Official LYNX Power Amp builder’s thread

[sryun]

First, C 101 and C109 is removed, and the second, the new capacitor (Cnew)is needed.
C101 is a kind of high cut filter (low pass filter) but its presence makes LYNX very unstable and reults oscillation almost always.
There are 2 kinds of oscillation waves. Around 2 MHz and around 200~300kHz.



The 2 MHz oscillation wave is about 2 Volts (in my memory, 4 v p-p) and behaves as mild form. The 200~300 kHz oscillaiton looks like sqare wave, is more critical, has very high voltage and always fries many transistors (Q105, Q106, Q107, Q110, Q111) and resistors (see below schematics and figures). I could catch a snapshot just before turning off power switch (using both hands).



C109 is located in feedback circuit but it makes LYNX also unstable. Unstability means it manifestes various unusual phenomena,example, even the "Pre-test" passed but after the output transistors wiring, abrupt oscillation is developed, and another example is that oscillation is developed when decreasing bias voltage (maybe because of change of output transistor contribution?) I performed many trials with the values of C 101 and C109, from 0pF to 330pF, I reached the conclusion that it is best when both capacitors (C101and C109) are abscent!



Cnew is between pin 2 and pin 6 of OP amp. Acording to OPA627 datasheet they recommend some capacitors for stabilty.



Other DIYer, BV also mentioned it in this thread. You also applied it in OPA134.

The best value of Cnew is around 1,000pF (below figures). C102 and C103 value are 5pF.
Below the level of 220pF of Cnew, it also produce oscillation.







I wish my experience could be a help to other DIYer.
And also wish I can see your schematics in your site again.
Thanks for your effort to introduce original, very great sound-producing LYNXs amp.

[sryun]

I forgot to mention the test square wave in oscilloscope screen.
It's 10kHz square wave.

Thanks again, Jan.

[AndrewT]

What are in the images that don't open?

[sryun]

There are critical errors in image loadings in Post number #660 and #661.
I feel really sorry for inconvenience.

But you can see the files by mouse right clicking and saving image ("Save target as...") in post #660 directly.

I attach the image files of #660.

[sryun]

I attach the image files #661 (1st~6th).

You can also see the files by mouse right clicking and saving image ("Save target as...") in post #661 directly.

[sryun]

I attached the image files of #661 (7th~14th)

All images are the result of 10kHz square wave input.

The value of Cnew (between pin 2 and pin 6 of OPA627) is marked in each image. You can also see the C102 and C103 values in the image.

I think the 1000pF (with C102 and C103 5pF) image is the best looking figure.

Cnew value 220pF was the threshold (I see severe ringing).
Below the 220pF, oscillation occurred.