Troubleshooting SS amp...30vdc on OUTPUT!

[Prattacaster]

I've replaced almost everything on the output board of this amp. Shematic here: http://s1122.photobucket.com/albums/...ter/Yamaha/GFS NYII?action=vGFS NYIIiew&current=Yamaha_G100_series_I2.jpg&newest=1
At point test point "5" I have about 30vdc, and after the output coupling cap I also have about 30vdc. The other speaker wire has no DC present, I did not replace this cap, is this a sign that it is leaky? It is a 2200uF 100v cap. When I got the amp to fix it almost all the transistors were blown due to a short from the output transistors mounting screw touching the copper tracing. When I fire it up with a bulb limiter it is a perfect dim glow with 100w bulb. All other voltages check out exept on the Output.

[R.G.]

I've replaced almost everything on the output board of this amp. Shematic here: http://s1122.photobucket.com/albums/...ter/Yamaha/GFS NYII?action=vGFS NYIIiew&current=Yamaha_G100_series_I2.jpg&newest=1
At point test point "5" I have about 30vdc, and after the output coupling cap I also have about 30vdc. The other speaker wire has no DC present, I did not replace this cap, is this a sign that it is leaky? It is a 2200uF 100v cap. When I got the amp to fix it almost all the transistors were blown due to a short from the output transistors mounting screw touching the copper tracing. When I fire it up with a bulb limiter it is a perfect dim glow with 100w bulb. All other voltages check out exept on the Output.

Point 5 must be exactly half the power supply voltage at pin 6. The output side of the cap must be 0Vdc.

The cap could be leaky or failed. Open the + terminal at point 5, and measure resistance from the - terminal to ground. This should measure about 830 ohms to ground. If the - to ground is open, one of the two resistors is burned open and this could account for the problem. If the resistance is about right, suspect the output cap. Be *sure* it's put in with polarity correct (i.e. + to point 5) because inserting it backwards will not only cause a voltage to appear on the output, it will permanently damage the cap.

[Jazz P Bass]

Have you hooked up a load yet?
I have seen on these Yamaha amps where the VDC will be present until you have a load hooked up.

[R.G.]

Have you hooked up a load yet?
I have seen on these Yamaha amps where the VDC will be present until you have a load hooked up.

Wow. That's a little bizarre. The 820R and 10R resistors are not enough to pull the 2200uF cap down to nearly ground?!?

I'd say if it has to have a speaker on it to get that terminal to ground, the cap is too leaky.

[Jazz P Bass]

I was thinking of the old EM-100 mixer.
Maybe they are different with the STK output IC.

[Prattacaster]

Thanks for replying guys, I checked those resistors before I started the thread, and they are both perfect,
A resistance test on the cap shows open, and my meter goes up to 30 or 40megs, if the cap is significantly leaky wouldnt it show much less resistance?

The cap is a chassis mount hog of a cap, it might be hard to find one that is the same size and mounts the same, any advice on that would be great,

another thing ...would it be ok to sub it with maybe a 1000uF instead of 2200uF?, I wouldnt think I would be a big deal but I'm not an SS guy.

[bkahuna]

Thanks for replying guys, I checked those resistors before I started the thread, and they are both perfect,
A resistance test on the cap shows open, and my meter goes up to 30 or 40megs, if the cap is significantly leaky wouldnt it show much less resistance?

The cap is a chassis mount hog of a cap, it might be hard to find one that is the same size and mounts the same, any advice on that would be great,

another thing ...would it be ok to sub it with maybe a 1000uF instead of 2200uF?, I wouldnt think I would be a big deal but I'm not an SS guy.

That would be fine to troubleshoot, but might affect your low end response.

[Prattacaster]

I found a cap that would fit pretty good, a couple more questions:

1)If I indeed have a severly leaky cap, wouldn't a resistance test be less than 10megs, I have run and rerun a resistance test and it seems that I get some conflicting readings, It would start at about 330k for a few seconds then it jumps to about 30 meg and then steadily decreases to under 6megs(couple minutes) and probably further if i left it clipped up.

2) When testing transistor via DMM diode drop, can a transistor test good but still be defective in the circuit, and wouldn't a defective transistor throw voltages off or maybe blow a fuse, (or draw too much current). Im using slightly underated fuses while fixing the amp, ive had absolutely no trouble with fuses or bulb indication.

3)When I first fired the amp up I was using a 60 w bulb, the voltage on the output read -20vdc(yes, negative), I then stuck in the 100w bulb, voltage read -10VDC, so it seemed as I let more current thru that my voltage on ouptut got closer to 0vdc. So I plugged straight into the wall and thats when I got 5vdc on output, I shut it off turned it on again and got closer to 30Vdc on ouput. I then hooked back into the bulb and then still got 30vdc on output, it seems like something went OPEN or malfunction at some point, sorry i didnt state this in the original post, I know how critical these timeline informants are to troubleshooting, especially via web.

[R.G.]

1)If I indeed have a severly leaky cap, wouldn't a resistance test be less than 10megs, I have run and rerun a resistance test and it seems that I get some conflicting readings, It would start at about 330k for a few seconds then it jumps to about 30 meg and then steadily decreases to under 6megs(couple minutes) and probably further if i left it clipped up.

The way a good capacitor should test is that it should show a very low resistance at first, then steadily increase until it goes off the top end of the scale. This is because most DMMs feed a current to the probes and read out the voltage the current causes. A capacitor eats all the current to zero volts initially and then charges up. Your readings make me deeply suspicious of the capacitor. The initial dip to 330K is all right, but then going up and back down are bad signs.

