PS ripple vs output impedance for PP AB1

[boywonder]

or.....CLC vs CLCLC in PP AB1

The basic question is this: Is adding a second LC (or RC) section to a power supply to reduce ripple worth the trade of increasing PS output impedance?

It seems that for AB1 operation, low PS output impedance is important due to changing current requirements, whereas class A being relatively steady state current draw would benefit from additional LC/RC sections at the expense of output impedance.

Is there a "sweet spot" point of diminishing returns for this? Are less sections "better" for AB compared to class A?

My question above assumes that different transformers can be specified to get whatever B+ is needed regardless of the number of LC/RC sections..

[SY]

Have your cake and eat it, too- use a regulator.

That said, I'd trade off somewhat increased ripple for lower supply impedance if I were on a desert island and couldn't get pass transistors. A decent push pull circuit does a pretty good job of ripple rejection, at least in the class A region, where it counts.

[janneman]

Originally Posted by boywonder

or.....CLC vs CLCLC in PP AB1

The basic question is this: Is adding a second LC (or RC) section to a power supply to reduce ripple worth the trade of increasing PS output impedance?

It seems that for AB1 operation, low PS output impedance is important due to changing current requirements, whereas class A being relatively steady state current draw would benefit from additional LC/RC sections at the expense of output impedance.

Is there a "sweet spot" point of diminishing returns for this? Are less sections "better" for AB compared to class A?

My question above assumes that different transformers can be specified to get whatever B+ is needed regardless of the number of LC/RC sections..

I think if you look at it in detail you will find that the current ripple from class A and class AB doesn't differ that much in amplitude. It does differ in shape, thus frequency content, so for class AB you'd want to decrease the higher harmonics on the lines.

But I agree with SY, a regulator does both.

jd

[cerrem]

Keep in mind that if these are pentodes you need not worry too much about plate voltage variation, ripple....Since you should be operating in the saturation region of the curves you should not see much issues....
The screen supply is another matter and this is what should be regulated and will make all the difference....
Many times to often I see chokes that are too small for the LC filter section for the screens... At idle things are OK....but at full power output the choke becomes void and the ripple on the screen is HUGE....
For example, 4 EL34 valves ate idle pulling roughly 4 mA per screen....
AT full power output this can be about 15mA to 25mA per valve and the choke is saturated or partially saturated and the L value drops off, making for some big 120Hz ripple of the screens....The screens in theory would cancel all this in P-P operation, but how many places match for screen gm in addition to input grid

Chris

[boywonder]

Originally Posted by cerrem

Keep in mind that if these are pentodes you need not worry too much about plate voltage variation, ripple....Since you should be operating in the saturation region of the curves you should not see much issues....
The screen supply is another matter and this is what should be regulated and will make all the difference....
Many times to often I see chokes that are too small for the LC filter section for the screens... At idle things are OK....but at full power output the choke becomes void and the ripple on the screen is HUGE....
For example, 4 EL34 valves ate idle pulling roughly 4 mA per screen....
AT full power output this can be about 15mA to 25mA per valve and the choke is saturated or partially saturated and the L value drops off, making for some big 120Hz ripple of the screens....The screens in theory would cancel all this in P-P operation, but how many places match for screen gm in addition to input grid

Chris

Great info, thanks. My particular application presently is 4x EL34 or KT88 PP Triode, so I'm assuming screen current is not an issue.