I am surprised
That i haven't seen more quantitative info here for the OP. Age doesn't matter, none of them are old enough to suffer strictly from age. A newer Sowter for example (or EP, or MQ etc etc) could have just as easily been left sitting right on top of a loudspeaker, or the runt of an amplifier gone smoking. What matters is how the transformer was made. Google is your friend for this kind of stuff, where one thing we all can not argue is that information is easier to come by than it has been in the past.
SAC Thailand
This is a great write up where actual real world tests are shown. Look for yourself, you can see the quality, and lack of quality, in some big name brand transformers.
Push-Pull Transformer Test
This is a very comprehensive set of tests done a few years back. We all have heard about how good dynaco transformers are for example. Well, look. Here you can see just how good, and why. Especially the Z-565 from the SCA-35, which stands out from the crowd much, much more than the A-470.
Personally I have had very good luck with older transformers, when they are from well taken care of amps. This encompasses for more than not of the old ampos and transformers I have come across. so many of those beauties were very well taken care of, and only fell by the wayside because of the oncoming of the "Newer, better, faster, smaller" transistorized stuff. Remember, a lot of what we now call "amazing equipment" was just what was available. Do you think Aunt Rose really cared if she had a Dynaco SCA-35 hooked up to the record player or a Sears and Roebuck transistorized receiver? Ya, she cared that the Sears receiver wasn't hot, and claimed it would work forever without any adjustment.
I recently did a 6BG6GA amp with old Scott transformers, from a 299-C. Came off an ampo sitting rusting on the back of a somewhat leaky bus. Rusted like something from the titanic. Brought them home, cleaned them up. opened them, perfect inside still. Built the amp, and the sound is far superior to offerings from Hammond or Edcor or a few of the other cheaper brands.
Does it mean they are "better" than comparable transformers today? No, that's what the word "comparable" means. But you get what you pay for with new stuff. You'll notice that Edcor, Hammond, James, these are cheap companies (compared to the other OPT makers). And, as you'll see when you look at that silk link I posted, well, they don't perform nearly as well do they.
The joy of some of the older stuff is that sometimes you can get way more than you paid for. Like these fantastic Scott transformers I got for free. It's good to listen to everyone's opinion because everyone hears differently, and it's important to factor in all the info you can to a final decision.
But look at the links, get some real world information, and make your own choice. I think it's safe to say that you can be confident transformers that test very well sound better in A/B testing to ones that don't test very well.
Loren
KB2WYL
That i haven't seen more quantitative info here for the OP. Age doesn't matter, none of them are old enough to suffer strictly from age. A newer Sowter for example (or EP, or MQ etc etc) could have just as easily been left sitting right on top of a loudspeaker, or the runt of an amplifier gone smoking. What matters is how the transformer was made. Google is your friend for this kind of stuff, where one thing we all can not argue is that information is easier to come by than it has been in the past.
SAC Thailand
This is a great write up where actual real world tests are shown. Look for yourself, you can see the quality, and lack of quality, in some big name brand transformers.
Push-Pull Transformer Test
This is a very comprehensive set of tests done a few years back. We all have heard about how good dynaco transformers are for example. Well, look. Here you can see just how good, and why. Especially the Z-565 from the SCA-35, which stands out from the crowd much, much more than the A-470.
Personally I have had very good luck with older transformers, when they are from well taken care of amps. This encompasses for more than not of the old ampos and transformers I have come across. so many of those beauties were very well taken care of, and only fell by the wayside because of the oncoming of the "Newer, better, faster, smaller" transistorized stuff. Remember, a lot of what we now call "amazing equipment" was just what was available. Do you think Aunt Rose really cared if she had a Dynaco SCA-35 hooked up to the record player or a Sears and Roebuck transistorized receiver? Ya, she cared that the Sears receiver wasn't hot, and claimed it would work forever without any adjustment.
I recently did a 6BG6GA amp with old Scott transformers, from a 299-C. Came off an ampo sitting rusting on the back of a somewhat leaky bus. Rusted like something from the titanic. Brought them home, cleaned them up. opened them, perfect inside still. Built the amp, and the sound is far superior to offerings from Hammond or Edcor or a few of the other cheaper brands.
