I have read that when people talk about good/great tube amps, among all the other things important, it seems one of the things at the top of the list is the quality of the “iron.” And of that “iron” (transformers), the output transformers seem to impart the most difference in sound-quality. I would like to have a discussion on what makes a good/great OT. I know many of the top names: Hashimoto, Tango, Tamura, Lundahl, old Peerless, et.al., but this post is about what is inside those that make them so good/great. I am definitely not an expert in this field, so I am open to your information, and also your corrections. Hopefully, I can bring in a few Tube afficionados with experiential knowledge, that have owned different types of OT’s.
I’ll start with what I have:
1. There is something about the nickel content of the iron core. Interesting, since nickel is undesirable on speaker binding post, but of course, this is a different application.
2, The quality of the copper in the windings, (or the silver winding in Steve McCormack’s half a million dollar amps.)
3. The particular ways of winding.
Ok, that’s all I know. Feel free to elaborate on any of these, and what I might have left out. Sound quality observations are appreciated, and so are part and SQ comparisons between brands. Thanks!
I’ll start with what I have:
1. There is something about the nickel content of the iron core. Interesting, since nickel is undesirable on speaker binding post, but of course, this is a different application.
2, The quality of the copper in the windings, (or the silver winding in Steve McCormack’s half a million dollar amps.)
3. The particular ways of winding.
Ok, that’s all I know. Feel free to elaborate on any of these, and what I might have left out. Sound quality observations are appreciated, and so are part and SQ comparisons between brands. Thanks!
Although it will be taken as snake oil, as a transformer craftsman and listening tests since 2015, my top criteria goes to building materials that are featured inside a transformer.
During this time periods, I studied the influence on materials on all types of transformer and chokes, but also on all audio devices building, including the chassis.
I'm a sworn enemy to the dogma of "absolute best sound", because it is a varying subjective criteria among every listener. Many audio building enthusiasts will try to convince of best and worst for the sake of marketing. So when it comes to transformers sound, I built my own understanding of favorite sound, which leans towards the dark, mellow side. It is similar to Tamura's signature, but more towards the bottom end with a bigger extent of energy and punch.
I achieve this type of sound by adding different materials to the transformer assembly. To achieve mellowness and fighting potential harshness, I impregnate the coils with a sticky chosen ceresin. If I need a transformer to sound more "bright", I don't impregnate the coils.
I also add chosen materials to the core and coils assembly, depending of where I need the signature redirected. Such materials can be gold, silver, bismuth alloys, tonewoods, bitumen.
As for wire sound goes, I find more important the hardness of the wire, compared to purity mumbo. The wire can be annealed for a more mellow, darker tone, however it is hard to impossible to do without damaging the enamel coating. However, some winding wires are softer than others.
The dielectric is also part of the transformer assembly and important for the signature. Paper is usually my top preference, where one can change the sound depending of the impregnant, or not. Mylar can be harsh is a no go, unless combined with a mellower sounding core, such as nickel containing or very dampened assembly.
If you do follow this philosophy, you can already tweak a cheap set of transformer by adding different materials and/or dampening them.
During this time periods, I studied the influence on materials on all types of transformer and chokes, but also on all audio devices building, including the chassis.
I'm a sworn enemy to the dogma of "absolute best sound", because it is a varying subjective criteria among every listener. Many audio building enthusiasts will try to convince of best and worst for the sake of marketing. So when it comes to transformers sound, I built my own understanding of favorite sound, which leans towards the dark, mellow side. It is similar to Tamura's signature, but more towards the bottom end with a bigger extent of energy and punch.
I achieve this type of sound by adding different materials to the transformer assembly. To achieve mellowness and fighting potential harshness, I impregnate the coils with a sticky chosen ceresin. If I need a transformer to sound more "bright", I don't impregnate the coils.
I also add chosen materials to the core and coils assembly, depending of where I need the signature redirected. Such materials can be gold, silver, bismuth alloys, tonewoods, bitumen.
As for wire sound goes, I find more important the hardness of the wire, compared to purity mumbo. The wire can be annealed for a more mellow, darker tone, however it is hard to impossible to do without damaging the enamel coating. However, some winding wires are softer than others.
