DC measurement is magnetizing core ?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

PermO

Well-known member
Joined
Mar 16, 2011
Messages
792
Location
Hilversum, Netherlands
Hello,

I know that DC voltage can permanently magnetise the core of a transformer.

But if I have an unknown (audio) transformer and I use a Fluke multimeter to measure the resistance to somehow figure what is what on the transformer...

Will this 'damage' the transformer ?

I ask this as someone mentioned I screwed up two transformers using my Fluke on the windings.
Does a multimeter for a few secs produce enough current to permanently magnetize a transformer core ?

Is this really so ?

 
PermO said:
Hello,

I know that DC voltage can permanently magnetise the core of a transformer.
Current, not voltage.

Does a multimeter for a few secs produce enough current to permanently magnetize a transformer core ?
It would take a brute of a multimeter to do that.
Nothing can "permanently magnetize a transformer core". It's always possible (and easy) to demag.
 
Tape recorder heads often come with a warning about applying DC , Im not sure a high impedence modern meter would be a problem , but probably not appropriate to connect something like an AVO 8 to measure the coil resistance in a tape head , well you could always de-gauss afterwards I susppose .
 
I have received messages on ebay from people telling me I had permanently destroyed transformers I had listed for sale with DCR measurements. I try to gently convince them that even if my meter could magnetize the core, the transformer could still be de-magnetized. Nobody ever writes back.

And yes, there are people who degauss small transformers for mics and mic inputs. Sometimes they even take measurements to show improvement in the low frequencies. Never tried it myself...yet.
 
In transformers, ( never tape heads) ....Sometimes a small magnetic offset, or a DC offset is deliberately applied, to shift the magnetic reversal hysteresis gap away from
zero crossing and up into large signal where its swamped and made inaudible. You do loose some maximum signal, by the amount of off-set you are adding.
 
Putting such a DC bias in a transformer winding simply causes a shift of the spectral mix of harmonics in the distortion. Without DC bias, distortion is almost purely odd-order (3rd, 5th, etc.) due to the symmetrical nature of the magnetic non-linearity. But DC bias introduces a non-symmetrical component, adding even-order (2nd, 4th, etc.) distortion products to the mix. With enough DC, even-order becomes the dominant distortion. This is responsible for the "warm" sound of transformers driven by single-ended class-A drivers (as well as the asymmetrical distortion in the driver itself). Since even-order distortions have an octave-up relationship to the original note, they're usually considered more "pleasant." But the other price-to-be-paid for significant DC in a transformer winding is the reduction in inductance due to introducing an air-gap in the magnetic circuit to prevent magnetic saturation. At Jensen, we considered the loss of inductance (and resulting loss of low-frequency response) unacceptable to our fidelity standards - so Jensen never made, for example, single-ended plate to push-pull grid transformers. It's also why we recommend an AC-coupled "totem-pole" vacuum-tube driver for the JT-10K61 output transformer (instead of a push-pull transformer with a center-tap).
 
In transformers, ( never tape heads) ....Sometimes a small magnetic offset, or a DC offset is deliberately applied, to shift the magnetic reversal hysteresis gap away from
zero crossing and up into large signal where its swamped and made inaudible. You do loose some maximum signal, by the amount of off-set you are adding.
I don't think there is any benefit gained from this, axcept the introduction of a supposedly euphonic "harmonic enhancement".
Just another dirtyfier IMO.
 
I don't think there is any benefit gained from this, axcept the introduction of a supposedly euphonic "harmonic enhancement".
Just another dirtyfier IMO.
An engineer I know in Canada has done exhaustive THD+n measurements at very low signal levels and proven it. For rock and roll at near saturation it is stupid.. for a classical recordist ever fraction of improvement at low signal levels is critical. I'm not saying it should be done always, only pointing out that blanket generalities are stupid as there often exceptions that are meaningful in certain outlier cases. Never say 'never'... say instead, 'most often'. Magnetic Hysteresis distortion is audible and its real and its measurable, and has been for ages.
 
An engineer I know in Canada has done exhaustive THD+n measurements at very low signal levels and proven it. For rock and roll at near saturation it is stupid.. for a classical recordist ever fraction of improvement at low signal levels is critical. I'm not saying it should be done always, only pointing out that blanket generalities are stupid as there often exceptions that are meaningful in certain outlier cases. Never say 'never'... say instead, 'most often'. Magnetic Hysteresis distortion is audible and its real and its measurable, and has been for ages.
I must admit I did not understand what you were pointing at, now I got it.
Similar in a way to DC bias in magnetic tape recording.
I believe ultrasonic bias would be a much better solution, but I wouldn't kill brain cells over the subject in an age where xfmrless mic preamps have reached near perfection. The best xfmr is no xfmr.

