Any voltage over 300 Vdc is a dangerous voltage,
and such voltages at significant power are found in every tube power amp out there.
So that critique is not really effective:
ALL tube amps are dangerous, and any false sense of safety or security in their regard
is itself a dangerous belief to encourage.
My design, like that of all tube amps,
requires absolute care in both the design, building and testing.
Tube amp circuits are not safe for the naive, careless, or incompetent, in any way.
I doubt there are very many 'purists' who insist on no caps.
My amp design has only two caps in the signal path.
That is surely minimal enough for most tube enthusiasts.
Ultimately its performance and enjoyment that counts,
not deferring endlessly to the 'purist' minority or the dangerously stupid.
I agree with you.
Why bridge medium power stereo amps?
Just build the amp you need and buy only one transformer.
My SRPP concept (see post #153) is push-pull, not bridged. Also it has no capacitors. Another advantage is the low turns ratio of the output transformer due to the signal coupled from the cathodes. I would appreciate if someone helped me in SPICE analysis of the circuit, as I have not much experience in it.
Last edited:
I understand what you're getting at in terms of keeping the design clean and with minimal phase shifts - I like that. And I am sure it really belongs in a separate discussion somewhere, but the bit in me that "watches where the electrons go" says that power supply capacitors are always in the signal path as well. (Not trying to wind anyone up, here, just observing...)
I agree with you.
Why bridge medium power stereo amps?
Just build the amp you need and buy only one transformer.
Can you decipher what you mean?
Incorrect.
The harmonic signature of a successful implementation of the anti triode is exactly the same as a good single ended triode amplifier.
Welcome back Nazaroo.
As Michael said, the anti-triode setup should give the same harmonic signature as a good SE amp. But preservation of the 2nd, or even harmonics, (versus the cancellation in a normal P-P setup) depends on one device having higher gm than the other for the anti-triode setup.
If both devices are the same (a nominal SRPP+ version), then the characteristics should be near identical to a normal P-P stage (cancellation of even harmonics). Both variants look like current sources to the B+ if a Mosfet or pentode are used up top.
------
Some of the Hammond OTs have separate secondaries for 4 Ohm versus 8 or 16 Ohm. Could put a CCS on the separate 4 Ohm secondary (for DC current cancellation in the OT), instead of using the Mu follower (or SRPP+) totem pole, to avoid the high B+.
As Michael said, the anti-triode setup should give the same harmonic signature as a good SE amp. But preservation of the 2nd, or even harmonics, (versus the cancellation in a normal P-P setup) depends on one device having higher gm than the other for the anti-triode setup.
If both devices are the same (a nominal SRPP+ version), then the characteristics should be near identical to a normal P-P stage (cancellation of even harmonics). Both variants look like current sources to the B+ if a Mosfet or pentode are used up top.
------
Some of the Hammond OTs have separate secondaries for 4 Ohm versus 8 or 16 Ohm. Could put a CCS on the separate 4 Ohm secondary (for DC current cancellation in the OT), instead of using the Mu follower (or SRPP+) totem pole, to avoid the high B+.
Parafeed vs PP:
As far as I can see the only difference in the mu follower over an SRPP/SRPP+ is that the top tube is driven much harder by the resultant AC signal - and so you get a better push-pull operation.
What I'm failing to see is why a single ended parafeed of say a 600V totempole in push-pull (say a mu-f) over both primaries is so different in terms of impedance etc than a proper push-pull in class A. I.e. a 300V driving a single primary vs 600V driving both of them - both circuits biased to the same idle current.
Surely the parafeed of 600V is effectively 300V idle +/- 300V extent, and the PP scheme is 300V idle with +/- 300V extent?
Am I being daft with the maths here (it's quite possible!)?
Note:
Also a note about the parafeed capacitor - if you use anode feedback you can put the feedback after the capacitor so it moves inside the/a feedback loop.
As far as I can see the only difference in the mu follower over an SRPP/SRPP+ is that the top tube is driven much harder by the resultant AC signal - and so you get a better push-pull operation.
What I'm failing to see is why a single ended parafeed of say a 600V totempole in push-pull (say a mu-f) over both primaries is so different in terms of impedance etc than a proper push-pull in class A. I.e. a 300V driving a single primary vs 600V driving both of them - both circuits biased to the same idle current.
Surely the parafeed of 600V is effectively 300V idle +/- 300V extent, and the PP scheme is 300V idle with +/- 300V extent?
Am I being daft with the maths here (it's quite possible!)?
Note:
Also a note about the parafeed capacitor - if you use anode feedback you can put the feedback after the capacitor so it moves inside the/a feedback loop.
"As far as I can see the only difference in the mu follower over an SRPP/SRPP+ is that the top tube is driven much harder by the resultant AC signal - and so you get a better push-pull operation."
The SRPP+ and anti-triode have current sensing resistors below the top cathode, so output is in current proportional to the drive signal from below. (and the sum of top current + bottom current is held constant.) The Mu F and SRPP drive the output directly off the top cathode, so are low Z voltage outputs from the top. The SRPP has a little resistance above the bottom tube plate so is relatively moderate Z from below, while the Mu F has a lot of resistance above the bottom plate, so is high Z from there. Result: Mu F has maximum current assymmetry, and SRPP+/anti-T has maximum current symmetry.
-------------
"What I'm failing to see is why a single ended parafeed of say a 600V totempole in push-pull (say a mu-f) over both primaries is so different in terms of impedance etc than a proper push-pull in class A. I.e. a 300V driving a single primary vs 600V driving both of them - both circuits biased to the same idle current."
For a CT'd OT, in class B P-P one tube side sees 1xZ. For class A P-P one tube side sees 2xZ (due to the tube conduction overlap). And for single tube parafeed driving the full primary, it sees 4xZ (ie, 2X turns squared).
The SRPP+ and anti-triode have current sensing resistors below the top cathode, so output is in current proportional to the drive signal from below. (and the sum of top current + bottom current is held constant.) The Mu F and SRPP drive the output directly off the top cathode, so are low Z voltage outputs from the top. The SRPP has a little resistance above the bottom tube plate so is relatively moderate Z from below, while the Mu F has a lot of resistance above the bottom plate, so is high Z from there. Result: Mu F has maximum current assymmetry, and SRPP+/anti-T has maximum current symmetry.
-------------
"What I'm failing to see is why a single ended parafeed of say a 600V totempole in push-pull (say a mu-f) over both primaries is so different in terms of impedance etc than a proper push-pull in class A. I.e. a 300V driving a single primary vs 600V driving both of them - both circuits biased to the same idle current."
For a CT'd OT, in class B P-P one tube side sees 1xZ. For class A P-P one tube side sees 2xZ (due to the tube conduction overlap). And for single tube parafeed driving the full primary, it sees 4xZ (ie, 2X turns squared).
Last edited:
Member
Joined 2009
Paid Member
Welcome back ! - I figure I haven't been 'pushing the envelope sufficiently' if I don't get into a sin bin once in awhile. Unfortunately, my wife doesn't agree
I don't have relevant experience with SS at high voltages to offer insight into reliability. My feeling is that the problem is not so much the high voltages per se providing you operate the parts well within their ratings. The problem arrises when using 'fake' parts or when unanticipated voltages arise - not so uncommon in a circuit that includes inductors. Wavebourn has relevant experience here, he uses SS on top of his tubes and has done so for many years to great effect. I particularly admire his Gyrator anode load. I'll dig out a link if necessary but I'm inviting him to jump in here....
One way or another, the danger level scales with voltage - this is common wisdom that I don't argue with.
By the way, my own current tube amp uses 280V B+, it's not much below 300V, but it is lower
p.s. I used to do Judo when I was a kid, the long name it was given to me as Ju-Jitsu. That's way I was 'complaining' about Ju Jutsu - looked like a typo ?
Sorry Bigun; I pass this time. I am working on a commercial project (hopefully) now.
Member
Joined 2009
Paid Member
Wow, historical thread! Brain storm, birth of bunch of ideas that now are common for hi-end design.
Thank you Bigun!
I think I would agree with you generally,
in that with most 'ordinary' amp designs,
which incorporate relatively low internal impedance Power Supplies,
the PS caps seem to contribute a huge amount to performance.
But this is I think in part due exactly to the low-impedance nature of a typical, simple 'constant voltage' supply.
In our circuit here, the 'top tubes' act as powerful feedback driven CCS regulators,
which work by supplying LARGE impedances in series with the PS/amp connection.
The result of this is not only the highest PSRR,
but extreme isolation from the primary stages of the Power Supply.
I'm not suggesting that bad caps will not have an effect,
or that good PS design can be avoided,
but only that with this type of design,
the PS variations due to cap quality in the PS are minimized.
So yes, with most designs, even PS caps are "in the signal path",
but with this unique design (double complimentary Mu-Followers in every stage),
the effect of PS caps is indeed satisfactorily very small.
One of the reasons I suggested building these circuits instead of 'armchair analysis',
is that computer simulations are often way off,
or lack key elements of actual functions in a real-world circuit.
Bob Pease, (Troubleshooting Analog Circuits) the man who invented Op Amps,
had the same important advice to all theorists
and especially grads who tended to rely on computer simulations:
Build it. Test it. Find out what it really is doing,
and then re-examine the whole thing.
Don't rely on calculations, or especially computer simulations.
is that computer simulations are often way off,
or lack key elements of actual functions in a real-world circuit.
Bob Pease, (Troubleshooting Analog Circuits) the man who invented Op Amps,
had the same important advice to all theorists
and especially grads who tended to rely on computer simulations:
An externally hosted image should be here but it was not working when we last tested it.
Build it. Test it. Find out what it really is doing,
and then re-examine the whole thing.
Don't rely on calculations, or especially computer simulations.
Attachments
Last edited:
Yep, true dat. Or rather, a truism.
But the only way to judge the final quality of Hi Fi amplification is to listen to it.
You can't do that on a computer.
- Status
- Not open for further replies.