"On second thoughts, you could just get away with the one pentode stage and cap couple the HF transformer at the anodes.
I will stop derailing the thread now. "
That's likely just what Wavebourn has in mind. I don't think it's too off topic!
I will stop derailing the thread now. "
That's likely just what Wavebourn has in mind. I don't think it's too off topic!
Sounds like a ferrite core would be great for the HF. It could be a class A amp so no worries about leakage inductance messing up the sound. Might take some experimentation to see what Bmax one can get away with. Pot cores easily can be gapped for DC if one is so inclined.
For mids most off the shelf power toroids are quite adequate. Parafeed, class A also.
And a nice hybrid circlotron switching OTL for the bass.
😎
For mids most off the shelf power toroids are quite adequate. Parafeed, class A also.
And a nice hybrid circlotron switching OTL for the bass.
😎
"And a nice hybrid circlotron switching OTL for the bass."
You mean with a switching DC power supply(s) I assume.
--------------
Thinking about the Berning OT scheme, where the power supply does the impedance conversion, I can't help but wonder if there is some strange variant that would work for the 60 Hz transformer B+ power supply typically already in tube amps. Suppose each P-P tube used B+ power from a separate half wave rectifier off the HV xfmr. Then add a second HV xfmr and wire it up the same way but so each tube gets the opposite 60 Hz phase half wave rectified. Now on the primary sides, the current draw from the 60 Hz line reflects the audio signal modulated on the 60 Hz. So wire the two xfmr primaries in series with the speaker coil such that each tube pulls current in one direction thru it (remember, each tube pulls current on a half cycle). And finally put a 60 Hz filter across the OT to short that out. Hmmm.... And then we use square wave 60 Hz power from one of those 60 Hz constant voltage xfmrs to power this mess. So becoming DC continuous (square wave sums being continuous time coverage), like in the Berning setup.
Darn. Looks like it needs de-modulation on the primary side, so instead of shorting out the 60 Hz, maybe have to put 60 Hz on the speaker's field coil.
Ohh, wait, if the 60 Hz constant voltage xfmr has two secondaries to power the two HV xfmrs separately, we can arrange a Circlotron like circuit thru the speaker coil without any need for de-modulation. The 60 Hz even nulls out in the speaker, no filter needed.
How do you like that one? I managed to get DC audio thru a 60 Hz power xfmr. And I think it is still obeying the Maxwell laws for flux in the xfmrs. Their is no DC voltage going thru the HV xfmrs, just current. The voltage is what has to behave, and it does. Um, I guess the speaker voltage drop shows up as subtracting from the primary voltages, so it won't work down to DC, still has to obey a minimum frequency for the xfmr flux.
You mean with a switching DC power supply(s) I assume.
--------------
Thinking about the Berning OT scheme, where the power supply does the impedance conversion, I can't help but wonder if there is some strange variant that would work for the 60 Hz transformer B+ power supply typically already in tube amps. Suppose each P-P tube used B+ power from a separate half wave rectifier off the HV xfmr. Then add a second HV xfmr and wire it up the same way but so each tube gets the opposite 60 Hz phase half wave rectified. Now on the primary sides, the current draw from the 60 Hz line reflects the audio signal modulated on the 60 Hz. So wire the two xfmr primaries in series with the speaker coil such that each tube pulls current in one direction thru it (remember, each tube pulls current on a half cycle). And finally put a 60 Hz filter across the OT to short that out. Hmmm.... And then we use square wave 60 Hz power from one of those 60 Hz constant voltage xfmrs to power this mess. So becoming DC continuous (square wave sums being continuous time coverage), like in the Berning setup.
Darn. Looks like it needs de-modulation on the primary side, so instead of shorting out the 60 Hz, maybe have to put 60 Hz on the speaker's field coil.
Ohh, wait, if the 60 Hz constant voltage xfmr has two secondaries to power the two HV xfmrs separately, we can arrange a Circlotron like circuit thru the speaker coil without any need for de-modulation. The 60 Hz even nulls out in the speaker, no filter needed.
How do you like that one? I managed to get DC audio thru a 60 Hz power xfmr. And I think it is still obeying the Maxwell laws for flux in the xfmrs. Their is no DC voltage going thru the HV xfmrs, just current. The voltage is what has to behave, and it does. Um, I guess the speaker voltage drop shows up as subtracting from the primary voltages, so it won't work down to DC, still has to obey a minimum frequency for the xfmr flux.
Last edited:
Possibly not relevant, but I remember some amp abusing a normal
sized output transformer for L+R, and smaller transformer for L-R.
2L=(L+R)+(L-R), 2R=(L+R)-(L-R)
L-R usually having has far less bass content than L+R.
sized output transformer for L+R, and smaller transformer for L-R.
2L=(L+R)+(L-R), 2R=(L+R)-(L-R)
L-R usually having has far less bass content than L+R.
By "switching amp" I only meant class AB but it could certainly use OTS 48 volt SMPS especially if it's only for bass.
As for the Berning scheme at 60 Hz, I don't know how the switching frequency could be so close to the audio being amplified. Oh I just noticed you have a square wave in there... yeah maybe...
Yeah, the Berning scheme should be able to work even with an audio carrier freq., if everything works "perfectly". (no residuals from imbalance, or diode charge storage etc.) I've just moved the scheme down to 60 Hz effectively.
-----------------
"Possibly not relevant, but I remember some amp abusing a normal
sized output transformer for L+R, and smaller transformer for L-R.
2L=(L+R)+(L-R), 2R=(L+R)-(L-R)"
My parents used to have a Westinghouse console that did that. Still got the 6BQ5s out of it.
There is also the Matrix amp scheme where the main channel (L+R) goes thru as normal P-P, and the difference channel goes thru as common mode (SE). Requires a gap in the center leg of the E lams. The P-P can go around the outside of the lams, and the SE common mode goes thru the center. (the two plate windings are on the outer parts of the E lam )
-----------------
"Possibly not relevant, but I remember some amp abusing a normal
sized output transformer for L+R, and smaller transformer for L-R.
2L=(L+R)+(L-R), 2R=(L+R)-(L-R)"
My parents used to have a Westinghouse console that did that. Still got the 6BQ5s out of it.
There is also the Matrix amp scheme where the main channel (L+R) goes thru as normal P-P, and the difference channel goes thru as common mode (SE). Requires a gap in the center leg of the E lams. The P-P can go around the outside of the lams, and the SE common mode goes thru the center. (the two plate windings are on the outer parts of the E lam )
Well, not too practical for HiFi obviously unless one can come up with a power notch filter for the 60 Hz residual that will inevitably leak thru.
Huh....., I was thinking that this would be another way around the Berning patent if the carrier freq were moved up to ultrasonic with ferrite xfmrs, but its not workable that way. The xfmrs have to handle the audio freq effectively, since the speaker V drop unbalances the carrier V thru the xfmrs (positive and negative half cycle voltages go out of balance). Strictly a LF scheme. Can sell it to people who have a notch filter in their hearing spectrum I guess.....
Huh....., I was thinking that this would be another way around the Berning patent if the carrier freq were moved up to ultrasonic with ferrite xfmrs, but its not workable that way. The xfmrs have to handle the audio freq effectively, since the speaker V drop unbalances the carrier V thru the xfmrs (positive and negative half cycle voltages go out of balance). Strictly a LF scheme. Can sell it to people who have a notch filter in their hearing spectrum I guess.....
Last edited:
Synchronus modulation and demodulation?
I mean drive the grids of two pentodes as-if parallel SE,
but under each cathode a switching MOSFET at >50KHz.
Each tode working an alternate half cycle into PP toroid.
One could use triodes, but Miller limits highest frequency.
The output would need demodulated by mosfets driven
from the same clock source. With Schottky diodes in
parallel to handle any gaps in the synchronous handoff.
Could use Schottkys alone (half wave bridge) on the
cheap.
There is a DC component to be canceled somehow...
Circlotron certainly an option. Or maybe some crazy
H-Bridge of four demodulated toroids?
I mean drive the grids of two pentodes as-if parallel SE,
but under each cathode a switching MOSFET at >50KHz.
Each tode working an alternate half cycle into PP toroid.
One could use triodes, but Miller limits highest frequency.
The output would need demodulated by mosfets driven
from the same clock source. With Schottky diodes in
parallel to handle any gaps in the synchronous handoff.
Could use Schottkys alone (half wave bridge) on the
cheap.
There is a DC component to be canceled somehow...
Circlotron certainly an option. Or maybe some crazy
H-Bridge of four demodulated toroids?
Last edited:
I still have some interest in Piezoelectric transformers.
Prices on Rosen types are getting reasonable, but don't
find many above 5W. Transoner rings have insane power
density, but samples I've seen $150 and up, out of my
price range for an experiment that probably won't work.
Prices on Rosen types are getting reasonable, but don't
find many above 5W. Transoner rings have insane power
density, but samples I've seen $150 and up, out of my
price range for an experiment that probably won't work.
- Status
- Not open for further replies.