You can do a better test by putting a 10K resistor in series with the (-) side of the cap, then applying a voltage comparable to the full working voltage across the (+) terminal of the cap and the free end of the resistor. The voltage across the resistor will jump to nearly the full applied voltage, and then will begin decreasing as the capacitor charges up, taking almost two minutes to stabilize if both cap and resistor are good. The eventual voltage across the resistor tells you the leakage of the cap. If the end voltage is 1/10 of the applied voltage, the capacitor's internal resistance is 9X the value of the 10K, or 90K. If the end result is 1/100 of the applied voltage, the capacitor's internal leakage is 99 times the 10K, or 990K. Pretty much, if the end result is not more than 1M, the cap isn't something I'd leave in the amp.

Note that you will have to discharge the cap safely after charging it, by putting a RESISTOR, not a dead short, across its terminals until the voltage declines to nearly zero. Shorting a cap can damage it from the big current pulse.

2) When testing transistor via DMM diode drop, can a transistor test good but still be defective in the circuit, and wouldn't a defective transistor throw voltages off or maybe blow a fuse, (or draw too much current). Im using slightly underated fuses while fixing the amp, ive had absolutely no trouble with fuses or bulb indication.

Yes, but if you're not losing fuses or showing overcurrent, I would not sweat that right now.

[Prattacaster]

Thanks for the reply, I wired in a 2200uF 50v cap and still have 38VDC on the white speaker wire, I tried a different cap that I know for a fact works and got the same, 38vDC. I doesnt look like the cap is the problem, so where do I go from here..

Like I mentioned earlier, i replaced all but one transistor, what I couldnt get the exact replacement for I cross-referenced and found the closest replacement, could this be a small quirk in the substituted transistors, I also replaced the current bias trimmer, the others trimmers tested ok, I tested all resistors on the board replaced the bad ones, i replaced all but one of the electrolytic caps, i'll replace that too.

You can see that the voltage on the output is rising, it is near 40vdc now. It seems like everytime I flip the switch it creeps up a little more, maybe its the cap i havent replaced. I didnt mess with any of the poly and ceramic caps, I guess one of these could be leaky.

The voltage at pin 5 is about half (+/- 3v) of pin 6.

[R.G.]

Thanks for the reply, I wired in a 2200uF 50v cap and still have 38VDC on the white speaker wire, I tried a different cap that I know for a fact works and got the same, 38vDC. I doesnt look like the cap is the problem, so where do I go from here...
The voltage at pin 5 is about half (+/- 3v) of pin 6.

What this says is that the amp is balancing internal to the output cap, so the amp itself is probably OK, at least as far as offsets.

Here's the problem - this is a single supply amp, with only a single polarity power supply. The output **has** to sit at half the power supply. That output capacitor is there to block the DC but let the AC signal through. It must sit at half the power supply on point 5 but have it's output at ground for the speaker, which will die if there is substantial DC across it.

There is another possibility. It is possible that the 820R and 10R resistors don't connect to the capacitor (-) terminal, or to ground at the other end. With power off, measure resistance between the capacitor (-) terminal and the (-) terminal of the 2200uF/100V power supply capacitor. This should be 830 ohms. If the wire from the resistors to the power filter cap (-) is open, then it could cause this all by itself.

Like I mentioned earlier, i replaced all but one transistor, what I couldnt get the exact replacement for I cross-referenced and found the closest replacement, could this be a small quirk in the substituted transistors

No, it's not a small quirk of the subbed transistor.

[Prattacaster]

Here's the problem - this is a single supply amp, with only a single polarity power supply. The output **has** to sit at half the power supply. That output capacitor is there to block the DC but let the AC signal through. It must sit at half the power supply on point 5 but have it's output at ground for the speaker, which will die if there is substantial DC across it.

Good info, I wasnt ever about to hook the speaker up to it, from what ive read you want almost no dc at output, the speakers arent getting hooked up until that is acheived.

There is another possibility. It is possible that the 820R and 10R resistors don't connect to the capacitor (-) terminal, or to ground at the other end. With power off, measure resistance between the capacitor (-) terminal and the (-) terminal of the 2200uF/100V power supply capacitor. This should be 830 ohms. If the wire from the resistors to the power filter cap (-) is open, then it could cause this all by itself.

Sure ill check that , pin 5 voltage is adjustable via the center voltage trimmer, generally speaking all my voltages are slightly higher as should be expected, so I need to set the Co to half pin 6 b+. Cool.

[Jazz P Bass]

After you set the center voltage, measure the idle current.
This will tell what the output stage is doing.
The service manual specifically states that the Waveform, Center Voltage & The Idle Current should be tested with an 8 ohm dummy load.

[Prattacaster]

Yea, I dont have a dummy load, just a dummy. The idle current looked to be really high, it measured around 110 milltvolts from what i remember, this is between ground and the TP pin on the board. Ill double check that, like I said before the vdc on the output gets a little higher everytime I flip the switch, I dont think it will get much higher as it cant exceed pin 5 voltage.
I suppose making a dummy load for a 100 watt amp isnt something cheap, for small amps that would be an issue, just a 8 ohm high wattage resistor and some other goodies im sure. Is this a loss without a dummy load?

[Jazz P Bass]

Before you do anything else you should verify what R.G. said:
"There is another possibility. It is possible that the 820R and 10R resistors don't connect to the capacitor (-) terminal, or to ground at the other end. With power off, measure resistance between the capacitor (-) terminal and the (-) terminal of the 2200uF/100V power supply capacitor. This should be 830 ohms. If the wire from the resistors to the power filter cap (-) is open, then it could cause this all by itself."