An amp capable of 20MHz full power bandwidth to pass it, is a very slim chance. How high was the oscillation peak to peak on the Pumpkin output anyway?
Drop the Alps in favor to a cheap DACT type switch pot. That will make it sharper subjectively.
Work finished early today 
Good, at least you now know the oscillation can be stopped.
Because the Pumpkin is balanced in and out, it therefor has two feedback paths, only one half needs to play up to cause problems on both sides.
What you need to do next is to turn it into an unbalanced pre - easily done.
With the inputs grounded and no links on R1 & R3, fit a shorting link across R7 (obviously with the power off) this will bias off Q7, the feedback will be via R5 only, power up and check J4, there should be no output unless there is something wrong with Q7, if there is - don't proceed with the next bit; check that Q2 gate is at 0V, if not - fit a link across R4; check for oscillation at J3.
Remove the links on R7 R4, repeat the above procedure with R8 R2 J4.
Don't feel too
You will get there in the end
Good, at least you now know the oscillation can be stopped.
Because the Pumpkin is balanced in and out, it therefor has two feedback paths, only one half needs to play up to cause problems on both sides.
What you need to do next is to turn it into an unbalanced pre - easily done.
With the inputs grounded and no links on R1 & R3, fit a shorting link across R7 (obviously with the power off) this will bias off Q7, the feedback will be via R5 only, power up and check J4, there should be no output unless there is something wrong with Q7, if there is - don't proceed with the next bit; check that Q2 gate is at 0V, if not - fit a link across R4; check for oscillation at J3.
Remove the links on R7 R4, repeat the above procedure with R8 R2 J4.
Don't feel too
You will get there in the end
Last edited:
R1 and R3 are the input resistors. When you short these out, and then short the inputs, the amplifier is running at very high closed loop gain, so in effect the loop gain is very low. Under these conditions I would expect the oscillation to stop. This seems to suggest you have a loop stability problem.
at the moment Vbe from Q3 is 0,6V divided by R7 (0,47) equals 127mA for one side
you get 100mA by 0,6V divided by 100ma gives 0,6 Ohm......
you can read it in the chapter how shunty works......
http://dl.dropbox.com/u/20665608/The Pumpkin & Shunty.pdf
in much better English than mine.....
you get 100mA by 0,6V divided by 100ma gives 0,6 Ohm......
you can read it in the chapter how shunty works......
http://dl.dropbox.com/u/20665608/The Pumpkin & Shunty.pdf
in much better English than mine.....
Last edited:
The components that are not hugging the boards are those that have been added during this conversation. I am trying very hard not to damage the delicate PCB.
Attachments
0.6V divided by 127mA is....0.6 divided by 0.127, ...is 6 divided by 1.27,...is 4R7.
0.6 divided by 0.1 is ... 6 divided by 1, is...?
(but then in much better English)
Last edited: