Yes, the OPT effectively becomes an autoformer with a higher turns ratio and a bit of extra primary inductance. In cathode biased output stages with CFB the DC can be kept out of the speaker by connecting the decoupling cap from the cathode to the secondary winding while the bias resistor connects to ground as usual.
Aha, interesting - I never had the curiosity to compare : this may explain why the EL85/6BN5 operates as best on the 5K UL connection than with the required 10K primary load, hence to the cathode included winding then :
T
I am thinking: a 5k to 16r OPT has a turn ratio of circa 17.7 to 1. If the 1 is added in series with the 17.7, one gets 18.7, which becomes the new turn ratio. With a load of 16 ohms, this would make for abut 5k6 primary. Is my calculation right, or am I missing some other part that influences the change in primary impedance?
best regards,
Erik
OK @Fuling : some news about the TT tt-otg5V3 output transformer...
I shifted to fixed bias my 10Y output stage.
B+ = 425V
Vp = 422V
Vg = -40V
Ik = 18mA at idle, 20mA at Pmax
NFB loop = 2dB
Power output = 1.5WRMS/8R load
primary 5K / secondary 4R, making the transformer behave as a 10K/8R.
The bandwidth of the circuit measured at 1WRMS is not better than :
35Hz - 29kHz @-3dB
72Hz - 17kHz @-1dB
Considering that the driver is a 12AX7 SRPP with a 4.7µF coupling cap to the 10Y grid, I find this quite a poor performance...
With my two LCR meters, I measured the primary inductance of the TT Xfo and found only 13H, which is quite low given the size of the laminations, probably due to the circa 1mm air gap !
As is, obviously this transformer doesn't meet Hi-Fi requirements - at least for me.
So what about your results with the PIEMME transformer you were waiting for ? Are the test better or worse than with the TT transformer ?
https://shop.piemme-elektra.it/Categories/93/Products/8104
T
I shifted to fixed bias my 10Y output stage.
B+ = 425V
Vp = 422V
Vg = -40V
Ik = 18mA at idle, 20mA at Pmax
NFB loop = 2dB
Power output = 1.5WRMS/8R load
primary 5K / secondary 4R, making the transformer behave as a 10K/8R.
The bandwidth of the circuit measured at 1WRMS is not better than :
35Hz - 29kHz @-3dB
72Hz - 17kHz @-1dB
Considering that the driver is a 12AX7 SRPP with a 4.7µF coupling cap to the 10Y grid, I find this quite a poor performance...
With my two LCR meters, I measured the primary inductance of the TT Xfo and found only 13H, which is quite low given the size of the laminations, probably due to the circa 1mm air gap !
As is, obviously this transformer doesn't meet Hi-Fi requirements - at least for me.
So what about your results with the PIEMME transformer you were waiting for ? Are the test better or worse than with the TT transformer ?
https://shop.piemme-elektra.it/Categories/93/Products/8104
T
Aah. 72Hz-17k -1dB is definitely nothing to write home about. Though, using this transformer as a 10k plate load for a high rp triode is probably somewhat a worst case scenario, you can probably put these transformers to a much better use together with some bigger tubes that can handle a lot more plate current. PL519 comes to mind, perhaps with Schade feedback to overcome the low Lprim.
I'm currently working on the 807 monos with the Piemme iron, and as I've already "borrowed" some parts from the breadboard I cannot provide any more test results until the monoblocks are done.
I don't know if it already has been mentioned but Lundahl has a small 16k SE transformer that can be ordered with customized airgaps, probably at least twice as expensive as the Piemmes but perhaps worth it considering the pricey output tubes?
I'm currently working on the 807 monos with the Piemme iron, and as I've already "borrowed" some parts from the breadboard I cannot provide any more test results until the monoblocks are done.
I don't know if it already has been mentioned but Lundahl has a small 16k SE transformer that can be ordered with customized airgaps, probably at least twice as expensive as the Piemmes but perhaps worth it considering the pricey output tubes?
Ah, OK - Don't hesitate to let me know how it turns when your amps will be operationals !
I am still making transformer tests with my 10Y...
T
Yes. This model from PIEMME is probably interesting for a 10Y : https://shop.piemme-elektra.it/Categories/93/Products/8377
T
T
Yes @Fuling, it was the case yesterday evening, but this morning the Piemme website is accessible as usual :
https://shop.piemme-elektra.it/Categories/93?page=1
T
Yes, now it's working again.
