I am waiting for quotations.
Could you indicate what cores you'd be interested in? (look at SG, SE or SU series of c-cores).
If you can tell me/us of any, I'm interested!
_-_-bear
Read my post carefully....
Hi Bud,
We firmly do not agree on this point.
I have no problems at all winding 1:1 interstage transformers with ruler flat curves up to several octaves above 20K without the resonances that your IT's have (Gary Pimm's measurements are revealing).
I have a clue how you wind them; I do it different, more like how Dave Slagle is winding them (Dave's measure pretty much better). I don't wind them bifilar.
For me winding a good SE to PP interstage is the biggest challenge; it is not easy to improve on Tango's NC14.
PLease note that my comments are on measurements, not sonic properties.
Z winding in output transformer
Does anyone have any experience with Z winding instead of U winding in a transformer design. On the Randall amp page at One Electron Niyam has chosen this style for the interstage windings to increase frequency response. Is the benefit of the Z make for a major increase in the frequency. With care and diligence it would just be a matter of mechanics to perform this type of wind. Of course the engineering aspects of the winding would need to be explored.
Any links or input.
Tad
Does anyone have any experience with Z winding instead of U winding in a transformer design. On the Randall amp page at One Electron Niyam has chosen this style for the interstage windings to increase frequency response. Is the benefit of the Z make for a major increase in the frequency. With care and diligence it would just be a matter of mechanics to perform this type of wind. Of course the engineering aspects of the winding would need to be explored.
Any links or input.
Tad
P,
U winding is finishing a row and starting back the way you came. Z winding is ending that row and starting a new row from the first row's starting point. You then connect the wires with jumpers usually arched over the complete wind. In Z winding all of the signal will be in the same direction as it progresses around the core. There is suppose to be less reverse EMF or possible signal cancellation. I know very little of this subject that is why I asked.
I would like to make the very best transformers for my amp that I can afford and there seems to be very little information posted on the web about audio output transformer winding configurations. The load the gauge of the wire and the size and type of core are all easy to find. The exact way you position the wire on the core is a black art and they seem to want it kept that way. The number of possible permutations for any given layout is enormous. Something that by trial and error could take many years to sort out.
There is a website selling transformer attenuators that states it takes a full week to wind one of the top of the line units. That must be some attenuator.
If you have time log on to the One Electron site and scroll to the Randall amplifier section. You will see a tube amp constructed by Nayim that is as good as it will ever get. Handmade caps, transformers, machined gold plated chassis. Just very nice. The interstage trannys are huge and Z wound.
Tad
U winding is finishing a row and starting back the way you came. Z winding is ending that row and starting a new row from the first row's starting point. You then connect the wires with jumpers usually arched over the complete wind. In Z winding all of the signal will be in the same direction as it progresses around the core. There is suppose to be less reverse EMF or possible signal cancellation. I know very little of this subject that is why I asked.
I would like to make the very best transformers for my amp that I can afford and there seems to be very little information posted on the web about audio output transformer winding configurations. The load the gauge of the wire and the size and type of core are all easy to find. The exact way you position the wire on the core is a black art and they seem to want it kept that way. The number of possible permutations for any given layout is enormous. Something that by trial and error could take many years to sort out.
There is a website selling transformer attenuators that states it takes a full week to wind one of the top of the line units. That must be some attenuator.
If you have time log on to the One Electron site and scroll to the Randall amplifier section. You will see a tube amp constructed by Nayim that is as good as it will ever get. Handmade caps, transformers, machined gold plated chassis. Just very nice. The interstage trannys are huge and Z wound.
Tad
Tad,
Ah, that's it.
Well I can tell you that I experimented with these different techniques a lot over the years.
IMHO the extra effort for Z winding does not pay; my advice is do it U.
Actually we are talking capacities, and it's only one of the (important) things.
Also, and that is also very important (but you don't hear much on this because it requires bigger cores and it seems to "appeal" less), keep DC resistances low!
Try to keep primary DCR under 10% of the tube Rp for a start, and secondary DCR preferably not over 0,2 ohm.