The dielectric is also part of the transformer assembly and important for the signature. Paper is usually my top preference, where one can change the sound depending of the impregnant, or not. Mylar can be harsh is a no go, unless combined with a mellower sounding core, such as nickel containing or very dampened assembly.
If you do follow this philosophy, you can already tweak a cheap set of transformer by adding different materials and/or dampening them.
Characteristics of a good transformer.
Flat freq response 1hz to 100khz
zero insertion loss
light weight.............
Keep on dreaming....
Flat freq response 1hz to 100khz
zero insertion loss
light weight.............
Keep on dreaming....
In the days of yore it was mostly a question winding technique, since no one used exotic materials worth mentioning. But the materials they did use were well-made and carefully employed.
Among the many useful materials on transformer design, there's a lot of good information at Tim Robbin's Williamson resource page, including detailed measurements of an original Williamson output transformer and an article by a fellow who wounds his own replacement.
https://dalmura.com.au/projects/Williamson.html
Speaking of exotic materials, some years ago I built a pair of Williamson monoblocks for a friend, using Heyboer's magnificent copy of the Peerless S-265-Q, which I consider to be one of the great steals in DIY audio for less than $500 a pair. Recently he was interested in seriously upgrading the amps, and ordered a pair of custom Japanese transformers from a company that specializes in fancy single-ended transformers. The price was 10 times that of the Heyboers. I spent some time corresponding with the distributor about the schematic, and the amount of feedback (20dB) employed in the design. The potted transformers finally arrived but due to careless packing were hopelessly dinged up and flagged for a refund. I clipped them into the amps anyway and they were completely unsuitable, exhibiting a 10dB peak at 300kHz which rendered the amps completely unstable and which no amount of compensation could tame. The open-loop bandwidth was very good, however. ;-) I just mention this to point out that classic transformer design, especially one that will be used in a feedback amplifier, is an art that no amount of exotic materials can make up for.
Among the many useful materials on transformer design, there's a lot of good information at Tim Robbin's Williamson resource page, including detailed measurements of an original Williamson output transformer and an article by a fellow who wounds his own replacement.
https://dalmura.com.au/projects/Williamson.html
Speaking of exotic materials, some years ago I built a pair of Williamson monoblocks for a friend, using Heyboer's magnificent copy of the Peerless S-265-Q, which I consider to be one of the great steals in DIY audio for less than $500 a pair. Recently he was interested in seriously upgrading the amps, and ordered a pair of custom Japanese transformers from a company that specializes in fancy single-ended transformers. The price was 10 times that of the Heyboers. I spent some time corresponding with the distributor about the schematic, and the amount of feedback (20dB) employed in the design. The potted transformers finally arrived but due to careless packing were hopelessly dinged up and flagged for a refund. I clipped them into the amps anyway and they were completely unsuitable, exhibiting a 10dB peak at 300kHz which rendered the amps completely unstable and which no amount of compensation could tame. The open-loop bandwidth was very good, however. ;-) I just mention this to point out that classic transformer design, especially one that will be used in a feedback amplifier, is an art that no amount of exotic materials can make up for.
mike567 and prairiboy,
I hope you are both joking.
Mike,
Have you never heard a good vacuum tube amplifier that uses an output transformer?
prairieboy,
Hobbies are expen$ive.
Perhaps either OTL amplifiers, or solid state amplifiers are in your futures. But good ones are also expen$ive, and not perfect for some loudspeakers.
I hope you are both joking.
Mike,
Have you never heard a good vacuum tube amplifier that uses an output transformer?
prairieboy,
Hobbies are expen$ive.
Perhaps either OTL amplifiers, or solid state amplifiers are in your futures. But good ones are also expen$ive, and not perfect for some loudspeakers.
50AE,
I think you wind some of your own output transformers, Right?
Good going!
Now where did I hide my ARRL Radio Amateurs Handbook?
It is a great reference book, with a chart of dielectric constants.
Paper versus Teflon . . .
I found my ARRL handbook.
Dielectric Constant: Paper 3.0, Teflon 2.1
That means more Capacitance between layers of winding, if you use Paper
But for a 0.001 inch thick sheet, the puncture voltage of Paper is only 200V, and for Teflon 1000v, to 2000V.
For high voltage, paper has to be thicker.