Does it mean they are "better" than comparable transformers today? No, that's what the word "comparable" means. But you get what you pay for with new stuff. You'll notice that Edcor, Hammond, James, these are cheap companies (compared to the other OPT makers). And, as you'll see when you look at that silk link I posted, well, they don't perform nearly as well do they.
The joy of some of the older stuff is that sometimes you can get way more than you paid for. Like these fantastic Scott transformers I got for free. It's good to listen to everyone's opinion because everyone hears differently, and it's important to factor in all the info you can to a final decision.
But look at the links, get some real world information, and make your own choice. I think it's safe to say that you can be confident transformers that test very well sound better in A/B testing to ones that don't test very well.
Loren
KB2WYL
I'm the author of the P-P transformer tests mentioned in post #21. These tests mainly showed the frequency response and smoothness, and only the grossest low-frequency distortion. However, there is something that makes a transformer sound good or bad, regardless of the frequency behavior. I'm pretty sure this is the quality of the core material as well as the dielectric behavior of the insulation. Based on many years of testing and listening, here is what I have found. A lot of this is subjective, but others have noted the same things.
I've auditioned some amplifiers from the 1920s and 1930s (some as part of radios), often with excellent tubes, but have never found them to sound great - the problem usually being a "muddiness" or lack of detail. I have also noticed this in some replacement-grade output transformers. In particular, I hooked up a Thordarson "universal" output transformer into an excellent SE amp. It sounded terrible (same muddiness, etc.). I later tested it and found that there was significant distortion, presumably from the core.
During World War II significant advances were made in core materials, especially in grain-oriented C-cores and laminations. When the first "hi-fi" amplifiers came out in the late 1940s, they used these newer materials, and the results were often outstanding. Transformers from this "Golden Era" of hi-fi that I have personally found especially good are: Acrosound, certain Peerless transformers, Partridge, and the transformers found in Marantz, H-K Citations, and the Dynaco ST-35. Just slightly below in quality, but still very good are the transformers in the Scott and Fisher receivers, the Dynaco ST-70 and Mark III, and the smaller Heathkits. There are some "sleepers", such as the output transformer in the EICO HF-50. The qualities I listen for are: tonal balance, detail, and an ability to transfer the emotion of the music.
An interesting counter-example is the UTC "Linear Standard" series. I had a push-pull 6V6 amp that used an LS-55 output transformer, and I could never get it to sound good, despite all sorts of circuit changes. I have other modest P-P 6V6 amps that sound great, and I ended up suspecting the output transformer being the limiting problem, despite the beautiful construction and marketing hype. The Linear Standard product line came out in the late 1930s, so may have some problems in common with the pre-war designs. This is not a sweeping indictment of the Linear Standard series, just one observation.
Another observation: I've restored many theatre and PA amps (Altec, Ampex, Bogen, etc.) and despite decent frequency response and THD specs, have never found them as good as good hi-fi amps. The issues are lack of detail, grainyness, or lack of emotion. I've also used Altec and Ampex output transformers in my own tube amp designs and found the same deficiencies. These output transformers are not quite so good.
With modern transformer technologies, we have new core materials (metglas aka amorphous), better and different insulating materials (mylar, kapton, nylon wire coating, etc.), and nylon bobbins. I've heard modern tube amps that sound excellent, but in using both new and good vintage transformers in new amp designs, I still find the good vintage transformers hold their own. As long as the transformer has not been abused or kept in a bad environment (next to the ocean or in the tropics), old transformers should still work well. Virtually all American transformers were potted or vacuum impregnated, which mostly protects them. I've seen output transformers in German radios from the 1950s that were not impregnated - these would be questionable in a humid environment. It is a good idea to fuse the B+ to keep a shorted tube from burning out the output transformer.
Some people feel that the insulation used in transformers make a big difference. In particular, some have a preference for the kraft paper or similar cotton/paper mixtures versus the modern mylar films. This is in the same category as preferences in capacitor dielectrics. I won't wade into this swamp, but the use of vintage dielectrics may be a reason some vintage transformers sound good.
TL;DR - Many vintage hi-fi transformers are fine - try them out!
- John Atwood
I've auditioned some amplifiers from the 1920s and 1930s (some as part of radios), often with excellent tubes, but have never found them to sound great - the problem usually being a "muddiness" or lack of detail. I have also noticed this in some replacement-grade output transformers. In particular, I hooked up a Thordarson "universal" output transformer into an excellent SE amp. It sounded terrible (same muddiness, etc.). I later tested it and found that there was significant distortion, presumably from the core.
During World War II significant advances were made in core materials, especially in grain-oriented C-cores and laminations. When the first "hi-fi" amplifiers came out in the late 1940s, they used these newer materials, and the results were often outstanding. Transformers from this "Golden Era" of hi-fi that I have personally found especially good are: Acrosound, certain Peerless transformers, Partridge, and the transformers found in Marantz, H-K Citations, and the Dynaco ST-35. Just slightly below in quality, but still very good are the transformers in the Scott and Fisher receivers, the Dynaco ST-70 and Mark III, and the smaller Heathkits. There are some "sleepers", such as the output transformer in the EICO HF-50. The qualities I listen for are: tonal balance, detail, and an ability to transfer the emotion of the music.
An interesting counter-example is the UTC "Linear Standard" series. I had a push-pull 6V6 amp that used an LS-55 output transformer, and I could never get it to sound good, despite all sorts of circuit changes. I have other modest P-P 6V6 amps that sound great, and I ended up suspecting the output transformer being the limiting problem, despite the beautiful construction and marketing hype. The Linear Standard product line came out in the late 1930s, so may have some problems in common with the pre-war designs. This is not a sweeping indictment of the Linear Standard series, just one observation.
Another observation: I've restored many theatre and PA amps (Altec, Ampex, Bogen, etc.) and despite decent frequency response and THD specs, have never found them as good as good hi-fi amps. The issues are lack of detail, grainyness, or lack of emotion. I've also used Altec and Ampex output transformers in my own tube amp designs and found the same deficiencies. These output transformers are not quite so good.
With modern transformer technologies, we have new core materials (metglas aka amorphous), better and different insulating materials (mylar, kapton, nylon wire coating, etc.), and nylon bobbins. I've heard modern tube amps that sound excellent, but in using both new and good vintage transformers in new amp designs, I still find the good vintage transformers hold their own. As long as the transformer has not been abused or kept in a bad environment (next to the ocean or in the tropics), old transformers should still work well. Virtually all American transformers were potted or vacuum impregnated, which mostly protects them. I've seen output transformers in German radios from the 1950s that were not impregnated - these would be questionable in a humid environment. It is a good idea to fuse the B+ to keep a shorted tube from burning out the output transformer.
Some people feel that the insulation used in transformers make a big difference. In particular, some have a preference for the kraft paper or similar cotton/paper mixtures versus the modern mylar films. This is in the same category as preferences in capacitor dielectrics. I won't wade into this swamp, but the use of vintage dielectrics may be a reason some vintage transformers sound good.
TL;DR - Many vintage hi-fi transformers are fine - try them out!
- John Atwood
thanks John, awesome work, do you have links to your test setup so that we can do our own?
Same thickness Nomex should give roughly same results as Presspahn (treated transformer paper)
Presspahn - 100 Presspahn
Presspahn - 180 Nomex
You should never ever use regular untreated paper in a transformer.
Presspahn - 100 Presspahn
Presspahn - 180 Nomex
You should never ever use regular untreated paper in a transformer.
Is there any relation to the listening test for those less performing OPT to the uneven FR measurements and the abrupt inductance change?
Is the LS-55 the older Black Case or the newer Gray Case?
AJT: The transformer tests linked to above were done 26 years ago using a modified Heath W-6M amp as the test bed and a conventional sine/square-wave oscillator and AF voltmeter. More details can be found at my web site: Push-Pull Transformer Test.
Since then, I acquired an Audio Precision System 2, later upgraded to a SYS-2522 "Cascade". This allows me to make some very detailed analysis of harmonic distortion vs frequency and amplitude, among other things. I have run tests on a variety of input and output transformers, but the results are not at hand at the moment. There definitely is a correlation between measured distortion behavior and subjective quality, but you need to look deeper than just a simple THD measurement.
megasat16: The amplifier with the LS-55 was disposed before I got my Audio Precision set-up, so can't give detailed measurements. The LS-55 was the grey type probably made around 1947. I've got a few pre-war LS transformers in the black cast-iron cases - if I have time, maybe I can test them.
