressively weirder acting. Opened it up and the power circuit board looks fine but I noticed two 470mF capacitors that had just slight bulges compared to the rest which were perfectly flat. Is this likely to be my issue? I ordered from digikey tonight hopefully Ill have a few new caps here in a few days. Whenever I've had caps fail and ruin electronics, they've been power supply filter caps that failed to a short; measuring the resistance across the caps (which sat between a +V rail and ground) would give 0.0 ohms, or a dead short on the supply rail. I'm not sure if that's universally true, but it's easy to measure real quick and see.
|
|
#
?
Dec 13, 2010 09:08
|
|
|
For some reason I thought you couldn't check caps for continuity or something. the two caps I thought were bad read as infinite ohms, the cap on a motherboard that's slightly bulging (about the same as the LCD caps) reads infinite, those exact same caps on the mobo that are flat give a value. Thanks for the tip they're almost certainly bad. Lets hope the parts coming from digikey is all that it needs. EDIT: All the caps I ordered are panasonic, is this a decent brand? EDIT 2: Located the fuse on the LCD's PS board and it tested good. I didn't think it wouldn't but its best to check. AnomalousBoners fucked around with this message at 16:34 on Dec 13, 2010 |
|
#
?
Dec 13, 2010 16:12
|
|
|
AnomalousBoners posted:For some reason I thought you couldn't check caps for continuity or something. the two caps I thought were bad read as infinite ohms, the cap on a motherboard that's slightly bulging (about the same as the LCD caps) reads infinite, those exact same caps on the mobo that are flat give a value. Thanks for the tip they're almost certainly bad. Lets hope the parts coming from digikey is all that it needs. You know you could've just gone down to radio shack and picked a couple of these up in person, right? Don't forget that a lot of capacitors, I think its electrolytic (I just scoured the wiki on it so I could give you something you could quote me on but if anybody knows better please correct me) are polarized; the have a positive-only and a negative-only electrode. Theirs usually a white line running from the top of the cylinder shape to the bottom of it, and is adjacent to a particular lead; this is usually the negative* electrode (that's correct, isn't it?) So for one make sure you are measuring the resistance with the test leads of the multimeter touching the right leads, and number two, if the caps were bad its more characteristic for them to show 0 ohms resistance, as opposed to the infinity you were getting, which is more indicative of an open circuit.
|
|
#
?
Dec 13, 2010 18:16
|
|
|
Not really sure where to ask this, so this seems as good as any... I'm looking for a bunch of that chinese-finger-trap style wire sheathing, for keeping a bundle of wires together, like they often put on the wires coming out of PSUs and such. I need to do three 10-foot lengths. Does anyone know where I might find such a thing? Or a suitable alternative; zip-ties do not interest me because they tend to catch once installed and have sharp edges where you clip off the excess.
|
|
#
?
Dec 13, 2010 18:31
|
|
|
I talked to a well learned man about this today and he suggested getting an analog multimeter to test the caps. The reason being it allows you to see a quick spike and then the discharge... Though apparently we have cap testers here at school....so yea I will probably be bringing them in here. Also yes these are polarized "can type" caps.
|
|
#
?
Dec 13, 2010 18:39
|
|
|
Bad Munki posted:Not really sure where to ask this, so this seems as good as any... This is not EXACTLY what you asked for, but have you ever heard of doing this? http://www.matronics.com/aeroelectric/articles/cable_lace/cable_lace.html Specifically, check out the last imagine on that page.
|
|
#
?
Dec 13, 2010 18:50
|
|
|
Hmm. That's not a bad option, but I'll probably only do that as a backup. Definitely something to keep in mind, though.  e: Oh hey, found some! Now to just find it locally because I'm an impatient rear end... Bad Munki fucked around with this message at 20:10 on Dec 13, 2010 |
|
#
?
Dec 13, 2010 18:56
|
|
|
Bad Munki posted:Hmm. That's not a bad option, but I'll probably only do that as a backup. Definitely something to keep in mind, though. Mouser sells stuff like that too: http://www.mouser.com/Wire-Cable/Tubing-Sleeving/_/N-5ggy?Keyword=braided Depending on where you live, it might get there the next day with the cheapest shipping.
|
|
#
?