So, thicker paper may have less capacitance than thin teflon that can withstand high voltage.
I could not find a chart for the dissipation loss of Paper versus Teflon . . .
Which ever material has more loss would de-Q the transformers resonances of capacitance and inductance.
Tradeoffs, which apparently can be used as an advantage to get what you want.
I think you wind some of your own output transformers, Right?
Good going!
Now where did I hide my ARRL Radio Amateurs Handbook?
It is a great reference book, with a chart of dielectric constants.
Paper versus Teflon . . .
I found my ARRL handbook.
Dielectric Constant: Paper 3.0, Teflon 2.1
That means more Capacitance between layers of winding, if you use Paper
But for a 0.001 inch thick sheet, the puncture voltage of Paper is only 200V, and for Teflon 1000v, to 2000V.
For high voltage, paper has to be thicker.
So, thicker paper may have less capacitance than thin teflon that can withstand high voltage.
I could not find a chart for the dissipation loss of Paper versus Teflon . . .
Which ever material has more loss would de-Q the transformers resonances of capacitance and inductance.
Tradeoffs, which apparently can be used as an advantage to get what you want.
Last edited:
6A3 yes I have listened to macs and a few others. Unfortunately these old ears and tinnitis can't really appreciate the high end stuff.
With that said i will build what i can still appreciate. Yes tube audio has gotten really expensive since covid and the great bailout.
With that said i will build what i can still appreciate. Yes tube audio has gotten really expensive since covid and the great bailout.
6A3,
I was aware of dielectric properties since 2015-16 and back then, I started measuring them on my own, instead of relying blindly on textbooks, due to variations between the manufacture of different materials. Using two 10x10cm metal plates, I measured different types of paper, Nomex 410, Nomex-Mylar-Nomex composite, Mylar alone, polycarbonate, Mylar-paper composite. I also developed my own air dielectrics. I also made two transformers with a 5% difference with capacitance, aiming for Mylar vs paper close epsilon.
I'm not persuaded that transformer sound has to do with the electrical parameters you mentioned, although, having a low capacitance as possible is never a bad thing.
In my work, despite tailoring the sound based on the choice of materials only, I still try to aim for the best frequency response extension, mostly for the sake of versatility with different tubes and conditions, hence sometimes, I'm even using air dielectrics in my biggest transformers.
I hope this might add to the discussion.
I too are interested in what to look/select for in a output transformer. I've been exploring the possible replacement a set of existing transformers in a SE 300b amplifier. I've read multiple threads concerning this subject. Lots of opinions. Maybe those with experience in multiple applications with different materials can also expanded upon the following issues also.
Using 7w, vs 20w, vs 30w transformers with a amp only realistically capable of 6 or less watts.
Transformers with single vs multiple output windings.
I too are interested in what to look/select for in a output transformer. I've been exploring the possible replacement a set of existing transformers in a SE 300b amplifier. I've read multiple threads concerning this subject. Lots of opinions. Maybe those with experience in multiple applications with different materials can also expanded upon the following issues also.
Using 7w, vs 20w, vs 30w transformers with a amp only realistically capable of 6 or less watts.
Transformers with single vs multiple output windings.
My knowledge of winding OPTs is on a caveman level. So from previous experiences with lower power SE OPTs and from reading posts here I specified the following for a new pair...
Open to correction for any of the above....
- small size
- smaller gap, just appropriate for the current, in my case 60mA
- higher primary impedance, so for 1.2K Rp I wanted 5K not 3.5K
- HiB core not amorphous which I find too lean, however detailed
- double C core.... not sure how much this matters though
Open to correction for any of the above....
It has been done, I've also experimented with this. One usually gets a more controlled tone with a bigger transformer of the same kind. My logical explanation is the higher weight dampening the chassis and surroundings. I've done similar test with smaller transformers by adding weight, using ingots of different metals - lead, bismuth, copper, brass scrap, etc. Subjectively, I got quite similar results.
50AE,
Correct. Capacitance is only one of many factors that affect output transformers.
You are one who does his research, pays attention to details, etc.
Trust but Verify!
Transformer design details:
number of windings; interleaved or not; etc.
lamination size and materials
winding techniques, including reversing winding direction, etc.
wire size
Insulation
shellac, varnish, nothing, etc.