The transformer in your link is probably the best choice for your needs of the ones offered by Piemme. I picked their other 8k transformer because of its lower Dcr and 16R tap (for CFB). The lower Dcr, higher DC rating and different freq response indicates a lower Lprim which isn't a big issue for me as my front speakers only reach 50-55Hz and are supposed to be crossed over to a pair of subs at 70Hz or so with an active XO.
Both versions of Piemmes 8k transformers (there is also a third one with less interleaving, intended for guitar use) are probably a much better match for your 10Ys than the 2,5/5k OPT from TubeTown.
The transformer in your link is probably the best choice for your needs of the ones offered by Piemme. I picked their other 8k transformer because of its lower Dcr and 16R tap (for CFB). The lower Dcr, higher DC rating and different freq response indicates a lower Lprim which isn't a big issue for me as my front speakers only reach 50-55Hz and are supposed to be crossed over to a pair of subs at 70Hz or so with an active XO.
Both versions of Piemmes 8k transformers (there is also a third one with less interleaving, intended for guitar use) are probably a much better match for your 10Ys than the 2,5/5k OPT from TubeTown.
Any news?
I'm quite happy with my 807 amps with Piemme OPTs, in fact so happy that I'll probably buy another pair to use in an upcoming transmitter tube project.
Possibly 7-8 dB. The 808 has a relatively hi amplification factor similar to the 811a. DF should be 2 or better, so in line with most zero fbk triode amps. Distortion figures will be good as well if the driver is capable, better than average.
Yes @Fuling !
Here is below the comparison of the square wave response of different configurations :
A = 10Y with large Chino output transformer of excellent performance with KT120 (3K5).
B = 10Y with tt-otg5v3 output transformer (set at 10K).
C = 10Y with PIEMME output transformer (set at 14K).
D = 300B with STS HH25B output transformer (2K5).
E = 6V6S in triode connection with PIEMME output transformer (set at 5K).
The 10Y was in fixed bias operation, with 400VDC plate and 20mA current.
The coupling cap was 2.2µF and the grid resistor was 220K.
The power output was 1WRMS.
Tested, those 10Y met their datasheet specs plentifully at 250, 350 and 425VDC, no flaw.
Below, the square waves with the KT120 and the Chino excellent performance output transformer, at 4.5WRMS :
As you can see, the square waves result is quite puzzling : the 10Y is defeated by a mere 6V6 or KT120... Despite being a true triode, direct heated, with low plate current and reputedly good linearity (according to the datasheets curves).
With an Audio friend, who like me held the 10Y in high consideration, we made listening tests with the PIEMME transformer using the 6V6S in triode connection and the 10Y, with suitable load impedance each time.
NOTE : my Audio friend was the "Candide", as he did not knew anything about my waveform test done previously, he was just here to judge by his ears. His conclusion was that the 10Y sounded very pleasantly, more "rounded", but also a bit more "shy" in the extremes.
My friend and I were waiting for a better listening performance - not to say amazing - compared to a 6V6S triode-connected, and we came to something finally not really significative, I mean : no undisputable gap in tone turning at the advantage of the 10Y...
Given the hassle to heat a 10Y in a compact tube amp design, the project is not really worth the expense, and at best is postponed, if not discarded.
That said, these little PIEMME SE-EI168038CV-N transformers are far from being an average one, and could be used suitably with other tubes !
T
Very interesting results, thanks for sharing!
I have no personal experience of either type of tube* but as far as I know 10Y is notoriously difficult to get right due to its high Rp (5k?) while 6V6 is known to sound excellent connected as triodes.
* I did play around a bit with EL90/6AQ5 when I was young, ~6V6 in a 7 pin miniature bulb. That was over 20 years ago when I was new to tubes and had very little knowledge and no budget for even half-decent transformers so I can't say much about the sonic results of those experiments.
I have no personal experience of either type of tube* but as far as I know 10Y is notoriously difficult to get right due to its high Rp (5k?) while 6V6 is known to sound excellent connected as triodes.
* I did play around a bit with EL90/6AQ5 when I was young, ~6V6 in a 7 pin miniature bulb. That was over 20 years ago when I was new to tubes and had very little knowledge and no budget for even half-decent transformers so I can't say much about the sonic results of those experiments.
