Hello! I have a Crate PCM-8 that had no output. I have not found an exact schematic for this amp but have found a few close enough to diagnose the obvious problem. This version of the PCM-8 has a single ne5532 opamp at the front end of the power amp board. Next to the opamp there is a J111 jfet that is shorted. This short is feeding the output back into the inverting input effectively silencing the amp. Pulling the jfet out the circuit gets the amp working again. I can tell that someone has previously replaced this jfet as well as a nearby j113 jfet. This tells me that there is either another problem or a design flaw of sorts. I'd like to fix this amp and have it be dependable. Does anyone has a schematic for this amp with a single ne5532 and a feedback (protection) jfet in the opamp circuit? Also any insight on why this failure happens or how to repair it for good would be appreciated.
I reassembled the amp enough to try to put a signal through it. I could get some sound but it became obvious that the preamp has issues as well After pulling the preamp board, the only obvious issues is a 220 ohm resistor in circuit that lights the protection led was burned and open circuit. Replacing it didn't effect any changes.
The biggest issues is that the main volume pot has little to no effect on sound level and often leads to a loud howling oscillation when trying to adjust it. The associated opamp tests ok in terms of voltage drop on the semiconductor junctions. I haven't pulled it or tried another yet. The other thing that has me a bit concerned is that the 16V + and - rails aren't very even, +17.1 and -14.7 respectively. I've removed all the load from the voltage supply circuit and the discrepancy remains. There are 16V zeners that connect to the base of the voltage supply transistors and I've considered swapping the diodes to see if the discrepancy follows. I wanted to get opinions if I should pursue the opamp rail voltage discrepancy before going through the hassle. Any other ideas appreciated.
The biggest issues is that the main volume pot has little to no effect on sound level and often leads to a loud howling oscillation when trying to adjust it. The associated opamp tests ok in terms of voltage drop on the semiconductor junctions. I haven't pulled it or tried another yet. The other thing that has me a bit concerned is that the 16V + and - rails aren't very even, +17.1 and -14.7 respectively. I've removed all the load from the voltage supply circuit and the discrepancy remains. There are 16V zeners that connect to the base of the voltage supply transistors and I've considered swapping the diodes to see if the discrepancy follows. I wanted to get opinions if I should pursue the opamp rail voltage discrepancy before going through the hassle. Any other ideas appreciated.
Is the J111 actually shorted, or did it just lose gate drive? It’s a depletion mode device, and if the gate is just floating (or shorted to source) it will read a low resistance. If left in the circuit without a negative bias on the gate, it will act as a short. The circuit driving it may actually be at fault.
I was /am a bit unsure about that. I get ~30 ohms across the j111 and IIRC its supposed to be 250-300 ohms. I did find the burned resistor going to the protect led in the preamp so something bad definitely happened in the protect circuit. I havent looked much further into the protect circuit as of yet. I am holding out hope to turn up a relevant schematic before I draw it out and miss an important detail.
What does it measure with Vgs set to -25 volts? If the Vgs won’t hold that voltage thru say a 100k resistor (too leaky) or if it still measures low ohms then I’d say Houston has a problem.
Houston definitely has a problem. I socketed the opamp at the front of the power amp. With the j111 back in the circuit and the opamp out, I have ~-3.5V on the gate and within .5V on the drain and source. With the opamp back in, all 3 pins of the j111 get driven to the opamp's negative rail voltage.
Unfortunately, no. I bought it not working for cheap with the story that it came from a local high school.
In principle "everything is repairable" in some way or another.
The problem I justifying time and expense.
An experienced Tech will definitely be able to repair that amp, as is, but at $65/85 an hour bench time, $120 up in NY or Nashville, plus parts, it quickly surpasses amp worth.
Short practical answer: if amp works fine without J111, just remove it and call it a day.
It is only a turn ON/OFF thump remover, nothing essential.
It would be "nice to know" what happened but budget does not allow that, so be it.
The problem I justifying time and expense.
An experienced Tech will definitely be able to repair that amp, as is, but at $65/85 an hour bench time, $120 up in NY or Nashville, plus parts, it quickly surpasses amp worth.
Short practical answer: if amp works fine without J111, just remove it and call it a day.
It is only a turn ON/OFF thump remover, nothing essential.
It would be "nice to know" what happened but budget does not allow that, so be it.
If it’s just a turn-on delay for the signal chain, then no big deal. in most cases I’ve seen, they also serve as the compressor, which you usually still want.
