It's always a good idea to install FE first, test it with power, before installing output. By doing things in series, you slowly work to the end, checking as you go. That way if something is wrong, you wliminate certain steps. Check gate voltages of the mosfets. Pots could be switched. Basically you have one side contributing most of the output, almost like its in AB mode.
check that P1 is ok. you can do so by massure resistance over R5 without any power. you will want to see close to 0ohm.
I'm in a spot of trouble. Here's what I spent the evening doing.
While trying to bias the above F5t channel I noticed a few problems.
I tested the power supply, it works fine, unloaded voltage hovers right around 35v.
I set P1 and P2 all the way counter-clockwise and P3 somewhere in the middle.
Power on and nothing smokes which is a good sign.
I measure voltage across R11 and R12. R11 is at 930 mV, R12 is close to zero.
Uh oh.
Turning P1 doesn't affect the voltage across R11. P2 adjusts R12 just fine.
Offset at the output was quite high, between 180 and 230 mV.
I say screw it, attach a test load and source and I get music. Why did that work?
Now I'm sitting here scratching my head.Best thing I can come up with is a transistor is gone. Neither P-ch heatsink or N-ch got any warmer than ambient using an infrared gun.
Anyone have something for me to measure or try?
Best,
Nelson
@Peter, you were a lot quicker than my editing skills. Yes, it is the side which goes to high bias with a pot turned fully CCW that seems to be the issue.
Depending on how the board is laid out, the pot dials may need to be either facing the same way, or mirror-imaged. For example in the original F5 boards, most boards including Cvillers required similar orientation, but Peter's required mirror-imaging. I didn't know this, but measured the actual resistance from supply rail to the JFET drain to ensure I was starting off right.
Depending on how the board is laid out, the pot dials may need to be either facing the same way, or mirror-imaged. For example in the original F5 boards, most boards including Cvillers required similar orientation, but Peter's required mirror-imaging. I didn't know this, but measured the actual resistance from supply rail to the JFET drain to ensure I was starting off right.
The reason I question whether it is the pot or not is the fact that it should survive anything other than some fault that would push too much current through it. He should be able to have it fully on or off with nonissue. So the question becomes is it the pot , of so what fried it. Unhook the power fets and work on just the FE.
yes. but if the whiper is shorted to the outer leg that is not linked to the whiper , then it will massure 0ohm at any time.
EDIT: never mind. then it would be 0mV too.
Last edited:
i was thinking of solder error. but never mind
How's yours coming along . Songbird yet?
Ah ha! You are very correct. I've now got 400mV across R11 and R12 but there doesn't seem to be any warming up, that is, everything is still room temp. Is it likely I blew up the transistors upon startup at full bias? I can replace the jfets pretty quick but the outputs require a bit of disassembly and I would have to match more p-ch.
You guys are beyond helpful.
not yet. i'm goining to cut the sinks tonight. so i guess it will be singing tomorrow. i hope
If RS is the source resistors I may be out of luck. Here are my measurements:
Rails: 35.9 V
Across R11: 466 mV
Across R12: 488 mV
Across every source resistor (1 ohm each): 0 mV
P-ch gates: 34.26 V
N-ch gates: -34.26 V
Output offset: 20 mV
Been on for at least half an hour and there's no temp change.
Should I start pulling parts or is there something else to check?