Differences in copper losses between mediocre and quality transformers are easily audible.
Thanks for the Randall link, I'll take a look.
Ah, that's it.
Well I can tell you that I experimented with these different techniques a lot over the years.
IMHO the extra effort for Z winding does not pay; my advice is do it U.
Actually we are talking capacities, and it's only one of the (important) things.
Also, and that is also very important (but you don't hear much on this because it requires bigger cores and it seems to "appeal" less), keep DC resistances low!
Try to keep primary DCR under 10% of the tube Rp for a start, and secondary DCR preferably not over 0,2 ohm.
Differences in copper losses between mediocre and quality transformers are easily audible.
Thanks for the Randall link, I'll take a look.
I am certain we could have a very interesting and friendly conversation about audio transformer design, only to find that we do not agree about anything!!!! One of the delights about audio transformers. I do regard your work highly and recommend your input splitters to folks who build for performance. Gary Pimm did say that you were very interested in the amp he brought to the ETF so long ago and seemed very interested in the OPT's, so you have heard my work also. There are so few of us left for this work.....
You are correct that I am much more interested in information transform, rather than measurement perfection and will put up with a peak at 35 kHz to get what I am looking for. I wind PP IT's in two wells, P S P, balance the DCR in the two primary sectors and X cross connect those primaries to obtain the minimum capacitive losses, with maximum coupling. Using more sections is not a good idea here.
To help solve the SE to PP IT's, try increasing wire diameter as you section up. Something like a P 1/3 #35, S 1/2 #35, P 1/3 #35, S 1/2 #34, P 1/3 #34. Designed for about a 75% fill this sort of scheme will balance the DCR among the primaries, providing near equal flux from each and balance the secondary DCR. It is also useful to narrow the secondary winding length with a narrow width (3mm) tape build up against the bobbin walls, or use of a shorter traverse on the secondary windings in paper section construction. Makes for a uniform and linear flux transform without incurring any leakage inductance.
As for Z winding, I use it all of the time in OPT's, but it is not useful for IT's unless other forms of normalization are not utilized.
Bud
I am not sure what you mean here Tyimo. On a torrpoid a Pye winding is a short segement repeated a few times, either as a secondary over a comntinuous primary or as completly seperate primary windings with sectioned secondaries. I am also having difficulty imagining a 90 degree winding, but I suspect you meant 180 out of physical phase (turning the winding around when winding interleaved sections). I have never bothered with this, though David Hafler was an ardent supporter of this technique.
Bud
Bud,
Look at the "education" section of Lenard Audio (output trannys).
There you find a description and picture.
IMHO not for audio....
I agree, unless I was working with amorphous material and a single loop section of cut core. The real problem with this winding format is that you are throwing away 1/2 of the information, well, there is some capacitive coupling between sectors but it is weak. I doubt that winding primary and secondary in each sector would allow enough usefull coupling surface to overcome the distributed capacitence in the windings, these are antenna events after all.
For core material that actually passes energy efficiently across the full frequency bandwidth, as amorphous cut cores will, you must eliminate the capacitive coupling to obtain flat frequency response. This strips the fine detail from note and transient specific, internal gradient strucutres, and this is going in the exact oposite direction to my interests.
I do have great respect for Per's work. Their OPT's have superb leading edge, tonal and color information. I just like more detail.
Bud
Just one of the many benefits gained from using C Core!
For those of you who utilize E/I core you must use either the winding layout from the 30's RCA PA amplifier, and all of the early and famous Fender and Marshall guitar amplifiers, or you must use the following scheme.
Primary A1 #28
Secondary 4 ohm
Primary A2 #28
econdary 8 & 16 ohm
Primary B1 #28
Secondary 4 ohm
Primary B2 #27
The primaries are split (A1 / A2) at the CT, so, series connected, and each set has the full primary turns. Set A and set B are parallel connected outside of the winding. Doing this, along with the 75% to 85% build rule, will always provide overall DCR balance within 1 ohm, usually withion 0.05 ohm. The RCA PA winding scheme is the first scheme shown in the RDH4 section on OPT winding, and it is criminal not to have this reference, and will provide exact DCR balance, but is really not acceptable for modern audio reproduction.
