DCR and Impedance - modern audio transformers

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Sc

Well-known member
Joined
Feb 3, 2020
Messages
62
Location
Switzerland
EmRR brought up this in another post where I left topic by complaining about having lost a precious GZ34 due to low DCR of modern power transformers:

I haven't thought about that, it's certainly the case that audio transformers now versus then use larger wire with lower DCR, and it makes things like tube compressors behave and sound differently than the originals they 'clone'.
https://groupdiy.com/index.php?topic=76670.msg974470#msg974470

The upper GZ34 issue is easy to fix with 2 fat serial resistors of 50 - 100 R (and probably has no big effect on proper function).
But had the same experience EmRR mentions with modern audio transformers (and plate chokes). I suppose, it's not that easy just to use two parallel resistors to reach the original DCR .... and then have original behavior and sound. Of course core material is different today and I think frequency response on modern transformers is usually better. But how to get to original behaviour in theory ?   



 
I don't think there's so much difference between vintage and new xfmrs. M6 steel has not significantly changed, copper neither. Perhaps filling coefficient has increased marginally, due to better and thinner varnish and insulating materials, but I don't think it's significant. Xfmr designers still use the same charts as their fathers.
Indeed, toroidal xfmrs are quite different.
 
Measure DCR and inductance.  New transformers are totally unlike old ones. 

New transformers are much lower DCR and generally much lower inductance as they assume bridging drive.  Many (most?) fail the matching test for frequency response.   

The Hammond 800 and 850 series are the modern exception, because they've been in continuous production since probably the 1950's. They measure like vintage transformers.  They act like them.  No one uses them in DIY because they are EXPENSIVE.  But maybe not really more expensive than Sowter or Lundahl.  Sowter stuff doesn't look remotely like they transformers they claim to reproduce. 

Quality vintage transformers are spec'd for matching impedances and as such deliver full response under matching conditions, many can have source Z's higher than spec with little response change. 

Higher DCR and inductance in an interstage is required to really make a modern recreation of a vintage vari-mu deliver 'same/close as' performance due to differences in the DC voltage and impedance swings going in and out of compression.  They just don't sound/behave the same with modern lower resistance/inductance transformers. 

The downside to vintage style smaller wire with more turns and higher resistance is an earlier high frequency roll-off, so more advanced winding methods come into play to keep response out as far as possible.  Again, if you're 'reproducing' some old thing, that's part of it.  Thinner wire is also more likely to break over time from temp/humidity changes, though there's obviously plenty of vintage iron doing just fine.  Also more likely to fail from DC current imbalances.
 
EmRR said:
The Hammond 800 and 850 series are the modern exception, because they've been in continuous production since probably the 1950's. They measure like vintage transformers.  They act like them.  No one uses them in DIY because they are EXPENSIVE.  But maybe not really more expensive than Sowter or Lundahl.  Sowter stuff doesn't look remotely like they transformers they claim to reproduce. 

Interesting. I wonder why more people don't use the Hammond 850 series for DIY? This is the first I've been made aware of them.

The 800 series advertises slightly less bandwidth, but the 850 series advertises 20-20k +/- 0.5dB for impedances below 5k, and above 5k "may be down 1dB at 20kHz"

I'll certainly be interested in trying them in the future.
 
soapfoot said:
Interesting. I wonder why more people don't use the Hammond 850 series for DIY? This is the first I've been made aware of them.

The 800 series advertises slightly less bandwidth, but the 850 series advertises 20-20k +/- 0.5dB for impedances below 5k, and above 5k "may be down 1dB at 20kHz"

I'll certainly be interested in trying them in the future.

More money than most, and not 'exotic overseas'.  Part of it.  Measured specs are always better than stated, those are conservative ratings from what I've seen. 
 
soapfoot said:
Interesting. I wonder why more people don't use the Hammond 850 series for DIY? This is the first I've been made aware of them.

They come up in discussions on the forum now and then. I’ve used them in several projects and keep a little stockpile of used and NOS 850s around for vintage designs. As EMRR says, there’re no modern alternatives, or at least there weren’t until Crimson started cloning old UTC standards.

Here’s an interesting thread where the higher DCR issue is explored a little deeper. There are a couple of others as well that i’ll hunt down.

https://groupdiy.com/index.php?topic=68722.0
 
Thanks for that!

In my 20s when I was just starting to tinker with my own guitar amps, all the snobbiest amp people turned their noses up at Hammond (even for power transformers), insisting on the more expensive options by companies like Mercury Magnetics.

I didn't know any better, so I internalized some of that prejudice and I guess it persisted until now--I've never even considered Hammond for signal transformers, and never investigated their product line (I did begin using them for power transformers, however).
 
I say above there are no modern alternatives other than Hammond and Crimson. I have to add to that Cinemag, which makes several transformers that are built from their and others’ vintage designs, and which stick to the original specs without modification.
 
I am not sure we are comparing apples with apples. Certainly there are many transformers with superficially similar specifications that have lower dcr and inductance than similarly specified vintage ones. But that is as much due to the generally lax nature of manufacturers specifications of their transformer products.

