Hi, please excuse the thread revival, but post #2 in this thread is the closest I've seen to being representative of the S300 I'm working on.
(I should note that the one I'm working on has MPS-U10/-U60 pre-drivers, though, which I think might make it a little older than the schematic. The number on the PCB is 550486, if that's useful.)
Owner's complaint was hot running. Both channels respond appropriately to offset adjustment - and seemingly to bias adjustment, too, as long as the offset has been zero'd first. Both channels can produce some output, but one distorts severely.
The distorted channel also exhibits a behavior I've never encountered before: When the output swings + (either dynamically, or just setting the offset to +100-200 mV or so), both halves of the complementary section (last 3 stages) conduct increasingly as the + voltage swing increases. (This is with no output load.)
As the + swing increases, the base-to-base voltage measured at the pre-drivers increases - I'm baffled about why that would be. The current in the CCS that loads the VAS stage does not change, at least not the emitter current. I suspected some weird leakage in the (-)-going pre-driver (labeled 2N5415 in post #2) , but I've tried replacing it with no behavior change.
I'm stumped. There's either some interesting difference in my unit's circuit vs the post #2 schematic (likely enough) or I'm missing something (very likely).
Thanks for any clues!
(I should note that the one I'm working on has MPS-U10/-U60 pre-drivers, though, which I think might make it a little older than the schematic. The number on the PCB is 550486, if that's useful.)
Owner's complaint was hot running. Both channels respond appropriately to offset adjustment - and seemingly to bias adjustment, too, as long as the offset has been zero'd first. Both channels can produce some output, but one distorts severely.
The distorted channel also exhibits a behavior I've never encountered before: When the output swings + (either dynamically, or just setting the offset to +100-200 mV or so), both halves of the complementary section (last 3 stages) conduct increasingly as the + voltage swing increases. (This is with no output load.)
As the + swing increases, the base-to-base voltage measured at the pre-drivers increases - I'm baffled about why that would be. The current in the CCS that loads the VAS stage does not change, at least not the emitter current. I suspected some weird leakage in the (-)-going pre-driver (labeled 2N5415 in post #2) , but I've tried replacing it with no behavior change.
I'm stumped. There's either some interesting difference in my unit's circuit vs the post #2 schematic (likely enough) or I'm missing something (very likely).
Thanks for any clues!
So, define running hot, the heat sinks should be around 50C on the SA/3. The power supply is simple but you just need to check the voltage rails are good +/- ? after the rail fuses! If so, then feed the input a sine wave and walk though the Front End board with an Oscope and see where the signal goes bad and look for problems in that area of the circuit. I have worked on many of the SA models and the output devices are not used to their limits so are very reliable.
Thanks for the suggestions. I don't think new front end boards are economically viable for a customer project, and I think I'm past the point where signal tracing will help to further narrow down what's wrong with the original board.
It feels like there is already a Big Hint in the observation that the base-to-base voltage of the predriver stage increases as the output voltage goes positive, but I can't figure out what could cause that when the current in the VAS stage doesn't appear to change. Any thoughts on that aspect?
It feels like there is already a Big Hint in the observation that the base-to-base voltage of the predriver stage increases as the output voltage goes positive, but I can't figure out what could cause that when the current in the VAS stage doesn't appear to change. Any thoughts on that aspect?