Bud
BudP,
Have you experimented with disc winding and did it have any benefits. The overall scheme seems to fit fairly well with high voltage transformers I just wonder about high frequency output trannys.
It is input from the knowledgeable such as yourself that makes this forum what it is.
Tad.
Have you experimented with disc winding and did it have any benefits. The overall scheme seems to fit fairly well with high voltage transformers I just wonder about high frequency output trannys.
It is input from the knowledgeable such as yourself that makes this forum what it is.
Tad.
Tad,
Not for audio, no. I am going in exactly the opposite direction. Wide winding width, low dielectric constant dielectric barrier across the antenna event and a shallow winding depth for the primaries.
In my obscenely expensive level three OPT's, I have Z wound, single layer secondaries, paired up in two layers per sector with each layer having the total # of turns needed for that ohmage. The primaries are generally exactly as deep as the secondaries and there are 16 windings with three complete primary windings paralleled outside of the coil. The high current secondaries are all paired together with the higher voltage secondaries adding on in series. This provides the least possible leakage inductance and the least possible distributed capacitance from the all important coil. Also takes forever to wind.
The spiral wound, narrow width coils do have a minimum of capacitive coupling, along with a fair amount of distributed capacitance and very high inductance. This format is perfect for high voltage.
Pieter,
Gary's Tabor amplifier, with solid state devices and IT and OPT and CCS input splitter, is one of the two best sounding amplifiers I have experienced. The man is just a superb designer and his personal sound system is staggeringly good. I am delighted to live within driving distance (3 hours) of Gary and Maggie (who might just be even smarter than he) and try to visit in person once or twice a year.
Bud
Not for audio, no. I am going in exactly the opposite direction. Wide winding width, low dielectric constant dielectric barrier across the antenna event and a shallow winding depth for the primaries.
In my obscenely expensive level three OPT's, I have Z wound, single layer secondaries, paired up in two layers per sector with each layer having the total # of turns needed for that ohmage. The primaries are generally exactly as deep as the secondaries and there are 16 windings with three complete primary windings paralleled outside of the coil. The high current secondaries are all paired together with the higher voltage secondaries adding on in series. This provides the least possible leakage inductance and the least possible distributed capacitance from the all important coil. Also takes forever to wind.
The spiral wound, narrow width coils do have a minimum of capacitive coupling, along with a fair amount of distributed capacitance and very high inductance. This format is perfect for high voltage.
Pieter,
Gary's Tabor amplifier, with solid state devices and IT and OPT and CCS input splitter, is one of the two best sounding amplifiers I have experienced. The man is just a superb designer and his personal sound system is staggeringly good. I am delighted to live within driving distance (3 hours) of Gary and Maggie (who might just be even smarter than he) and try to visit in person once or twice a year.
Bud
Maggie is one of those women you occasionally meet, who have a vast mentation. SY recently married himself one and I was fortunate to find one who would agree to be my life partner, 40+ years ago.
Do you mean the long E core that was used in telephone line isolation transformers or do you mean EE laminations? I currently make an audio transformer for a government agency sponsored communication network for first responders. A tiny 0.1875 center leg M6 EE core thing. Actually worse than an E/I core. At least the I has the core grain going in the correct direction, but it is a major pain to put together in that small a size and these folks don't care about audio quality. They subscribe to the same audio parameters found in NASA's voice transmission. Total distortion so you cannot ignore it, is the goal as near as I can tell.
Do you mean the long E core that was used in telephone line isolation transformers or do you mean EE laminations? I currently make an audio transformer for a government agency sponsored communication network for first responders. A tiny 0.1875 center leg M6 EE core thing. Actually worse than an E/I core. At least the I has the core grain going in the correct direction, but it is a major pain to put together in that small a size and these folks don't care about audio quality. They subscribe to the same audio parameters found in NASA's voice transmission. Total distortion so you cannot ignore it, is the goal as near as I can tell.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.