small signal cfp?

[mrkramer]

Hi all

I see CFP/Sziklai Pair used in output stages all the time but much less elsewhere. Is there a reason for this? Seems like an appealingly simple circuit that would perform well as a single ended buffer.
Am I missing something?

Cheers
MrKramer

 

[djoffe]

It was quite common in equipment designed in the 70's and 80's. At some point, opamps became cheap enough that most people went the opamp route.


www.akitika.com

 

[mrkramer]

Hard to compete with cheap.
I wonder how they might compare in a real world circuit.

 

[bwaslo]

I've been using a cfp in the input circuit of a power amp, works extremely well, distortion of the stage is below what I can measure (and SPICE predicts about -100dBc for half volt inputs). But probably not the best input topology if you're using global feedback.

 

[Mark Johnson]

I see CFP/Sziklai Pair used in output stages all the time but much less elsewhere. Is there a reason for this?

Perhaps because the CFP is best used as an improved emitter follower.

Output stages want improved emitter followers. Previous stages don't want or need emitter followers nearly as often. However CFPs do find application in bias spreaders / VBE multipliers, where their quadratic current gain is useful in reducing the end-to-end impedance, deltaV/deltaI.

 

[Bigun]

I've tried the cfp in most places in an amp.

In the input stage it always sounded harsh, couldn't get along with it.

In the VAS I had mixed results - once it was OK, the other time it was not nice.

In the output stage I've used it single ended (headphone amp) and push pull (power amp) and liked it a lot, even at low power.

 

[Calvin]

Hi,

You might have seen this Thread about Preamp Buffers that deals to a great extent with CFP structures.
From about #52 on (and #87) the circuit basically reached its final design.
The CFP functions very well also in I/V converter stages as grounded base Input stage.
It keeps the input impedance very low and constant over frequency.
You may also search for user WrenchOne´s threads and posts (#51 e.g)
He used CFPs as gain stages also and his posts are generally a fine read.

jauu
Calvin

 

[DF96]

CFPs can be used as an improvement on a Darlington pair when you need a simple buffer, but they need to be properly designed as the improvement is only seen at medium current levels. Current too small and the CFP doesn't really work; current too large and it is no better than a Darlington at the same current - and probably less stable.

 

[mrkramer]

What would you consider small current?

 

[CBS240]

Degeneration (feedback) is needed with a current feedback pair. The overall Gm is quite high even for small signal devices in this arrangement. Degeneration should be relative to the particular transistor in question and relative to it's re; linearizing overall GBW and setting a limit to output goss. Playing with the amount of degeneration for a particular situation might yield you interesting results.

 

[Mark Johnson]

... linearizing overall GBW ...

That's a keeper, thanks! Putting it on the list of DIYA howlers, right now.

 

[DF96]

What would you consider small current?

That depends on the value of the resistor across the base-emitter of the second transistor. You need enough total current that the second transistor has turned on, but not so much that the second transistor base takes nearly all the current (as then you might as well use a Darlington).