Es and Is, double C core, Toroid, etc.
I am sure there are more factors.
And . . .
The tubes, drive impedance, and circuit that drives the primary
The loudspeaker (and crossover if present), that loads the secondary
And . . .
Negative feedback that:
helps reduce the effect of some problems
Makes saturation worse
And . . .
That can cover over some micro details of the music that is coming from the tubes that drive it
That sometimes can make different quality levels of output transformers sound very much alike.
Negative feedback can sometimes cover over a multitude of "sins", and sometimes can make things worth, and sometimes can do Both.
Just my $.03
Correct. Capacitance is only one of many factors that affect output transformers.
You are one who does his research, pays attention to details, etc.
Trust but Verify!
Transformer design details:
number of windings; interleaved or not; etc.
lamination size and materials
winding techniques, including reversing winding direction, etc.
wire size
Insulation
shellac, varnish, nothing, etc.
Es and Is, double C core, Toroid, etc.
I am sure there are more factors.
And . . .
The tubes, drive impedance, and circuit that drives the primary
The loudspeaker (and crossover if present), that loads the secondary
And . . .
Negative feedback that:
helps reduce the effect of some problems
Makes saturation worse
And . . .
That can cover over some micro details of the music that is coming from the tubes that drive it
That sometimes can make different quality levels of output transformers sound very much alike.
Negative feedback can sometimes cover over a multitude of "sins", and sometimes can make things worth, and sometimes can do Both.
Just my $.03
optimationman,
Good questions!
Good Quest!
Can you attach the complete and accurate schematic of your 300B amplifier, please.
Things like output stage voltages, and output stage current, negative feedback, or No negative feedback.
Commercial amplifier, or diy amplifier.
What are the detials of the output transformers you are using now?
Details of the loudspeaker model or diy loudspeaker that you are using, including nominal impedance, DCR measured at the input terminals of the loudspeaker system, and minimum impedance, ported, closed box, open baffle, transmission line box; crossover, etc.
This has to be viewed and considered as a Systems optomisation problem.
Making things work better, requires us to know the target details you, or others are trying to hit on the Bull's Eye.
types of music, signal source(s), etc.
Good questions!
Good Quest!
Can you attach the complete and accurate schematic of your 300B amplifier, please.
Things like output stage voltages, and output stage current, negative feedback, or No negative feedback.
Commercial amplifier, or diy amplifier.
What are the detials of the output transformers you are using now?
Details of the loudspeaker model or diy loudspeaker that you are using, including nominal impedance, DCR measured at the input terminals of the loudspeaker system, and minimum impedance, ported, closed box, open baffle, transmission line box; crossover, etc.
This has to be viewed and considered as a Systems optomisation problem.
Making things work better, requires us to know the target details you, or others are trying to hit on the Bull's Eye.
types of music, signal source(s), etc.
6A3,
My top list goes as the following, that is, most details I'd pay attention to, that are most important for the sound signature.
1. Core type
2. Lamination adhesive type
3. Wire metal
4. Wire hardness
5. Core mass to wire mass ratio.
6. Total weight of the transformer
7. Dielectric acoustic tonality.
8. Coil impregnants properties
9. Potting compounds properties
10. Mounting elements properties
11. Lead out wires/output terminals properties.
12. Airgap material type.
My top list goes as the following, that is, most details I'd pay attention to, that are most important for the sound signature.
1. Core type
2. Lamination adhesive type
3. Wire metal
4. Wire hardness
5. Core mass to wire mass ratio.
6. Total weight of the transformer
7. Dielectric acoustic tonality.
8. Coil impregnants properties
9. Potting compounds properties
10. Mounting elements properties
11. Lead out wires/output terminals properties.
12. Airgap material type.
50AE,
Over 2 decades ago, I did not know much about output transformers: materials, winding techniques, etc.
I wanted to make a stereo single ended amplifier. A Dyna Stereo 70 was sacrificed, it gave up its identity.
The A470 push pull output transformers were modified, all the laminations were removed, then all the E's were inserted through one side of the coils, and all the I's covered the ends of the E's at the other side of the coils.
I used a layer of Kapton tape over the ends of the E's, and then put the I's down on the kapton tape.