BTW, when checking the manual for the Telefunken M15 tape machine, I found that the bias current of the 5534 goes through the PB head. there may be other examples.
Actually several mic preamps have the secondary carrying the bias current. Is that a deliberate attempt to shift the mag curve?
 
Last edited:
Magnetic Hysteresis distortion is audible and its real and its measurable, and has been for ages.
A couple places I worked at wound their own transformers. One of the two actually specified a special annealing process for the cores in a magnetically anechroic chamber. The other did not. The reasoning the place that did gave was low frequency distortion performance, and we (read as I) _did_ test for it in production. There were war stories of the time they 'forgot' to anneal the cores properly, and the distortion (and the rework involved) was pretty significant, I was told.

I'm not sure of the overall value to the end user, but less distortion is generally better IMO.

That said, one of the nice things about transformers, besides the CMRR, is the kind of natural limiting the core saturation gives. With a bias to start with, the distortion becomes asymetric, which I think just bugs me from a philisophical standpoint more that anything.

The best xfmr is no xfmr.

Of course too is: if you're shooting for 0.0000% distortion, then, no, you probably don't want a transformer anyway.

Anyhoo, the phenomenon we're talking about here is:

https://en.wikipedia.org/wiki/Remanence
 
Hysteresis distortion can be made vanishingly low. For example, look at the data sheet https://www.jensen-transformers.com/wp-content/uploads/2014/08/jt-11k8-apc.pdf, which is typical of Jensen input types with 80% nickel alloy core material. Look at the THD vs Signal Level on page 2. On the right you have the gradual approach to magnetic saturation but the slight bump in the -30 to -40 dBu region is hysteresis distortion. The frequency changes because the perm of the material is higher at 20 Hz than at 50 Hz. Anyway, the bump adds very little distortion. More importantly, it's lower than other brands because the annealing (heat treatment process) of the material is the one and only Jensen trade secret. It's extremely difficult to get our core material suppliers to do it "our way" instead of their standard, but it gives Jensen the lowest low-level THD in the industry. Most other suppliers don't even have the courage to publish such a detailed. Someone on the forum pointed to a Haufe T226 "data sheet" - it's so sketchy as to be useless!

Bill Whitlock
(for those who may not know, I was owner/chief engineer of Jensen Transformers from 1989-2014 after my friend and founder Deane Jensen killed himself in 1988)
 
Cheers Bill ,
There was an interesting discussion here before about a special furnace used for the anealing process of lams ,
Was an eastern European lad who seemed to have a backround in metalurgy , his name and nationality escapes me , as far as I remember it was oxygen and hydrogen but the exact methodology I cant remember , its around here somewhere .
 
It's a hydrogen annealing process but the tricky part is the temperature profile (the number of hours at each temperature plateau) so that the crystalline magnetic domains completely randomize in their orientations (think of it as "relaxation"). The temperatures are high enough to make the alloy glow a dull orange. Also, specific additions of minute amounts of molybdenum and chromium to the 80/20 % nickel/iron alloy can result in near-zero magnetic hysteresis (i.e., zero magnetic remanence).
 
Last edited:
Touch of Alchemy about it , although of course we do know how even minute amounts of rare earth metals and other elements can have profound effects on the properties of alloys .
 
The temperatures are high enough to make the alloy glow a dull orange.
A related story I find interesting: The samurai bladesmiths would go thru an elaborate process to get a sword that would both take a razor sharp edge, and yet remain a bit flexible, so it wouldn't crack in battle. Two competing design goals, actually. They achieved the balance by coating the blade in clay of varying thickness. When it was quenched after heating (tempered), the thicker clay would cause that part of the blade to cool more slowly, and the steel beneath the thinner clay would cool more quickly. This achieved the desired different hardness of the various sections.

The key was getting the blade's temperature right before it was dunked. Of course, they had no thermometers, no modern knowledge of crystal structures in metals, and really only (centuries of) experience of what worked and what didn't to guide them. The ritual (and yes, that's the right word for it) was to heat the blade to "the color of the morning sun" before quenching. Sound familiar? :)

Touch of Alchemy about it

I kinda feel like, like swordmaking, creating a good transformer is at once art, science, alchemy, and / or a bit of magick.

BTW: Jensen white papers were how I learned about CMRR and balancing circuits coming up. Thank you for writing those, sir.

If you haven't read them, that's probably where you should start with understanding audio transformers. Whitepapers | Jensen Transformers
 
Back
Top