I did some more digging into this and I think that the class D section that I had pictures of above is actually OK. I was using just a single probe without hooking the ground probe to anything. When I did an actual Vgs measurement using 2 different probes (Channel 1 - Channel 2), both gates looked good. I'm going to say that part is actually good now.
The amp still has an issue though. It will power up, but the power supply transformers do squeal a little bit. It draws about .5 amps at this point. If I hook a source up and play it, it seems to kick on to a higher state and pulls 1 amp. I do all of this without a load on the output. The drains of the power supply FET's have a very funny waveform on them (not square waves), but I am not sure if this is because the power supply is semi-regulated. The TL494 does not put out a 50% duty cycle pulse, so this is why I think the power supply is regulated. I checked the PS driver after the TL494, but they bug out as good both in and out of circuit. They are PNP only, not totem pole drivers.
If I drive an actual signal into the amp, I can see the output at the speaker terminals (using just the oscilloscope probe, no load), but once it reaches about 10V peak to peak, I get a large current current spike (hits the limit set on my supply), and it seems to go into a reset loop where it draws current, then drops. The power light seems to dim out during this point, but the protection light never activates. Besides just changing out the drivers and the TL494 because they seem to be in the problem area, what else should I be looking for? The power supply FET's themselves seem fine. I'm really hoping it is not from the missing resistors earlier in the post because I replaced them with what Perry had in his reference material. It has the same symtoms of a grounding issue somewhere because plugging and unplugging the RCA's make an audible difference in the transformer squeal and I've had "similar" issues when I left a ground strap disconnected on some kicker amps. I have not found anything obviously wrong in this regard though.
The amp still has an issue though. It will power up, but the power supply transformers do squeal a little bit. It draws about .5 amps at this point. If I hook a source up and play it, it seems to kick on to a higher state and pulls 1 amp. I do all of this without a load on the output. The drains of the power supply FET's have a very funny waveform on them (not square waves), but I am not sure if this is because the power supply is semi-regulated. The TL494 does not put out a 50% duty cycle pulse, so this is why I think the power supply is regulated. I checked the PS driver after the TL494, but they bug out as good both in and out of circuit. They are PNP only, not totem pole drivers.
If I drive an actual signal into the amp, I can see the output at the speaker terminals (using just the oscilloscope probe, no load), but once it reaches about 10V peak to peak, I get a large current current spike (hits the limit set on my supply), and it seems to go into a reset loop where it draws current, then drops. The power light seems to dim out during this point, but the protection light never activates. Besides just changing out the drivers and the TL494 because they seem to be in the problem area, what else should I be looking for? The power supply FET's themselves seem fine. I'm really hoping it is not from the missing resistors earlier in the post because I replaced them with what Perry had in his reference material. It has the same symtoms of a grounding issue somewhere because plugging and unplugging the RCA's make an audible difference in the transformer squeal and I've had "similar" issues when I left a ground strap disconnected on some kicker amps. I have not found anything obviously wrong in this regard though.
Next update on this amp. I replaced both PS drivers and that stopped the random excessive current draw. When out of the circuit, one of the drivers was good and the other was "OK". When I say that it was not obviously shorted or open, but outside the range of normal. I replaced both with a BD140-16 for improved thermal handling. The gate drive waveform still was not perfect on the falling edge, but this is because this amp does not use an emitter follower (PNP+NPN pair). It only uses a PNP to pull the gate down with a diode to drive it high. Once the gate hits ~0.7V, it slowly decays because the BJT is not fully on anymore.