If it came from a high school, somebody might have plugged a line in jack or effects return straight to the wall or the speaker out of another big amp. Don’t under estimate the power of stupid^H^H^H^H^H^H the force.
If it came from a high school, somebody might have plugged a line in jack or effects return straight to the wall or the speaker out of another big amp. Don’t under estimate the power of stupid^H^H^H^H^H^H the force.
These are the closest schematics I could find. There is another version of the power amp schematic that can be found but that version has a quad opamp where my version has the lone ne5532 in the power amp.
That one doesnt have the Jfet clamp, and is using an optocoupler for the compressor/limiter. Slower response time, but still fast enough to get rid of the worst of the clipping. Less audible artifacts. But not RoHS compliant (cadmium).
You show 2 schematics.
wgski mentions the one on the right without Jfets.
The one on the left has two 😲
One shorting Op Amp output to input , and for good measure shunting Op Amp output to ground, which is poor Engineering, they should add some resistor in between.
Both start unbiased so very low resistance and are negative biased by apparition of the -16V rail.
As wgski found, in both amps (typical of Crate), compression is achieved by a Led-Ldr Vactrol.
Led is turned on when input Op Amp "gets crazy" when amp clips and loses NFB
wgski mentions the one on the right without Jfets.
The one on the left has two 😲
One shorting Op Amp output to input , and for good measure shunting Op Amp output to ground, which is poor Engineering, they should add some resistor in between.
Both start unbiased so very low resistance and are negative biased by apparition of the -16V rail.
As wgski found, in both amps (typical of Crate), compression is achieved by a Led-Ldr Vactrol.
Led is turned on when input Op Amp "gets crazy" when amp clips and loses NFB
The resistor is already there. It should be on the other side of R3.
Similar designs use the FETs as variable resistors, where they assume the compressor function. Then it is sensitive to what TYPE is used, as it can go out of range. As clamps, if somebody moves the pinch off or cuts IDSS it’s not the end of the world. The linear region gets ignored (full on or full off) so type substitutions are less likely to create issues.
Bipolars actually work better as clamps, but they don’t default to “on” with the circuit unpowered. This relies on that happening.
Similar designs use the FETs as variable resistors, where they assume the compressor function. Then it is sensitive to what TYPE is used, as it can go out of range. As clamps, if somebody moves the pinch off or cuts IDSS it’s not the end of the world. The linear region gets ignored (full on or full off) so type substitutions are less likely to create issues.
Bipolars actually work better as clamps, but they don’t default to “on” with the circuit unpowered. This relies on that happening.
I finally made real progress on this amp. I figured out why the resistor for the limit LED was burned. Turns out that if you inadvertently connect the jumper between the preamp and the power amp backwards, that resistor becomes a fuse. Doh! Stupid mistake but mystery solved. So it would appear that somebody trying to go in and fix the original problem caused this "newer" problem. I feel confident now to order in replacement jfets and restore the proper function of that subcircuit. I'll also mark the board to keep it from happening again.
Back to the original problem which showed itself after pulling the shorted jfet, the master volume knob unresponsive and/or loudly popping and cracking with painful oscillation. Early on I had pulled that pot and disassembled it to clean and re-tension the wiper contacts. Much to my disappointment that yielded absolutely no improvement. The problem was the main volume pot, but not in the way I had anticipated. The pot has a board with the carbon traces printed onto it and the legs are riveted to the board. That rivet connection is also the electrical connection between the traces and the legs. The rivets had lost enough tension that the electrical connection was intermittently (mostly) failing. Removing and disassembling the pot a second time to put a bit more crush in these rivets has totally solved the problem. You can now turn the pot from stop to stop with absolutely no scratchyness in the output. With 350 watts into a 2x15, this thing is LOUD.
Back to the original problem which showed itself after pulling the shorted jfet, the master volume knob unresponsive and/or loudly popping and cracking with painful oscillation. Early on I had pulled that pot and disassembled it to clean and re-tension the wiper contacts. Much to my disappointment that yielded absolutely no improvement. The problem was the main volume pot, but not in the way I had anticipated. The pot has a board with the carbon traces printed onto it and the legs are riveted to the board. That rivet connection is also the electrical connection between the traces and the legs. The rivets had lost enough tension that the electrical connection was intermittently (mostly) failing. Removing and disassembling the pot a second time to put a bit more crush in these rivets has totally solved the problem. You can now turn the pot from stop to stop with absolutely no scratchyness in the output. With 350 watts into a 2x15, this thing is LOUD.