Mosfet Power Amp Problem

[Telnet100]

Hello!
Recently, I constructed another 'Pro Series One' Power Amp,
(from Electronics Australia), an Amp that I have built many times
in the past. This time, however, I have a problem, and thought that someone might be able to help.
The Amp sets up perfectly, DC wise. Very stable. But as soon as You try to drive it with a signal, even small, it goes totally Unstable.
I am using the E.A. designed PC Boards, as before.

The ONLY difference is in the Power Mosfet types.
Previously I used the 2SJ49 and 2SK134, or 2SJ50 and 2SK135.
this time, however, I have used the 2SJ56 and 2SK176, as the
Older Types are getting harder to get. These Types have exactly
the Same Characteristics, according to downloaded spec. sheets,
and are also TO3 mounting.

To try to cure this fault, I have tried fiddling the Values of the Gate
Supply resistors, and tried also with the values of Capacitance in the Gate/Source circuits. Then -- Discovery!
If You remove EITHER Pair of Output Mosfets, and just run it on One Pair, ie, one P and one N channel, the Amp works PERFECTLY, so this looks like a Current/voltage Sharing problem.
The Effect is worst at 70v supply voltage, and gets less at lower supply voltages. I will have a go at posting the Circuit Diagram, if
that will help, as it is a well known design, and sounds very nice.
Does Anyone have an Idea here?? I'm looking for answers!!

 

[The Saint]

Hi Telnet100

The problem you are having is typical of this type of amplifier using
Lateral MOSFETs.
Try changing the Gate resistors on the 2SK176 devices to 680-820 Ohms and the Gate resistors on the 2SJ56 to 470 Ohms.

Also place a 22pf ceramic cap across the Gate and Drain of each 2SK176 device and remove the Gate Source Caps if you have any
across the 2SK176 devices.

These mods should help with the stability heaps.

 

[jackinnj]

the input capacitance of the 2SK176 is ~ 1,100 pF so with 680R the -f3 point is 212kHz -- this is about the same I use with the Renesas 2SK1058 with no stability problems.

yes, the gate-source caps "hot up" the slew rate and can lead to instability.

i use "belt and suspenders" impedance trap with 0.7uH with AND 10R and 100nF to ground.

 

[Telnet100]

Hey, Thanks Everyone!

One of the things I had done was to change the Gate feed Resistors already, to 1K1 and 1K6 for the 2SK176.
These are values that I found in another Commercial amp.
This did not help, at that time.
I had tried using 470Pf across Gate/source of them all, with the
object of seeing if this made it stable, but still, no luck.
I had even tried using Non-inductive .22 Ohm Source Resistors!But remove the .22Uf Capacitance between the Two 2SK176
Source connections, I have not tried that. Nice one, I will do that next!

Thanks Again, I will post the results.

 

[jackinnj]

I had tried using 470Pf across Gate/source of them all, with the
object of seeing if this made it stable, but still, no luck.

Take the gate-source capacitors out (we are guessing that this is a source-follower amplifier). Then see what happens.

Here's another trick -- put an RC network in parallel with the feedback resistor -- something like 10pF and 330R -- this will cause the gain to roll off at HF -- it won't help the slew rate, however.