Thanks for the lengthy and constructive feedback, it helps having skilled professionals like yourself participating in these topics explaining in great detail the intricacies of this particular amplifier's inner workings.
Many modifications not good.
Just off 2,2pF
Set capacitors in parallel to 47kohm
The original schema.
All other modifications are unsystematic, confusing and unnecessary
First, to remove the oscillations, as I wrote earlier. I drew red.
Federmann - I have checked the feedback capacitors, when I had this amplifier at my workbench. Feedback capacitors have added HF oscillations.
@Federmann,
Despite the fact I've already evaluated your suggestion prior to you making it I'll repeat the exercise and post the screenshot below from stimulating the input with a squarewave (PULSE(-1 1 0 100n 100n 50u 100u)). It simply doesn't work, see below noted screenshot from the simulation with C//47K = 10pF.
However the suggestion I've provided and exhaustively evaluated in the simulator works, please refer to the attached screenshot simulating the same squarewave.
Furthermore this has also been verified by two people that implemented these changes on their ExtremA and have looked at the step response independently, their results match the results I've obtained in the simulator.
So I welcome your feedback, but it needs to be a bit more in depth and constructive in order to be worthwhile.
Despite the fact I've already evaluated your suggestion prior to you making it I'll repeat the exercise and post the screenshot below from stimulating the input with a squarewave (PULSE(-1 1 0 100n 100n 50u 100u)). It simply doesn't work, see below noted screenshot from the simulation with C//47K = 10pF.
However the suggestion I've provided and exhaustively evaluated in the simulator works, please refer to the attached screenshot simulating the same squarewave.
Furthermore this has also been verified by two people that implemented these changes on their ExtremA and have looked at the step response independently, their results match the results I've obtained in the simulator.
So I welcome your feedback, but it needs to be a bit more in depth and constructive in order to be worthwhile.
Attachments
SSassen, if negative feedback C // 47K, leads to instability, it is bad.
Error in search Boucherot 10R + 1 nF ... Total negative feedback must fulfill the function to stability.
Error in search Boucherot 10R + 1 nF ... Total negative feedback must fulfill the function to stability.
I'm starting to wonder whether you've been privvy to the whole discussion leading up to the changes I've proposed to the ExtremA design?
Thanks for spotting that, you're absolutely correct. Corrected stuffing guide is attached below.
I'd like to thank you again Bohdan as I've just simulated the implications of this L1/R6 error, and the large DC fluctuations that you've reported and which Pavel has also commented on are clearly caused by swapping these parts around.
It was enough to measure the wrong voltage T1 and T2 and not add anything unnecessarily.
That's it.
That's it.
As you know, there is a proverb:
After a battle, everyone is a general.
Applies to you at 100%.
Pavel, first checking offset, and the current input differential pair It can not happen to me so fatal error, so I changed the involvement. Always trust the author that was functional involvement.
Paul, a man learns from mistakes.
An externally hosted image should be here but it was not working when we last tested it.
The layout is different
L1-R5-R6-R64
now
R5-R6-L1 moved R64
When was the difference?
You are misusing someone else's design images.
The original placement is at
Amplifier and powersupply PCBs
Unfair, at least.
Last edited:
That picture is for illustrative purposes only, it is not the final PCB, but one of the prototype PCBs.
- Status
- Not open for further replies.