How big?
Big enough to be able to stuff bifilar windings and/or trifilar windings on them . . .
. . . not to mention the secondary winding(s).
Big enough to be able to stuff bifilar windings and/or trifilar windings on them . . .
. . . not to mention the secondary winding(s).
Last edited:
I think that the McIntosh output transformers were both an art and a science.
And if the re-winders have all the ingredients, and the exact procedures, they may be able to duplicate the performance.
I would not suggest that anybody try to wind a McIntosh output transformer as a first try project.
And if the re-winders have all the ingredients, and the exact procedures, they may be able to duplicate the performance.
I would not suggest that anybody try to wind a McIntosh output transformer as a first try project.
Remember that all the amplifiers you mentioned feature the Unity Coupling design, which was a unique selling point for McIntosh. So it is fair that they won't let any detailled information about the transformers creep outside the factory walls.
Best regards!
Best regards!
Its not impossible, just requires lot of skill and calculation. Definitely nothing for beginners. For example I managed to did mine maybe even better than original ones... if i remember correctly 200kHz -3dB here it is:
https://www.diyaudio.com/community/threads/photo-gallery.71300/post-4669726
https://www.diyaudio.com/community/threads/photo-gallery.71300/post-4669726
around 4x5cm of 1,2T grain oriented silicon steel for 80W / 20Hz at current density of 3A/mm^2
around 4x3,2 or better 4x4cm core for 40-50W
around 4x3,2 or better 4x4cm core for 40-50W
tjerneld,
Were you planning on either winding your own, or contracting someone else to do it?
Or is your original question just a curiosity?
Building a Mc275 or Mc240 from scratch might take a lot of time and effort.
And time is money (Time = Money).
The good thing about building from scratch is you can say you did it yourself (or with a professional winder's help).
It might take less time and money to find and purchase an Mc275 or Mc240.
What loudspeakers do you have that can take those amplifiers 20 Hz full power out without distorting . . .
Like the 2nd and 3rd harmonic distortion products are more audible (40Hz and 60Hz) when the speaker is driven by 20Hz at full power?
Personally, I think I would rather have a pair of Mc30 amplifiers.
The later McIntosh amplifiers are way too much power for me, and too heavy too.
I no longer listen at high volumes; only 1 ear remains.
And I have had 2 hernia operations already.
I am almost getting to be practical at my age.
Were you planning on either winding your own, or contracting someone else to do it?
Or is your original question just a curiosity?
Building a Mc275 or Mc240 from scratch might take a lot of time and effort.
And time is money (Time = Money).
The good thing about building from scratch is you can say you did it yourself (or with a professional winder's help).
It might take less time and money to find and purchase an Mc275 or Mc240.
What loudspeakers do you have that can take those amplifiers 20 Hz full power out without distorting . . .
Like the 2nd and 3rd harmonic distortion products are more audible (40Hz and 60Hz) when the speaker is driven by 20Hz at full power?
Personally, I think I would rather have a pair of Mc30 amplifiers.
The later McIntosh amplifiers are way too much power for me, and too heavy too.
I no longer listen at high volumes; only 1 ear remains.
And I have had 2 hernia operations already.
I am almost getting to be practical at my age.
Last edited:
For 50% CFB, so easy to just order an off the shelf OT for Circlotron mode. 600 Ohm, 800 Ohm or 1000 Ohm P-P OT for 2400 Ohm, 3200 Ohm or 4000 Ohm equivalent Circlotron. Then use two split-bobbin industrial xfmrs for the floating B+ supplies. Only 100 pF common mode capacitance, easy.
Since there is NO DC balance issue for Circlotron mode, you could even series connect some LV power xfmrs to make the Circlotron OT. Or wind it yourself (low turns).
Since there is NO DC balance issue for Circlotron mode, you could even series connect some LV power xfmrs to make the Circlotron OT. Or wind it yourself (low turns).
Last edited:
tjerneld,
Bifilar windings of secondary turns is one thing, even though the secondary wire size is larger than the primary wire.
That is because the number of turns is the same, whether they are bifilar wound or not.
But winding 2 complete push primaries and 2 complete pull primaries, plates and cathodes, no matter the wire size ...
takes 2X as much space as only one push plate and one pull plate primary winding.
Those extra cathode primaries take up Extra space, that most output transformers do not have.
Bigger lams, or more losses in the DCR of smaller wire.
