AJT: did you try adjusting tail currents?
RC: Yes, increasing big ccs current slightly reduces offset but only slightly, from .3 to .25v
Increased driver stoppers to 220r and the problem persists.
Now I will reduce feedback and see what happens
AJT: we have progress, you are on track now, can you post a pdf version of you schematics, these tired old eyes are hard up looking at your simulation pictures...
i remember i also asked you the polarity of the 0.3 mV,? is it positive or negative?
RC: Reduced feedback and oscilation remains now at 250khz
Running out of ideas
The offset is positive
I have 0.3Vdc and 10Vac at 250khz
The zobel resistor allready changed color because of the heat
RC: Yes, increasing big ccs current slightly reduces offset but only slightly, from .3 to .25v
Increased driver stoppers to 220r and the problem persists.
Now I will reduce feedback and see what happens
AJT: we have progress, you are on track now, can you post a pdf version of you schematics, these tired old eyes are hard up looking at your simulation pictures...
i remember i also asked you the polarity of the 0.3 mV,? is it positive or negative?
RC: Reduced feedback and oscilation remains now at 250khz
Running out of ideas
The offset is positive
I have 0.3Vdc and 10Vac at 250khz
The zobel resistor allready changed color because of the heat
offset is positive means reduction of tail currents is to be done,
you can do this two ways, via small ccs, or the big ccs...
either way, one or both needs reduction in current,
via increased value of emitter resistors...
again my recommendations are:
1. remove the variable pot at emitters of the top current mirrors.
hust two 220 ohms resistors and you are done...
2. small ccs, put emitter resistors in series with a fixed resistor...
you can use an off board variable port to adjust bias and then
when the offset is trimmed to 0 volts, you can replace that with a resistor
of the nearest standard value...
you can do this two ways, via small ccs, or the big ccs...
either way, one or both needs reduction in current,
via increased value of emitter resistors...
again my recommendations are:
1. remove the variable pot at emitters of the top current mirrors.
hust two 220 ohms resistors and you are done...
2. small ccs, put emitter resistors in series with a fixed resistor...
you can use an off board variable port to adjust bias and then
when the offset is trimmed to 0 volts, you can replace that with a resistor
of the nearest standard value...
While searching for sources of oscillations I found this paper that gave me an idea.
Maybe the cascode in the LTP is where the collpits is.
C6 is not refered to GND and as there is no zener to bias the cascode I now see this cap does not decouple noise.
I will remove it first.... then if needed I will place a cap from the cascode bases to gnd
Maybe the cascode in the LTP is where the collpits is.
C6 is not refered to GND and as there is no zener to bias the cascode I now see this cap does not decouple noise.
I will remove it first.... then if needed I will place a cap from the cascode bases to gnd
You are using old and probably bad simulation models. I suggest you delete them and use these:
CordellAudio.com - SPICE Models
CordellAudio.com - SPICE Models
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.