Thank you Bimo
Now I understand. In my sim, a 10meg resistor lets 10uA pass.
Would you care to explain what it does ?
Now I understand. In my sim, a 10meg resistor lets 10uA pass.
Would you care to explain what it does ?
In theory, it will corrected unbalance of VAS's base current. But I never try it. Please try in your amplifier to see if VAS current more stable or not.
This is not the final solution. With or without that 10meg resistor, the standing current of the VAS's is still ill defined.
Cheers, E.
Cheers, E.
Now I am listening and allready can state that this topology provides a very pleasant sound.
Voices are better defined than my blameless version although it does not have the same bass grip as the latter.
There is an impressive response to mid band transients.
Will try the mega ohm resistor latter.
Voices are better defined than my blameless version although it does not have the same bass grip as the latter.
There is an impressive response to mid band transients.
Will try the mega ohm resistor latter.
I have used TMC in a number of my CFA bipolar output amps (incl. a big 240 W commercial amp). It indeed provides significant loop gain enhancements at HF and improved distortion performance. At 200 W out into 8 Ohms the distortion is 7 ppm.
My simulation of this topology using PSU 63VDC. At low level output, 20kHz, 8 Ohm almost no distortion. I use OITPC compensation by Dadod.
Attachments
The VAS current depends on the offset matching between the LSK489 and the LSJ689. This is a common behavior for full-complementary input stages. The shunt resistor across the collectors of the IPS current mirror stabilize it. Without those resistors the VAS current is theoretically undefined. The smaller those resistors, the better defined it is, but at some expense of IPS gain. The choice in the design is thus a compromise. The VAS bias current matters little, and a deviation during operation (e.g., warmup) from 12.0 to 12.7 mA is quite good. This issue is covered in the IPS/VAS chapter in my book.
Cheers,
Bob
Sorry. I design it for a friend.
Other solution is you can add common mode control loop in HEC IPS (other amplifier of Bob Cordell) instead of resistor in current mirror's collectors.
Now I try to sim my idea.
Clipping is good even at 20kHz because using 2 diode anti parallel in collector of current mirror which series a small resistor. Sorry, I'm not using my computer now so I can not show the simulation result.
This is my 100W CFA OITPC (OPS serial combination of ClassA and ClassB) clipping, simulated and real amp clipping.
Damir
Attachments
Now I sim HEC input with OITPC compensation. LTP is JFET in CFP with BJT. The VAS did not cascode because loop gain is high enough. THD at 20kHz is slightly worst than symmetry topology that I sim before in my previous post.