I have a pair of A470 Dynaco Stereo 70 output transformers. They have been used in different projects.
If they are bifilar, perhaps the actual output taps are 4, 9, and 16 Ohms, not 4, 8, and 16 Ohms.
Bifilar windings of secondary turns is one thing, even though the secondary wire size is larger than the primary wire.
That is because the number of turns is the same, whether they are bifilar wound or not.
But winding 2 complete push primaries and 2 complete pull primaries, plates and cathodes, no matter the wire size ...
takes 2X as much space as only one push plate and one pull plate primary winding.
Those extra cathode primaries take up Extra space, that most output transformers do not have.
Bigger lams, or more losses in the DCR of smaller wire.
I have a pair of A470 Dynaco Stereo 70 output transformers. They have been used in different projects.
If they are bifilar, perhaps the actual output taps are 4, 9, and 16 Ohms, not 4, 8, and 16 Ohms.
What are folks' current thoughts on the relative merits of Circlotron vs. N. Crowhurst? Might be interesting to consider separate cases of output valves with and without separate cathodes, and their power supplies' parasitic capacitances. Always a moving target, and I'm always a decade late and a dollar short.
All good fortunes, and thanks for any thoughts,
Chris
All good fortunes, and thanks for any thoughts,
Chris
The total turns number (and cross sectional area) is the same (50% at the anode and 50% at the cathode)
Total primary turns is the same for McIntosh as for N. Crowhurst (series'd), but is scaled 1:4 for Circlotron (paralleled) (a Bridge too Far). So maybe the interesting question is which of these gain-of-two output stages has the best match to the copper geometry of an output transformer. Core iron given, what's the best copper distribution? Is there ever a better empirical solution where multiple paralleled conductors (plus their insulators) is better than a single conductor with the same insulation? Don't know.
The gain-of-two output stage for vacuum valves has a magical place in society, just like the gain-of-one output stage has for the sad, sad semi-con output stages of this modern world. It just falls out naturally, like, well,
All goodness and light,
Chris
The gain-of-two output stage for vacuum valves has a magical place in society, just like the gain-of-one output stage has for the sad, sad semi-con output stages of this modern world. It just falls out naturally, like, well,
All goodness and light,
Chris
In theory, the Circlotron wins, as the transformer can be wound as an autoformer. Hence, less effort for insulation material, a bit smaller core etc.
Best regards!
Best regards!
tjernold,
Just to make it simple, do a comparison for an SE transformer.
Push pull is very similar, but we have to think in terms of 2x more separate winding(s) and se[arate connection(s).
Consider the inductance of a plate primary that is 2000 turns (on a given size of laminations).
SE output, plate drive only.
Now, wind a bifilar primary, use 1000 turns for the plate, and 1000 turns for the cathode, on the same given size of laminations). SE output, Unity coupled, plate and cathode drive.
You can use the same wire size in both cases, and the fill size is the same for 2000 turns; versus 1000 turns plus 1000 turns.
Are you willing to put up with only 1/4 of the inductance on the plate, and 1/4 of the inductance on the cathode, versus 2000 turns plate only.
1000 turns plate and 1000 turns cathode;
2000 turns plate;
inductance is according to the square of the turns.
1000 squared = 1M;
2000 squared = 4M
Same wire size, same total DCR, same wire length, same volume of turns occupancy.
There is no free lunch today.
Just my opinions.
Just to make it simple, do a comparison for an SE transformer.
Push pull is very similar, but we have to think in terms of 2x more separate winding(s) and se[arate connection(s).
Consider the inductance of a plate primary that is 2000 turns (on a given size of laminations).
SE output, plate drive only.
Now, wind a bifilar primary, use 1000 turns for the plate, and 1000 turns for the cathode, on the same given size of laminations). SE output, Unity coupled, plate and cathode drive.
You can use the same wire size in both cases, and the fill size is the same for 2000 turns; versus 1000 turns plus 1000 turns.
Are you willing to put up with only 1/4 of the inductance on the plate, and 1/4 of the inductance on the cathode, versus 2000 turns plate only.
1000 turns plate and 1000 turns cathode;
2000 turns plate;
inductance is according to the square of the turns.
1000 squared = 1M;
2000 squared = 4M
Same wire size, same total DCR, same wire length, same volume of turns occupancy.
There is no free lunch today.
Just my opinions.