Dec 13, 2010 20:18
|
|
|
Sweet, thanks. Im in Iowa, so that's about two days by ground most likely. And even as impatient as I am, I can probably make myself wait a day or two. Now I just need to figure out what size I should be getting. 
|
|
#
?
Dec 13, 2010 20:22
|
|
|
AnomalousBoners posted:I talked to a well learned man about this today and he suggested getting an analog multimeter to test the caps. The reason being it allows you to see a quick spike and then the discharge... Though apparently we have cap testers here at school....so yea I will probably be bringing them in here. Also yes these are polarized "can type" caps. In general when measuring large caps with a diode tester you'll see the voltage start at 0 and then increase until the cap is charged to whatever voltage the tester uses, often 1,6-2V. If you then reverse the polarity it will start at -something, cross zero and move up. I think most capacitor testers work the same way, charging it up and measuring the time it took. I saw a capacitor in a TV PSU once where the capacitance was way down but it worked otherwise, in that case it was a filter cap for the secondary of the stand-by supply. Since it wasn't filtering properly the power was very unstable and noisy, which made the feedback system shut it down after a few seconds.
|
|
#
?
Dec 13, 2010 22:17
|
|
|
Bad Munki posted:Hmm. That's not a bad option, but I'll probably only do that as a backup. Definitely something to keep in mind, though. Go to your local advance auto or autozone. It's sold in the ricer isle. Ask specifically where the "ricer isle" is.
|
|
#
?
Dec 14, 2010 00:41
|
|
|
insta posted:Go to your local advance auto or autozone. It's sold in the ricer isle. Ask specifically where the "ricer isle" is. Damnit, I'd so successfully avoided such places until now.
|
|
#
?
Dec 14, 2010 01:39
|
|
|
Ok, so I've decided to forget about repairing the amplifier, and instead rewire the speaker itself so I can use it with my other bass amp. The amp in question is a 400W Ashton BH400. It has two 4ohm min outputs and an xlr for DI. Here are some pics of the speaker, and the inputs I purchased (yes I will only use one, I just bought 2 in-case I stuffed up).  [img]http://img.waffleimages.com/2b3e21b7f7f3d4ec1508b5ca211532f5671bc01b/image[1].jpeg[/img] [img]http://img.waffleimages.com/27ddb3a408752600bac1a36f49a17d68cc544821/image[2].jpeg[/img] [img]http://img.waffleimages.com/174acf1329266bbdb100235a2aad0a78715af757/image[3].jpeg[/img] Do I have the correct input jacks? I've been trying to search for the speaker model to ensure it is the correct impedance, but to no avail.
|
|
#
?
Dec 14, 2010 09:33
|
|
|
AnomalousBoners posted:I talked to a well learned man about this today and he suggested getting an analog multimeter to test the caps. The reason being it allows you to see a quick spike and then the discharge... Though apparently we have cap testers here at school....so yea I will probably be bringing them in here. Also yes these are polarized "can type" caps. It's a good idea to have both an analogue and a digital multimeter as there are situations for both types where one gives nonsensical measurements where the other doesn't.
|
|
#
?
Dec 14, 2010 10:32
|
|
|
I have an idea for something, but I figured I'd ask here because a good chunk of it is going to need a microcontroller, and whether it would be too much or not. I did a little bit when I was still in school, but otherwise haven't done much electronics stuff since. I'm going to make a Master Chief costume for Halloween, and I think a "fully functioning" assault rifle would round it out nicely. I want to: * Pull the trigger, and a firing sound plays while a counter ticks down ammo. * When the counter hits zero, a clicking "empty chamber" sound plays. * To reset the counter, work the slide (bolt? whatever) Besides the mechanics and fitting it into the drat thing, it doesn't sound too complicated. I think an ArudinoUno with the audio player shield will get it. Besides time to learn it all, does this seem like too much, or is there an easier way? EDIT: And before I do all this, is there a program I can sketch this out on? Fooley fucked around with this message at 00:48 on Dec 15, 2010 |
|
#
?