The most obvious example is the 600:600 transformer. By itself 600:600 tells us  nothing about the transformer other than if it is an output transformer it expects to see a 600 ohm load. The problem comes from the driving side. In the vintage era it was expected that the source would also be 600 ohms so the primary inductance had to be sufficient not to load 600 ohms at the lowest frequency of interest. When transistors came along, it was soon possible to design output stages with much lower output impedances. This meant the primary inductance of a typical 600:600 transformer did not need to be so high to meet the same low frequency response requirement. This meant you could get away with fewer turns and/or use thicker wire. There is an abundance of such transformers available now but they are no good for vintage circuits that expect the source impedance driving the transformer to be 600 ohms. There is nothing wrong with the transformer or the manufacturers; they are just responding to the market.

What would be wrong would be if a modern transformer with low inductance and DCR was touted as a direct replacement for a vintage transformer that required more inductance. I see no evidence of this.

Cheers

Ian
 
ruffrecords said:
I am not sure we are comparing apples with apples.
What would be wrong would be if a modern transformer with low inductance and DCR was touted as a direct replacement for a vintage transformer that required more inductance. I see no evidence of this.

Cheers

Ian

You're saying the same thing I am.

Except the part about the actual amount of inductance in vintage parts, compared with the reproductions made by Sowter/etc, which are much lower.  Not saying they don't work in the circuits, but they aren't comparable parts and as such behave differently. 

I'm addressing the average non-expert trying to use a modern 600:600 as if it were a vintage 600:600 in a matching condition, which is very much reality, they don't know WHY they are different.  They don't know there's a difference between bridging and matching source drive, and that the specs don't usually say what standard is being used, YOU have to know what the standard is for the era the data is from in most cases.  I've seen 600:600's that only passed full lows when driven with a 50 ohm source, which is still lower than most of what's out there amongst modern converters, many are 150. 

Or a modern 10K:10K (described) Sowter interstage in an RCA limiter clone that used something that would work fine 50K:50K with 5-6x the inductance.
 
EmRR said:
You're saying the same thing I am.

Except the part about the actual amount of inductance in vintage parts, compared with the reproductions made by Sowter/etc, which are much lower.  Not saying they don't work in the circuits, but they aren't comparable parts and as such behave differently. 

Not that I don't believe you but can you give me some concrete examples?

Edit: As a rule I have not found Sowter products to be shy on inductance

Cheers

Ian
 
I believe we could all agree that a modern 600:600 output xfmr is probably different than a vintage one.
In vintage parlance a 600:600 is designed to be driven from a 600r source, and deliver into a 600r load, when the modern counterpart is meant to be driven by a very low Z (typically an active stage with considerable NFB), and be loaded with a light load, of a few kohms.
Some manufacturers produce 600:600 xfmrs that are specifically made to address situations where the user has no control on the source impedance (typically 50-200r from a soundcard or other piece of gear).
In that respect, they are neither "modern" nor vintage.
 
ruffrecords said:
Not that I don't believe you but can you give me some concrete examples?

Edit: As a rule I have not found Sowter products to be shy on inductance

Cheers

Ian


Buy ANYTHING they sell as a repro of some old thing, directly compare against the old thing.  It's across the board with pretty much anything modern.  I could take the time to make a specific list, but it's hardly worth the effort. 
 
For modern transformers distortion tends to go up with source impedance.  Is the same true for vintage ones? Or will a vintage 600:600 have lower distortion at 600 ohm source than at 0 ohn source?
 
EmRR said:
Buy ANYTHING they sell as a repro of some old thing, directly compare against the old thing.  It's across the board with pretty much anything modern.  I could take the time to make a specific list, but it's hardly worth the effort.

No need for a list - just give me one typical example.

My own experience of Sowter has been very good. For example I originally used their Helios mic input transformer for my tube mic pre designs> I measured the inductances and they are more than adequate.

Cheers

Ian
 
EmRR said:
Buy ANYTHING they sell as a repro of some old thing, directly compare against the old thing.  It's across the board with pretty much anything modern. 
There is no technical reason for that. The modern materials are at least as good and the winding techniques have not regressed.
The only reasons are:
Economics: some mfgrs have found that they can pass inferior products to uneducated buyers
Incompetence: some mfgrs think they've got it right, but they haven't.
Worst case is when both factors are added.
As Ian said, and I concur, Sowter, until now, have always been capable of producing replacements for vintage magnetics that perform at least equally well.
I'm not sure this will continue under the OEP tutelage, since I have found them often guilty of one of the aforementioned shortcomings.
Now, with all respect due to Lars Lundahl, I would not expect them to produce equivalents to vintage designs, since they have their own twists.
I am not aware of any winder other than Sowter, that produces genuine copies of vintage magnetics on any significant scale.
Jensen, Cinemag, ... produce their own "replacements", which are capable of answering the same basic specs as the vintage unit they are meant to replace, but not up to the last detail.
Often, "modern" mfgrs make modifications that improve some feature over the vintage design, at the risk of loosing over some subtle aspect.
 

Latest posts

Back
Top