Kapton tape "was" the air gap.
The PCB went away, 2 EL34s went away, and the power supply was modified for better filtering to work with the non global negative feedback SE outputs. The EL34s were Triode wired.
The 16 Ohm tap was connected to the EL34 cathode, and the primary winding was connected so that was negative feedback, and not positive feedback. A 12AT7 dual triode became the stereo driver tube.
It sounded pretty good.
Over 2 decades ago, I did not know much about output transformers: materials, winding techniques, etc.
I wanted to make a stereo single ended amplifier. A Dyna Stereo 70 was sacrificed, it gave up its identity.
The A470 push pull output transformers were modified, all the laminations were removed, then all the E's were inserted through one side of the coils, and all the I's covered the ends of the E's at the other side of the coils.
I used a layer of Kapton tape over the ends of the E's, and then put the I's down on the kapton tape.
Kapton tape "was" the air gap.
The PCB went away, 2 EL34s went away, and the power supply was modified for better filtering to work with the non global negative feedback SE outputs. The EL34s were Triode wired.
The 16 Ohm tap was connected to the EL34 cathode, and the primary winding was connected so that was negative feedback, and not positive feedback. A 12AT7 dual triode became the stereo driver tube.
It sounded pretty good.
Last edited:
This is my first play into SE amplifiers. I have a boyuurange 50 amp that I've modified following skunkie exploration/mods plus a few more including an ultrapath mod(stock trans). I realize this is just an intro into SE 300b territory but also a learning path for me. I would like to try a better transformer set as a next step before I build a more "upscale" amplifier but would like to retain these new transformers for that future build. Speakers are Von Schweikert VR-5 HSE, Oppo- Meridian cd, VPI table etc. I mostly listen low volume blues, soft rock, easy jazz. Like others the SE presentation is unique, and I'm surprised how it has drawn me back into listening to music. Like most of us, we want full spectrum, smooth, detailed, life like presentation. "Iron" is touted as the SE road to this so here I am picking through the Iron minefield. Where next?
Why don't you plan the chassis layout for your new amp and leave the Reisong as it is, since it works. Like others, I'm a fan of separate power supply chassis - much more versatile for future builds and cuts down the overall weight. So start with the power supply - choose a chassis and then get a suitable B+ transformer, chokes, caps and a rectifier of choice, e.g. a 5U4. Once that's there go on to the output stage - the choice of output tubes and the OPTs. Then lastly figure out what driver tubes you like the sound of. Here you can think beyond the usual 9 pin tubes and look at octals e.g. 6SN7, 6J5 etc. If you have a modular structure with removable top plates that screw on you can make changes later.
I understand your logic/intent making good since. The modular structure/removal top plate is a great idea for the enthusiasts, tinkerers. Liken to a turntable/arm/cartridge journey.
I have been locked into this SE journey making a mod, then listening for changes driving a learning path. I've removed some past changes that weren't taking me where I wanted. I then became focused on the big step, better transformer as I had done the usual tube rolling, small circuit and component changes. I think we all know the infinite variations/combination of components and circuits possibilities. I'm to busy, maybe lazy to pursue half of them. I think this modded Chinese sounds good, not great, but surprises me. How much can be improved by selecting a reportedly much better output transformer? An expensive, but easily performed change. That's my curiosity speaking as I have no experience it this arena. Hoping for those experienced can add some light to what makes a great output transformer and what characteristics(windings, core, encapsulation, etc) can contribute to its sound. I haven't started into biasing changes etc.
I have been locked into this SE journey making a mod, then listening for changes driving a learning path. I've removed some past changes that weren't taking me where I wanted. I then became focused on the big step, better transformer as I had done the usual tube rolling, small circuit and component changes. I think we all know the infinite variations/combination of components and circuits possibilities. I'm to busy, maybe lazy to pursue half of them. I think this modded Chinese sounds good, not great, but surprises me. How much can be improved by selecting a reportedly much better output transformer? An expensive, but easily performed change. That's my curiosity speaking as I have no experience it this arena. Hoping for those experienced can add some light to what makes a great output transformer and what characteristics(windings, core, encapsulation, etc) can contribute to its sound. I haven't started into biasing changes etc.