There are still some things I do not like with the amp performance though. This amp has the positive output voltage rail tied with a divider into pin 1 (non-inverting error amp). I am not sure if the resistor (R186, 84k) Perry referenced above is from the V2 version of this amp. With that value of resistor in there, with no load attached I get about +/-55V on the output rails. The issue is that the +/-15V regulator only gets to about +/-8V on the inputs so their output is roughly a volt less than that. That means all the input section opamps are getting less voltage and the LDO's are not operating in their ideal area. In general, I don't know that this causes tons of problems, but the thing that still concerns me something is wrong somewhere else is that the power light either barely illuminates or does not illuminate at all. If I force the TL494 to put out max pulse width (ground pin 1), the rail voltages go to +/-80V and the +/-15V regulators are at regulation. Nothing seems wrong with the amp other than it runs a little warmer now because it is at max power supply output all the time. The power LED still do not seem to act correctly. Sometimes it will be at full illumination and other times it will be dim, even with the forced output voltage. The power LED is tied directly to the regulated +15V through a current limiting resistor. What appears to be happening is the cathode/"ground" voltage is varying but I cannot understand why. I might be missing something on this, but it is very basic. I do not understand why the LED is acting this way and it makes me wary that something else in the amp is not acting correctly either.
When I forced the power supply to run unregulated, I could get clean output up to about 180watts into a 4 ohm load. I was hitting the current limit of my bench supply then. Is there a danger in just changing the 84k resistor to something higher like 100k to change the regulation point so the power supply puts out wider pulse widths than before, but still not always running full tilt?
There are still some things I do not like with the amp performance though. This amp has the positive output voltage rail tied with a divider into pin 1 (non-inverting error amp). I am not sure if the resistor (R186, 84k) Perry referenced above is from the V2 version of this amp. With that value of resistor in there, with no load attached I get about +/-55V on the output rails. The issue is that the +/-15V regulator only gets to about +/-8V on the inputs so their output is roughly a volt less than that. That means all the input section opamps are getting less voltage and the LDO's are not operating in their ideal area. In general, I don't know that this causes tons of problems, but the thing that still concerns me something is wrong somewhere else is that the power light either barely illuminates or does not illuminate at all. If I force the TL494 to put out max pulse width (ground pin 1), the rail voltages go to +/-80V and the +/-15V regulators are at regulation. Nothing seems wrong with the amp other than it runs a little warmer now because it is at max power supply output all the time. The power LED still do not seem to act correctly. Sometimes it will be at full illumination and other times it will be dim, even with the forced output voltage. The power LED is tied directly to the regulated +15V through a current limiting resistor. What appears to be happening is the cathode/"ground" voltage is varying but I cannot understand why. I might be missing something on this, but it is very basic. I do not understand why the LED is acting this way and it makes me wary that something else in the amp is not acting correctly either.
When I forced the power supply to run unregulated, I could get clean output up to about 180watts into a 4 ohm load. I was hitting the current limit of my bench supply then. Is there a danger in just changing the 84k resistor to something higher like 100k to change the regulation point so the power supply puts out wider pulse widths than before, but still not always running full tilt?
I wanted to close out this repair. I ended up changing the reference point for the negative voltage rail and the amp works great now besides the LED (more on that later). I changed the R187 from 33k to 39k. This ensured that the power supply would raise the positive rail voltage slightly higher to keep all of the other ancillary circuits on. It just seems that the positive output rail has more current draw than the negative, so at idle not all of the circuits powered off the positive rail were getting sufficient voltage and staying shutdown.
The low input voltage on the 15V regs at idle seems to be OK as long as the input opamps have at least 7V on them. The main reason for concern for me was the power LED being inconsistent. It turns out it really just seems to be a function of that LED. I drove the LED off of a power supply separately and it would act normal at 10mA drive, but would sometimes light and sometimes not at anything 6mA and below. When the amp is idling, the current through the LED is normally 3-6mA based on measurement. So that explains why the LED was acting abnormally. Once the amp is actually producing power, the LED is on like it is supposed to and everything else is normal.
The low input voltage on the 15V regs at idle seems to be OK as long as the input opamps have at least 7V on them. The main reason for concern for me was the power LED being inconsistent. It turns out it really just seems to be a function of that LED. I drove the LED off of a power supply separately and it would act normal at 10mA drive, but would sometimes light and sometimes not at anything 6mA and below. When the amp is idling, the current through the LED is normally 3-6mA based on measurement. So that explains why the LED was acting abnormally. Once the amp is actually producing power, the LED is on like it is supposed to and everything else is normal.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.