I have a pair of F5T V2 monoblocks that I'm building, they are all assembled, but I am having issues with both amps in the amplifier section, power supply is working fine. Everything is wired up, and I've triple checked all component values and locations, polarities, all wiring arrangements.
Problem is that I'm getting .4V across the source resistors with only 20VAC going into the amp... P1,P2, and P3 do not change anything, DC on the output is about 1V with the same voltage in, power supply is about half the voltage it should be at 20VAC, but is fine when unloaded.
I don't think I have the right kind of brain for amplifier work and trouble shooting, so I want this to be the last amp I do, but that means it has to be finished first...
Anyway, I know some of you are more sadistic than I and would probably enjoy trouble shooting these, since it should be a pretty obvious solution that I'm just overlooking because I'm too dense.
Here's what we are dealing with.
Try not to give me too much **** about my wiring, it's only my second solid state amp
Don't have any money, but would be willing to help with a speaker project if there is anyone who can conquer this beast.
Problem is that I'm getting .4V across the source resistors with only 20VAC going into the amp... P1,P2, and P3 do not change anything, DC on the output is about 1V with the same voltage in, power supply is about half the voltage it should be at 20VAC, but is fine when unloaded.
I don't think I have the right kind of brain for amplifier work and trouble shooting, so I want this to be the last amp I do, but that means it has to be finished first...
Anyway, I know some of you are more sadistic than I and would probably enjoy trouble shooting these, since it should be a pretty obvious solution that I'm just overlooking because I'm too dense.
Here's what we are dealing with.
Try not to give me too much **** about my wiring, it's only my second solid state amp
Don't have any money, but would be willing to help with a speaker project if there is anyone who can conquer this beast.
Last edited:
Its possible the mosfets are too far from the PCB. Not saying that's the exact problem, but usually lead lengths go 3-5" max.
This might sound bad, but you are having the same exact problem on both sides, so its consistent.
Also, nothing burned up. That's a plus as well. You are half way there. Don't give up.
It probably a reversed component somewhere or a resistor value is too high.
Checking your wiring and component placement 2 or 3 time will cost nothing but time.
Check everything again with no power to amp.
This might sound bad, but you are having the same exact problem on both sides, so its consistent.
Also, nothing burned up. That's a plus as well. You are half way there. Don't give up.
It probably a reversed component somewhere or a resistor value is too high.
Checking your wiring and component placement 2 or 3 time will cost nothing but time.
Check everything again with no power to amp.
I am sure they are not shorted unless the painted case is electrically conductive, I could space them off of the heat sink just in case, right now the leads are bent so that when I install the board there is a bit of spring tension from the leads to keep it in contact with the heatsink. What would happen if they were shorted?
I really hope that's not the issue. If I really have to, I could mount all devices on the left heatsink using a heat spreader, no new holes needing to be drilled, but I would be wasting the right heatsink.
The leads are not as long as they look, the chassis is really tiny, they are at most 6", I knew I was pushing it when I did that, but with my boards there was no other option since all devices were on one board. Didn't know there were boards with separate N and P channel PCBs until it was too late.
I checked all of the component values and orientations against the silkscreening AND the schematic, everything is in the right place, is the right value, and is the right orientation. The output devices are all in the right place, and are the right polarity. None are shorted to ground from any pins.
I'll check them again.
I have operated the power supply independently from the amplifier board, and have a bit over 32V unloaded, don't remember exactly how much.
I think I figured it out!!!
I'm using the IRF9240/IR240 (I think, I'll double check those transistors) same transistors as many use in the F5, but the IRF 9240 is a P type and the FQA19N20 is an N type.
So it could be that the output transistors are in the wrong place, piece of cake, just need to resolder a couple of them, and remount two on the opposite side.
I'm using the IRF9240/IR240 (I think, I'll double check those transistors) same transistors as many use in the F5, but the IRF 9240 is a P type and the FQA19N20 is an N type.
So it could be that the output transistors are in the wrong place, piece of cake, just need to resolder a couple of them, and remount two on the opposite side.
Just on other thing (which probably might not solve your issues yet): Solder the gate stopper resistors directly to the gates of the output MOSFETs to prevent the amp from HF oscillation. As far as visible on your photographs, you didn't.
Nothing easier than that. Remove them from the PCBs and solder them to the gate pins.
Not sure if the leads are long enough to do that, that's what I'm going to check. I don't have the amp in front of me until tomorrow when I go to work.
Could you explain why it can cause oscillation? I'm not contesting your statement, just don't know much about this.
<simplified>
Every mm of wire lenght is ~1nH. There is also some capacitance by example in the mosfet itself. A LC-network has a resonant frequency. increase either L or C or both and you'll lower that frequency.
So, by default, ringing will start to occur. If you move the resonant frequency up, there won't be enough bandwidth to amplify it to problematic proportions. Also, if the damping properties of the circuit (such as gate stoppers) are sufficiently high, each starting oscillation will stop. But some point the resonant frequency is low enough to be amplified by the semiconductors and damping will get seriously difficult. From there-on the probability of oscillation is too high to ignore.
Your used wire length is by 'rule of the thumb' and 'gut feeling' and 'experience' way too much. Ofcourse you can calculate how much is 'too long', but you could also take the safe route and prevent time wasted debugging.
more reading material http://en.wikipedia.org/wiki/Parasitic_oscillation
Every mm of wire lenght is ~1nH. There is also some capacitance by example in the mosfet itself. A LC-network has a resonant frequency. increase either L or C or both and you'll lower that frequency.
So, by default, ringing will start to occur. If you move the resonant frequency up, there won't be enough bandwidth to amplify it to problematic proportions. Also, if the damping properties of the circuit (such as gate stoppers) are sufficiently high, each starting oscillation will stop. But some point the resonant frequency is low enough to be amplified by the semiconductors and damping will get seriously difficult. From there-on the probability of oscillation is too high to ignore.
Your used wire length is by 'rule of the thumb' and 'gut feeling' and 'experience' way too much. Ofcourse you can calculate how much is 'too long', but you could also take the safe route and prevent time wasted debugging.
more reading material http://en.wikipedia.org/wiki/Parasitic_oscillation
Last edited:
What would this manifest itself as? Will it prevent the amplifier from working? What do I use to measure if it is oscillating? I have an oscilloscope but not too much experience on it. I want to get the amp working first, then go from there.
Worst case I have to move the transistors all to one heatsink, but then I won't get as much power out of it in class A.
HF Oscillation could be recognised in a higher-then-expected power draw with increased heat but no discernable other malfunction. LF osclilation can manifest as 'motorboating'.. Increased distortion is also a logical effect for both.
If you have an oscilloscope, connect it to the amplifier output. If what you see on the screen is not a steady, sharp line, there's noise and/or oscillation. If you can see a sinusoid waveshape on screen, it's most definitly oscillating. There are plenty osciloscope tutorials on the net to familiarize yourself with the sensible range to do your measurements in.
But first, get the amp working.
If you have an oscilloscope, connect it to the amplifier output. If what you see on the screen is not a steady, sharp line, there's noise and/or oscillation. If you can see a sinusoid waveshape on screen, it's most definitly oscillating. There are plenty osciloscope tutorials on the net to familiarize yourself with the sensible range to do your measurements in.
But first, get the amp working.
Thank you for your input!
I built my last F5 (normal, stereo but dual mono, one chassis) with the output transistors on wires, and it worked, with no oscillation that I saw on the scope (had help checking that when I brought it up) so I'll be crossing my fingers that I don't have issues with these.
Should have more info tomorrow.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.