Regarding the 1990 transformer tests, the ones with bumpy square wave response were not always bad sounding, however, they would be problematical if designing them into high-feedback amps. The ones with high saturation frequencies indicate insufficient inductance or core size, which will color the low frequencies. However, some of the smaller transformers from inexpensive PP amps sometimes sounded great, others, not so much. Hence the quest for better test equipment.
- John
Since then, I acquired an Audio Precision System 2, later upgraded to a SYS-2522 "Cascade". This allows me to make some very detailed analysis of harmonic distortion vs frequency and amplitude, among other things. I have run tests on a variety of input and output transformers, but the results are not at hand at the moment. There definitely is a correlation between measured distortion behavior and subjective quality, but you need to look deeper than just a simple THD measurement.
megasat16: The amplifier with the LS-55 was disposed before I got my Audio Precision set-up, so can't give detailed measurements. The LS-55 was the grey type probably made around 1947. I've got a few pre-war LS transformers in the black cast-iron cases - if I have time, maybe I can test them.
Regarding the 1990 transformer tests, the ones with bumpy square wave response were not always bad sounding, however, they would be problematical if designing them into high-feedback amps. The ones with high saturation frequencies indicate insufficient inductance or core size, which will color the low frequencies. However, some of the smaller transformers from inexpensive PP amps sometimes sounded great, others, not so much. Hence the quest for better test equipment.
- John
Last edited:
Hmmm, I wonder if this was Dennis Boyle who used to be at the
old Dallas Works, Western Electric, Bell Labs, Lucent Technologies.
I think I may have worked at the same site with him and was referred to
him when I had some transformer questions.
My question at the time was benefits of using torroidal transformer.
His response (or the resident expert at W.E./Bell Labs with whom
I spoke) was to use the E.I. transformer, that its imperfections can
be well mitigated. By using a torroidal transformer will show that
some specs will be better, but there are a whole slew of short
comings and other problems they will introduce.
Thanks John and others for your thoughts and links on this.
John, another old acquaintance of mine made some of the same
observations as you about the old big PA amplifiers, the DynacoIII
and some others. He was building a lot of amps at the time and
checking out all the vintage iron to see what would work and
sound good. His observations/findings match yours. I can't
ask him for comments as he passed on last year.
I think I'm going to replace the output iron in my old Vibrolux reverb amp.
It never sounded as good as another I brought back to life and used
the W W brand. I've still got some in stock.
In summary, good advice, try what you have it might be good.
EDIT ADDED:
Haven't read the links yet, but if you are going to do some testing,
I'd be interested in performing some tests as well. If you have a
good methodology I can attempt to duplicate. I have a mix of
famous iron, new iron, autoformer, and some new never used
quality stuff along with a smattering of some good oscillators
and some simple to dial frequency generators, distortion analyzers
and FFTs.
I started evaluating 3 automotive ignition coils and found a few
reactances from them at 16kHz, 70kHz up to around 320kHz.
I could use the distortion analyzer to find reactances from
320kHz up to almost it's 5th harmonic. The distortion analyzer
petered out somewhere in the 900kHz, well above it's spec.
I ran into problems trying to find a HV probe to 50kHz
that didn't have high frequency attenuation problems.
I don't have the Skill or knowledge to build a proper
one currently and I started investigating that.
Then, I found reference to an old electronics text and
bought one...it's been outside since I received it feeling
and reeking of milldue. It's the old Page & Adams text
of Electricity and Magnetism.
Cheers,
Sync
Last edited:
You'd better asked for the benefits of using a c-core transformer.
The expert should have told you that c-core transformers combine the benefits of EI and toroidal transformers.
It is really striking that the vast majority of non US transformer manufacturers (Tamura, Tango, Lundahl, Audio Note, Monolith Magnetics, me to name just a few) use c-cores for their high quality transformers, whereas US manufacturers still stick to EI (Dave Slagle the positive exception).
McIntosh back in their good days used c-cores but that is already long ago.
@Pieter, He may very well have mentioned C cores during our conversation, but it was 20 or 25 years ago. Quite possibly described their limited
availability, problems sourcing them, etc. The advise at the time,
was best to stick with a good high quality EI as there is nothing
wrong with them.
I would be interested in learning more about the C cores and how they
are made.
Cheers,
Sync
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.