VSSA Circlotron

[UltimateX86]

yes I will add the cascade.
you use a current of how much?

 

[UltimateX86]

separate B+ supply line just for the VAS stages , but that separate B+ line can be derived from OPS B+/B+`lines, via two resistors and one smoothing electrolytic capacitor connected to ground ,it is somethig like this :

Good idee but why not the input ?

 

[banat]

x

by inserting positive fbck(bootstrap) on input JFET`s drains you got max. gain with minimum thd , I think that if you supply input JFET stages from separate B+ line than cascode can help very much perfomance wise , but maybe bootstrapped cascode input stage is even better performer ?

 

[N101N]

I am not sure, it depends on the supply voltage and the cascoding transistor, which will have the same current. A high supply voltage restricts the biasing of output stage in particular.

 

[UltimateX86]

for the moment bad performance...

circlotron1.5A.asc

 

[analog_sa]

where can a found this article ?

The outputs are of course darlingtons.

 

[N101N]

This is the topology I have been proposing.
Next, biasing..

 

[UltimateX86]

for the moment the best results are here:

Jfet with differential feedback
jfet_circlotron_1.1A.asc
http://circlotron.audio/data/simulation/img/jfet_circlotron_1.1A.png
http://circlotron.audio/data/simulation/img/jfet_circlotron_fft_1.1A.png

bjt with differential feedback
bjt_circlotron_1.1A.asc
http://circlotron.audio/data/simulation/img/bjt_circlotron_1.1A.png
http://circlotron.audio/data/simulation/img/bjt_circlotron_fft_1.1A.png

 

[UltimateX86]

I just found this :

Lateral MOSFETs SuSy circlotron - Page 4 - Solid State - diyAudio.rs

Juma already made the same differential feedback with Jfet on a circlotron

Banat, you are also on this forum

 

[N101N]

What would you like to achieve?
The main purpose of the cascode is to establish a low drain-source voltage, while the voltage amplification is being carried out by a common-base amplifier. The thermal situation of Q3, Q4 can be worrisome. I would connect R19 and R25 to ground.
I am having difficulty suggesting appropriate bias currents at 35 V supply voltage.

 

[banat]

X

I like the results , THD spectrum is so incredible clean ,
few question ,
- what is the OPS- IQ ( bias current) ?
- it is possible to make sim of 19Khz+20Khz at full output power blast ?
-------------------------------------------------------------------
for long time I don`t visit that site anymore .

 

[UltimateX86]

n101n, I do not question the cascode,
I just did not have time to put it in the diagram with differential feedback

on the other hand I think we must use a 20v regulated power supply for input and vas

 

[UltimateX86]

banat,

the output bias is 400mA

ok, I'm doing a simulation at 28V 20Khz and 19 if you want , but why 19khz ?

at full power the THD increases a lot that's why I was asking for help

 

[analog_sa]

19+20 for IMD

 

[banat]

19+20 for IMD

Yes

 

[UltimateX86]

Intermodulation, OK I learn every day

I will make a ZIP with all the simulation

 

[UltimateX86]

first of all, should we use an input filter?

I already use the same filter as VSSA for bjt input, but not on the jfet

 

[padamiecki]

Hi UltimateX86!
the link from your footer does not work (403 Forbidden)

 

[UltimateX86]

yes I just created the hosting, for now I use it only to share circlotron images and simulatons, but there will be web a site later

 

[banat]

x

you can also check this variant for connecting VAS emiters on oposite B+(AC) side ( VAS full bootstrap) , as is basically done by Juma`s circlotron amp , but than maybe you will need to change the values of some resistors to get best amp performance .

edit , in this variant you can also try to connect R6/R14 on the ground line , not as is shown on the output lines .