In the original post, there looked to be 8 active devices on the bar across the case, are those voltage regulators or?
Ah well, I found a source for them from my usual shop. what about 2sc2240/2sa970 as a substitute for the input devices? Pouring over the datasheets tell me yes but
Because the beta tester was me, it's probably some very simple mistake that I just could not see.
Thanks!
The Beta proto arrived around noon.
I was actually working on another phono preamp, but I was curious what was happening.
So it was oscillating, quite badly actually.
After trying a hundred different things all afternoon, including some micro-surgery on the PCB, I was actually at a lost.
But I happened to have my working example on the bench, so I made a visual comparison.
Note that the Public Version of the PCB is different from my own.
It is simplified, a bitmore spread out, and accomdates the original JLH auto-bias circuit.
In the end the one single cause of the instability is the long legs of the power transistors.
Morde was using 10mm stand-offs. Reducing them to 5mm, and the amp was stable.
And that took 8 hours !!!!!!
But unluckily I also tried all these other things to try to improve stability.
So now I know what has which effect, etc. Also educational for me.
The BoM will be updated later after Morde has finished the second proto successfully.
Too late now, so photos tomorrow.
Patrick
I was actually working on another phono preamp, but I was curious what was happening.
So it was oscillating, quite badly actually.
After trying a hundred different things all afternoon, including some micro-surgery on the PCB, I was actually at a lost.
But I happened to have my working example on the bench, so I made a visual comparison.
Note that the Public Version of the PCB is different from my own.
It is simplified, a bitmore spread out, and accomdates the original JLH auto-bias circuit.
In the end the one single cause of the instability is the long legs of the power transistors.
Morde was using 10mm stand-offs. Reducing them to 5mm, and the amp was stable.
And that took 8 hours !!!!!!
But unluckily I also tried all these other things to try to improve stability.
So now I know what has which effect, etc. Also educational for me.
The BoM will be updated later after Morde has finished the second proto successfully.
Too late now, so photos tomorrow.
Patrick
As promised, including frequency responses with or without load.
You can see when when not loaded the amp has a slight (1.5dB) overshoot at ~ 2MHz.
But it is now kept stable by the Zobel network.
At 30R load, it is stable on its own.
Patrick
.
You can see when when not loaded the amp has a slight (1.5dB) overshoot at ~ 2MHz.
But it is now kept stable by the Zobel network.
At 30R load, it is stable on its own.
Patrick
.
