I am working on troubleshooting what appear to be high (+10v) B voltages at the PS board outputs. What I have done so far.
1. Replaced all electrolytic caps and upgraded film caps
2. Replaced bridge rectifiers 3 bolt mount and 3 in board (1N4004's)
3. checked resistors and diodes in circuit and pulled a few. They all seem ok with DMM. Checked transformer output voltages 70 VAC and 33 VAC. Resistance at the B terminal on the driver/output boards is ~30M ohms. There is ~.5 V of AC measured at the B term. Measurements were made at idle with all boards plugged in and 8 ohm dummy at speaker terms.
Adjusted bias to exactly 50mV and output offset to 0 mV +-1mV and it is very stable. Scope shows perfect sine wave 100- 10,000 Hz. Clips at 46VAC.
My questions are:
1. Is a 15% voltage increase within design spec tolerance ? Seems high to me.
2. Can anybody tell me what the transform output AC voltages should be?
3. Can resistors measure properly at low voltages and fail at higher voltages?
4. What other checks can I do ? Is it possible the Driver/ output boards have too much resistance or draw too much current and I should go through testing them? What should the total idle resistance or current draw be by the amplifier? Seems odd that all 3 boards have issues, but I suppose if the PS electrolytics went bad then the boards may have all suffered equally... thinking more that the issues are in the input/output boards, which are one in the same on this design. I'll start by testing the transistors and move downstream. Just bought a Peak pro tester.
Other symptoms of the amp which seem off are the FETs get hot and the sound seems just ok, but should be much better. My updated GFA-555 sounds much better.
Update: After pushing it a bit (running 3 fans down drafting with a variable power supply) to the Dunlavy SC-IV's and SC-S2 sub it is sounding quite a bit better. I think all the polypropylene the film caps needed some burning in..... Wonder if these caps will lower their resistance after burning in ? - wouldn't think enough to affect the I*R parameters enough to affect B voltage.
Thanks
1. Replaced all electrolytic caps and upgraded film caps
2. Replaced bridge rectifiers 3 bolt mount and 3 in board (1N4004's)
3. checked resistors and diodes in circuit and pulled a few. They all seem ok with DMM. Checked transformer output voltages 70 VAC and 33 VAC. Resistance at the B terminal on the driver/output boards is ~30M ohms. There is ~.5 V of AC measured at the B term. Measurements were made at idle with all boards plugged in and 8 ohm dummy at speaker terms.
Adjusted bias to exactly 50mV and output offset to 0 mV +-1mV and it is very stable. Scope shows perfect sine wave 100- 10,000 Hz. Clips at 46VAC.
My questions are:
1. Is a 15% voltage increase within design spec tolerance ? Seems high to me.
2. Can anybody tell me what the transform output AC voltages should be?
3. Can resistors measure properly at low voltages and fail at higher voltages?
4. What other checks can I do ? Is it possible the Driver/ output boards have too much resistance or draw too much current and I should go through testing them? What should the total idle resistance or current draw be by the amplifier? Seems odd that all 3 boards have issues, but I suppose if the PS electrolytics went bad then the boards may have all suffered equally... thinking more that the issues are in the input/output boards, which are one in the same on this design. I'll start by testing the transistors and move downstream. Just bought a Peak pro tester.
Other symptoms of the amp which seem off are the FETs get hot and the sound seems just ok, but should be much better. My updated GFA-555 sounds much better.
Update: After pushing it a bit (running 3 fans down drafting with a variable power supply) to the Dunlavy SC-IV's and SC-S2 sub it is sounding quite a bit better. I think all the polypropylene the film caps needed some burning in..... Wonder if these caps will lower their resistance after burning in ? - wouldn't think enough to affect the I*R parameters enough to affect B voltage.
Thanks
Last edited: