He's in read-only mode for a few more days. If he chooses to follow forum rules like the rest of us, he'll be back shortly. If not, a design like this can be discussed and detailed by others- both parafeed and mu followers are very well understood circuits.
Interestingly, of the analyses out there, I don't find any example of calculating the output impedance for the mu-follower or mu-stage. I haven't checked Valley and Wallman or Mo Jo yet. Lots of assertions and even measurements; it's not that difficult but it is surprising that the derivation of output impedance is missing from the commentary.
---------------------------------------------------
Here Broskie demonstrates the superiority of "SRPP+" over the mu-follower for driving a load, i.e. the totem pole anti-triode is superior to the mu-follower as a power amplifier.
SRPP+ All-in-One & Mu Followers
The anti-triode seems to be an even better choice than the mu-follower. It would use both devices in the totem pole to provide power to the output e.g. current swing of 2X the idle current as opposed to only 1X. You know, 2X the output power for your 2X B+ voltage investment. That sounds even more Ultimate... Or at least way more awesome.
Also having my doubts about the PSRR of a mu-follower based on a triode connected pentode for the top device. I would expect that connecting the screen grid to the dirty B+ rail might not result in the ideal PSRR. Pentode connection or MOSFET will be better.
Anyhoo pretty good noise and ripple isolation can also be obtained by using a simple MOSFET follower stabilizer on the B+ that only needs to drop 20 or 30 volts. Just sayin...
---------------------------------------------------
Here Broskie demonstrates the superiority of "SRPP+" over the mu-follower for driving a load, i.e. the totem pole anti-triode is superior to the mu-follower as a power amplifier.
SRPP+ All-in-One & Mu Followers
The anti-triode seems to be an even better choice than the mu-follower. It would use both devices in the totem pole to provide power to the output e.g. current swing of 2X the idle current as opposed to only 1X. You know, 2X the output power for your 2X B+ voltage investment. That sounds even more Ultimate... Or at least way more awesome.
Also having my doubts about the PSRR of a mu-follower based on a triode connected pentode for the top device. I would expect that connecting the screen grid to the dirty B+ rail might not result in the ideal PSRR. Pentode connection or MOSFET will be better.
Anyhoo pretty good noise and ripple isolation can also be obtained by using a simple MOSFET follower stabilizer on the B+ that only needs to drop 20 or 30 volts. Just sayin...
Last edited:
MJ shows how to derive output impedance (and while you're at it, power) using AC loadlines. Very nice how-to guide.
The anti-triode (or SRPP+) certainly makes more sense for an output stage than the Mu follower, since it can pull current in both directions, double the power out, and looks like a current source up to the B+ instead of a resistor (ie, the triode top), for PSRR.
If one is going to put a Mosfet up top, one might consider putting it on the secondary side instead, using a low voltage supply for it. (sort of a Circlotron split between the primary and secondary sides. Primary connects to B+ then instead of ground) That would cut the B+ needed in half. P Mosfet drive could come from a current sense resistor in the tube cathode, with a current sum balancing sense resistor in the Mosfet source side. This will put a little DC offset from wire resistance in the output though (at least for class A op).
By using two such circuits (with inverted signal inputs) with secondaries connected in series (and floating LV supplies for each Mosfet counter current booster), the small DC output offsets can cancel. (and the DC still balanced in the OTs, between primary and secondary currents)
Or, to get rid of the DC output offset in one stroke, one can go full hog with a P Mosfet and an N Mosfet current driver to the single OT secondary, with current sense from the tube cathode(s) to drive them. Several ways to do it. Essentially the Mosfets act to minimise (or track with gain) the current change thru the tube(s) (near constant Mu tube operation then) when the tube is driven. (a tiny OT can then be used as well, while still getting high power output) Of course this is really a hybrid amp, but an unusual one since the tube(s) still sets the output voltage directly. (What Broskie calls an impedance multiplier)
If one is going to put a Mosfet up top, one might consider putting it on the secondary side instead, using a low voltage supply for it. (sort of a Circlotron split between the primary and secondary sides. Primary connects to B+ then instead of ground) That would cut the B+ needed in half. P Mosfet drive could come from a current sense resistor in the tube cathode, with a current sum balancing sense resistor in the Mosfet source side. This will put a little DC offset from wire resistance in the output though (at least for class A op).
By using two such circuits (with inverted signal inputs) with secondaries connected in series (and floating LV supplies for each Mosfet counter current booster), the small DC output offsets can cancel. (and the DC still balanced in the OTs, between primary and secondary currents)
Or, to get rid of the DC output offset in one stroke, one can go full hog with a P Mosfet and an N Mosfet current driver to the single OT secondary, with current sense from the tube cathode(s) to drive them. Several ways to do it. Essentially the Mosfets act to minimise (or track with gain) the current change thru the tube(s) (near constant Mu tube operation then) when the tube is driven. (a tiny OT can then be used as well, while still getting high power output) Of course this is really a hybrid amp, but an unusual one since the tube(s) still sets the output voltage directly. (What Broskie calls an impedance multiplier)
Last edited:
"Do you mean anti-triode, i.e. with split resistor in the source of the FET? "
The anti-triode Zout should be approx. 1/2 the output Z of the lower tubes alone, since the upper section just doubles the effective class A current. (and a 2X impedance multiplier for the load to the tubes)
The anti-triode Zout should be approx. 1/2 the output Z of the lower tubes alone, since the upper section just doubles the effective class A current. (and a 2X impedance multiplier for the load to the tubes)
No, actual mu follower (tube-tube). Split resistor between CF and common-cathode, plate of common cathode cap coupled to CF grid.
Take a look at post #37 (link below), the yellow trace P-P tube amp FFT. Could the small sidebands around the main test signal peak be the power supply intermod that has been hypothesized to cause unpristine sound?
http://www.diyaudio.com/forums/tubes-valves/213391-most-powerful-se-diy-design-4.html#post3044838
http://www.diyaudio.com/forums/tubes-valves/213391-most-powerful-se-diy-design-4.html#post3044838
Last edited:
This link The Tube CAD Journal,SRPP Where to start?
describes the idea of driving from a couple of totempole outputs.
I do think the thread has a point about the PSRR varying as the signal varies, but that's been noted before and not just by me 🙂.
The idea of converting a traditional PP amp into a push-pull parafeed has some interest - especially from the PSRR view - has anyone here ever done it successfully with some decent pentodes? It seems to suit only triodes as far as I can see.
BTW the trouble with PSRR in a traditional PP circuit (using a common PSU rail) is that it's impossible to correct for it via feedback unless you throw a long and clumsy global loop over it. I.e. you get a PSU rail sag for both +ve and -Ve going peaks.
OK so he bashed the Aikido, but apparently this topology has been covered by TubeCad.
As soon as somebody disagrees on the perfect functioning of the mu-follower output stage, the reply is
"Incorrect. Build it and see". No theory, no explanation.
Well, not much to see here. I am unsubscribing from this thread.
As soon as somebody disagrees on the perfect functioning of the mu-follower output stage, the reply is
"Incorrect. Build it and see". No theory, no explanation.
Well, not much to see here. I am unsubscribing from this thread.
Forgive me, just now I'm over here, I think slowly, and you write a lot.
You mean your scientific brain conceived the amplifier for WWII?😕
Is this the reason that you do not put the component values?
Is it classified information?😕
Last edited:
SRPP Based Audio Power Stage
Consider the output stage on the attached schematic. It is inspired by my previous I/V stage for current-ouput DAC, and the by the circuit of the OP.
http://www.diyaudio.com/forums/digi...s-valve-output-stage-lundahl-transformer.html
n1 to n5 of T2 is on a common iron core. Biasing of the output tubes is not shown. I haven't checked if this is an original idea, or has someone came with a similar topology in the past. Neither have I built it yet. Any comments welcome.
Consider the output stage on the attached schematic. It is inspired by my previous I/V stage for current-ouput DAC, and the by the circuit of the OP.
http://www.diyaudio.com/forums/digi...s-valve-output-stage-lundahl-transformer.html
n1 to n5 of T2 is on a common iron core. Biasing of the output tubes is not shown. I haven't checked if this is an original idea, or has someone came with a similar topology in the past. Neither have I built it yet. Any comments welcome.
Attachments
Built two and both of them sucked. No more.
Yes, you do. I think this kind of stupid posting (trolling) should be moderated. I'm not talking about mine.
I think in the forum just three of us enthusiasts SRPP, including me.
The fact that you can not run properly, does not mean it sucked.
English is not my language, and you may have noticed.
It gives me a lot of work to understand what you say.
My intention is not to offend anyone, and if I did I apologize.
After reading page after page, I still have not seen anything of theory, any measurement, nor the component values of circuit proposed by the OP.
I subscribed to this wonderful forum for learning, in this case every time I'm more confused.
Did you build a power stage or a preamp?
True. Normally I don't give out personal opinions, too many variables, etc. BUT, the amazing thing is that despite using different parts and tubes both designs sucked in the same manner. When something sounds OK I try to improve it; when something sucks, it just sucks and that's the end of the story.
I don't care about your intentions. Did you have anything valuable to say? Who cares if you're confused. The OP can't say a thing because he's doing time. Just wait for him to get back, if he gets back. I doubt it.
if you have tried to build a SRPP poweramp, I suppose you have some experience and idea about what you are doing 😉
but if a SRPP preamp is difficult to optimise and make work properly, why would a poweramp be any easier
and hey, guys, no real reason to get too upset and personal about this crazy stuff, please
Can you show us what you did and explain how it sucked? This is the kind of assertion without detail that makes this thread suck.
I was expecting this, now I have to explain myself. 
I built something along these lines:
http://www.diyaudio.com/forums/attachment.php?s=&postid=84773&stamp=1037816258
And a headphone amp with these values, you may say the output impedance was too high and that may be the cause but I did try with a matching transformer too. Same disgusting sound.
As for the reason why they sucked I can't say. My body had a natural rejection against both designs.
I built something along these lines:http://www.diyaudio.com/forums/attachment.php?s=&postid=84773&stamp=1037816258
And a headphone amp with these values, you may say the output impedance was too high and that may be the cause but I did try with a matching transformer too. Same disgusting sound.
As for the reason why they sucked I can't say. My body had a natural rejection against both designs.
Attachments
I was expecting this, now I have to explain myself.
http://www.diyaudio.com/forums/attachment.php?s=&postid=84773&stamp=1037816258
And a headphone amp with these values, you may say the output impedance was too high and that may be the cause but I did try with a matching transformer too. Same disgusting sound.
As for the reason why they sucked I can't say. My body had a natural rejection against both designs.
Without wishing to offend, but there are many circuits with the SRPP topology, not just the ones you mentioned.
I do not know much about software, but I think that + B = 290V for ECC88 is too much.😉
- Status
- Not open for further replies.