My update lin topology

[Lumba Ogir]

Gareth,
that`s very true.

[GEirin]

Hi
My know is,
"The most linear stage for large voltage swings is the grounded base or common base amplifier. Unfortunately interfacing his stage with others is not easy because of biasing problems".
My idea is a buffer.
I need a explanation.
Thanks.

[GEirin]

The VAS is a single folded = a common base.
A common base has a low impedance input and has a very high impedance output.
Why no to use a buffered VAS?

[Lumba Ogir]

GEirin,
because in all, the disadvantages of adding a stage are greater than the advantages.

[AKSA]

Interfacing grounded base VAS is achievable. If VAS runs at 10mA, each leg of the LTP at 1mA, run input transistor collector to rail via a 470R resistor and fb transistor collector via a 47R resistor to rail. Then connect non-rail end of the 47R to the emitter of the VAS, with the base held at 1.14 volts above rail with a voltage reference to keep LTP in balance. Use 10mA CCS to supply the VAS, of course.

Cdom is tricky, but can be effected across collector of the VAS to the collector of the LTP input side, using a small resistive load to rail (suggest 470R).

I've tried this and it works well. HF performance is very good, as you would expect.

Hugh

[GEirin]

Hi Hugh

Thank you very much for the important informations.
Thanks for helping,
I will try those suggestions

regards
Guillermo

[GEirin]

Hi
I´ve done several changes.
Tipology: CFP LTP input + single folded VAS + CFP driver - EF output.
Yours opnions...........
Thanks
Guillermo

[AndrewT]

Hi the RC of both the input high pass filter and the NFB are near enough the same. 22k+10uF =220ms, 470r+470uF = 221ms.
The input filter should limit the bandwidth of the amplifier, not the NFB.
Try changing either the 470uF to 1000uF or preferably the 10uF electrolytic to 4uF or 6u8F metalised plastic film or foil.

The stupid auto log out lost my posting.
When are the MODs going to sort this stupid system?

[GEirin]

Hi AndrewT
Thank you very much for yours suggestions
Guillermo

[AndrewT]

I'll try re-posting the bits I lost yesterday.
I hope the system does not **** me about again.

The LTP passes ~600mV/470r ~1.3mA per side.
T1 & T2 pass ~600mV/560r ~1.1mA per side.
That leaves ~0.2mA passing T3 & T4 per side. This is very low.
Try adding a third diode to D2a,b or changing to a red LED 1.6V to 1.8V to increase the LTP current.

Does the 10pF act like a Miller comp cap across T5?
If so, I like the small value.
The 33pF to LTP and the 150r+1nF are the compensation. Are these sufficient?

What are D1a,b doing?

T9 & T10 force ~5.5mA through T7 & T8.
T11 & T12 force ~5.5mA through the 220r.
That leaves both T9 and T10 passing near zero bias current. How do they operate in small signal condition. Do they constantly turn on/off with signal amplitude? I think you should look at changing the 220r to somewhere between 33r and 100r.

T6 could be replaced with a To92 dual transistor multiplier.
The first To92 could be on the common sink fitted to T7+T8, the second could be on the main outputs heatsink.

The 10r T11 & T12 base resistors could be reduced, maybe somewhere between 1r0 and 3r3.
Add in base resistors to T9 & T10. Maybe 10r here.