Hi there all,
Call me old & rusty [true] but I've only just become aware of Sziklai pair transistor arrangements.
I am hoping that with this Thread I can read and learn more about them.
Any/all info. would be great.
Call me old & rusty [true] but I've only just become aware of Sziklai pair transistor arrangements.
I am hoping that with this Thread I can read and learn more about them.
Any/all info. would be great.
Rod Elliot is a fan of Sziklai and he is from your neck of the woods;
https://sound-au.com/articles/cmpd-vs-darl.htm
https://sound-au.com/articles/cmpd-vs-darl.htm
This topology is more often called CFP, Complementary Feedback Pair, which will help with searching. Basically the circuit has a high level of local feedback and be more linear (with less headroom) than an EF output stage. Stability issues can be worse though, due to the feedback.
The lower headroom makes it a sensible choice for lower voltage amplifiers.
Thermal feedback for bias-stabilization only needs to work on the driver transistors, not the output devices, which can simplify bias compensation.
The lower headroom makes it a sensible choice for lower voltage amplifiers.
Thermal feedback for bias-stabilization only needs to work on the driver transistors, not the output devices, which can simplify bias compensation.
Rod Elliot does have a rather romantic affair with them.
If you read with his bias, they are the greatest things in the world LOL
Most tend to disagree.
There is ways to parallel them for more power. With additions to adding degen resistors or slightly lower the feedback
So it is not a stability time bomb.
Otherwise agree is limited more to low voltage amps and a single pair. To stay safe
You can also add additional gain to the output stage.
Was a trick for driving them with +/_ 15 volt opamps but add gain to have more power or higher voltage rails to the power section
Say +/_ 30 volts or more to get past the limited swing of the opamp.
Again if you indulge in Rods romance with the things. First they are the greatest thing in the World!! ( their not)
Then if you add gain it is the END of the World ahhhh !!! dont do it!! ( It isnt, stability is a known issue regardless)
Vertical Mosfets are easy to drive, but need higher voltage to bias. So you loose a lot of voltage swing to the rail.
You can add small amount of gain, so more stable. But overcome the losses and go closer to the rail like a BJT normally does.
There is a lot more than that
If you read with his bias, they are the greatest things in the world LOL
Most tend to disagree.
There is ways to parallel them for more power. With additions to adding degen resistors or slightly lower the feedback
So it is not a stability time bomb.
Otherwise agree is limited more to low voltage amps and a single pair. To stay safe
You can also add additional gain to the output stage.
Was a trick for driving them with +/_ 15 volt opamps but add gain to have more power or higher voltage rails to the power section
Say +/_ 30 volts or more to get past the limited swing of the opamp.
Again if you indulge in Rods romance with the things. First they are the greatest thing in the World!! ( their not)
Then if you add gain it is the END of the World ahhhh !!! dont do it!! ( It isnt, stability is a known issue regardless)
Vertical Mosfets are easy to drive, but need higher voltage to bias. So you loose a lot of voltage swing to the rail.
You can add small amount of gain, so more stable. But overcome the losses and go closer to the rail like a BJT normally does.
There is a lot more than that
My one and only experience with attempting sziklai/CFP output pairs was not good, in terms of stability. The output stage oscillated no matter what I tried. That was before decent cheap/free SPICE programs were readily available. Nowadays it would be easier to simulate different approaches toward achieving stability.