Dec 15, 2010 00:44
|
|
|
I have a piece of equipment (a water cooler for a TIG torch) that I want to turn on automatically when another piece of equipment (a TIG welding machine) is turned on manually. I'm envisioning wrapping a wire around one of the welding machine's power supply wires forming an inductive pickup, and using that signal to trigger some electronics which drive a relay. The welding machine draws 2 amps when turned on, and up to 60 amps at full load @ 220V AC. The water cooler draws 6 amps @ 110V AC. I know there is equipment out there to operate dust collection for woodworkers when they turn on a saw, but the last time I checked (a long time ago) it was way more then I am willing to spend. Does anybody have any easy to build circuits that can do this job?
|
|
#
?
Dec 15, 2010 06:02
|
|
|
Fooley posted:I have an idea for something, but I figured I'd ask here because a good chunk of it is going to need a microcontroller, and whether it would be too much or not. I did a little bit when I was still in school, but otherwise haven't done much electronics stuff since. Should be pretty simple. Two switches, one global variable, some functions to check for button presses and output the display, if you're doing that. Some LCD modules have pretty terrible datasheets that are a pain to figure out, though. dv6speed posted:I have a piece of equipment (a water cooler for a TIG torch) that I want to turn on automatically when another piece of equipment (a TIG welding machine) is turned on manually. A step-down current transformer driving a relay through some power diodes?
|
|
#
?
Dec 15, 2010 06:37
|
|
|
ante posted:A step-down current transformer driving a relay through some power diodes?
|
|
#
?
Dec 15, 2010 06:46
|
|
|
Fooley posted:I think an ArudinoUno with the audio player shield will get it. Besides time to learn it all, does this seem like too much, or is there an easier way? I think you could fairly easily do this with Arduino boards. You can sketch it out on the Arduino software available from their site. There is also a free program called Virtual Bread Board (VBB) that you can download and prototype with. It has some Arduino shields and gives you a visual of what you are programming, its decent for learning.
|
|
#
?
Dec 15, 2010 15:57
|
|
|
Fooley posted:...HALO AR... Check out this blaster thread specifically for a pretty incredible build.
|
|
#
?
Dec 15, 2010 23:30
|
|
|
I'm going to need to start making my own inductors for experiments. Are there any good resources on designing, building, and testing inductors?
|
|
#
?
Dec 15, 2010 23:30
|
|
|
BattleMaster posted:I'm going to need to start making my own inductors for experiments. Are there any good resources on designing, building, and testing inductors?
|
|
#
?
Dec 15, 2010 23:46
|
|
|
Does anyone have any experience with writing to SD cards? I've got my PIC talk to it, and I can send it CMD0 and get the proper response, but trying to send it CMD1 gives me a 0x05 in response, illegal command and idle. Pretty sure I'm just supposed to be getting a 0x01
|
|
#
?
Dec 16, 2010 00:02
|
|
|
BattleMaster posted:I'm going to need to start making my own inductors for experiments. Are there any good resources on designing, building, and testing inductors? http://info.ee.surrey.ac.uk/Workshop/advice/coils/ This is some pretty good advice. Making coils without an RCL meter is insanity, though. You could probably get by with a scope and signal generator if you were very patient.
|
|
#
?
Dec 16, 2010 01:36
|
|
|
sixide posted:http://info.ee.surrey.ac.uk/Workshop/advice/coils/ Nah, a decent scope and a power supply is all you need. 0.1ohm current sampling resistor in series with the inductor, then connect the power leads with the scope on single acquisition mode. V=Ldi/dt on the bit before the inductor hits saturation. I can get within less than 1% of a (really good) RCL bridge's measurements with my self-wound inductors.
|
|
#
?
Dec 16, 2010 01:45
|
|
|
Speaking of oscilloscopes, what can I get one for these days?
|
|
#
?
Dec 16, 2010 03:37
|
|
|
|
| # ? Jun 10, 2024 23:39 |
|
|
BattleMaster posted:I'm going to need to start making my own inductors for experiments. Are there any good resources on designing, building, and testing inductors? I've done a fair amount of custom magnetics design, mainly for switching power supplies and current transformers. Anything in particular you want to know? What's your application? ANIME AKBAR fucked around with this message at 04:54 on Dec 16, 2010 |
|
#
?
Dec 16, 2010 04:51
|
|