Ever made a MOSFET CFP/Sziklai that works?

[richie00boy]

I built an amp with CFP output stage using IRF540N and IRF9540N devices with TIP29C and TIP30C drivers. The upper CFP bursts into oscillation after a short random time after power-up 9/10 times.

I thought I had cracked the design of this amp, just looking if anybody has successfully made a MOSFET CFP/Sziklai, I know Millwood (long time gone) was having problems as well.

[MikeB]

Yes, but a "long" time ago, also using irf540n/9540n.
I used bd139/140 as driver for the first one, and mje243/253 for the 2nd one. I had 10pF CB-caps at the drivers.
It sounded very powerful... But i had bias problems and didn't follow further.

Mike

[Workhorse]

Hi Richieboy,

I have designed complemantary mosfet amp with sizklai pair, but it requires some modification though....
1. 220 ohm resistors at the base of driver transistors,
2. R_collector=220ohms and R_emitter=47ohms
3. Rsource=0.47ohms 3Watts
4. 22pF at C-B of drivers
5. UF4004 diodes at C-B of Drivers...
6. Mosfets=IRF540/9540N
7. Drivers MJE340/350
8. Rgate=100ohms+IN4148
9. Gate to Source 7.5V 1W Zeners

Hope this Helps...

K a n w a r

[richie00boy]

Thanks Kanwar, I had already implemented a most of what you suggest even with very similar values. A few things to try though, like driver base resistors and driver C-B caps and diodes.

How do the driver C-B diodes help?

I see you have an antiparallel diode on the gate resistor to help with cross-conduction - a good mod I will add if it gets past the not oscillating stage.

[Workhorse]

Originally posted by richie00boy
Thanks Kanwar, I had already implemented a most of what you suggest even with very similar values. A few things to try though, like driver base resistors and driver C-B caps and diodes.

How do the driver C-B diodes help?

I see you have an antiparallel diode on the gate resistor to help with cross-conduction - a good mod I will add if it gets past the not oscillating stage.

Hi Richieboy,

The C-B Diodes helps very much by facilitating fast recovery from clipping behaviour when the VAS drives the Driver stage much harder....

I haven't experience any thermal biasing problems nor any of the oscillations.....so far!

regards
K a n w a r

[Fanuc]

Have never seen it done reliably with DMOS/hexfets/vertical based mosfets to be honest - or amp designers advocating it.

Seen it done with lateral power mosfets but I get the feeling some DMOS fets are faster in common source mode than lateral's in source follower mode.

An IRF240 has a GBP above 100mhz and transit times of 1ns or under etc.

I would stick with laterals with a high gain video transistor grade driver. Sanyo's 2SC2911(T) and it's comp. springs to mind. Very linear device.

[Workhorse]

Hi Richie,

Have you successfully mentored the CFP amp, please do post your results , So that every one could benefit from it....


regards,
K a n w a r

[AndrewT]

Hi,
is SKA by Ampguru a Sziklai connected FET?

[richie00boy]

Thanks for the reminder Kanwar. I made some mods to the amp but have not yet tested due to some other things taking priority. I will keep this thread updated

[forr]

John Linsley-Hood and Doug Self have proposed bipolar input - Hitachi mosfet output Sziklai pairs for output stages. I once built the JLH circuit, no stability problem at all.