Welll Certainly an interesting puzzle ..If a repairman.
Much like a persistent oil leak on some beater car will baffle some mechanics.
But, and it seems to me a genuinely relevant But:
The thing is worth 'mebe' 100$ in resale, optimistically even 125$ to the right buyer..in saleable condition.
Customer gets to pay a 2/300$ repair bill ? or is this all Pro Bono ?
Much like a persistent oil leak on some beater car will baffle some mechanics.
But, and it seems to me a genuinely relevant But:
The thing is worth 'mebe' 100$ in resale, optimistically even 125$ to the right buyer..in saleable condition.
Customer gets to pay a 2/300$ repair bill ? or is this all Pro Bono ?
Is the FET above suspicion ? They do crop up as problem parts in other gear (generally, not necessarily that particular device). If you have small croc clips to use as a heat shunt and are handy with braid then its only two minutes work to swap the FET with the other channel.
Have you tried dripping (as in a drop at time) freezer spray on any suspect parts ?
Have you tried dripping (as in a drop at time) freezer spray on any suspect parts ?
Hi gmarsh,
The basic problem is bias instability. I'm not sure a noisy dual JFET would cause drifting bias current. Not in this circuit with close to a CCS for tail current control.
You know, without having the unit on our bench, we can't really see how the drift in bias is behaving - really. We can only throw out guesses as to the problem. You are right about swapping the parts (not normally something I'd agree with).
If you have a solder sucker, use that first. The suction removes the heat energy along with the solder. Solder wick is great for cleaning up with.
-Chris
The basic problem is bias instability. I'm not sure a noisy dual JFET would cause drifting bias current. Not in this circuit with close to a CCS for tail current control.
You know, without having the unit on our bench, we can't really see how the drift in bias is behaving - really. We can only throw out guesses as to the problem. You are right about swapping the parts (not normally something I'd agree with).
If you have a solder sucker, use that first. The suction removes the heat energy along with the solder. Solder wick is great for cleaning up with.
-Chris
The craziest things are possible. After exchanging a pre driver stage current source transistor an amp worked again but output offset DC kept changing between + and - 70 mV. After lots of testing I found out the symmetrical current source trans switched on and off every few seconds. New failure mode? BC 550 loaded to the max.
I start with a sucker, finish with wick. Otherwise I go through wick way too fast
At this point I've measured every passive component and changed anything that remotely looks suspect, changed the capacitors, etc... pretty much all that's left in the amp that's original, that would influence the measured bias, is the transistors in that stage. The amp did fail with a shorted output, which would have seen the VAS stage driven hard in one direction or another, and I know TR05/TR07 got hot due to the discolored board.
I could spend a long time hypothesizing and measuring to pick out exactly what component is causing all this... or I can just spend a couple extra dollars and replace the whole shot, and hopefully be done with it.
Found the offending culprit - TR03. Swapped it out for a KSA992, changed nothing else, bias is now stable.
Out of curiosity, I tied the base/emitter together and applied a voltage to the collector, measuring current with a DMM, to see if there was anything weird going on. Well, with the power supply at about 50V (the transistor's rated for 80V) I now have a blown mA fuse in the DMM.... sigh.
Out of curiosity, I tied the base/emitter together and applied a voltage to the collector, measuring current with a DMM, to see if there was anything weird going on. Well, with the power supply at about 50V (the transistor's rated for 80V) I now have a blown mA fuse in the DMM.... sigh.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.