demag transformers

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

desol

Well-known member
Joined
Jun 20, 2005
Messages
2,122
I didn't realize i could magnetize a small transformer. So, i used my DMM to test the windings on all my 2622 api transformers and neve inputs.  :'(  Stupid...

So now, i feel i need to demag them, i guess just to be sure. From what i've searched so far...it looks like i need 1-2v at 20-30hz sine wave for ?(5) seconds...then taper it down slowly.

I'm having a hard time finding a signal generator on ebay or elsewhere that won't tear a hole in my pocket and it's hard to ascertain what generator to even use? Hmm. Some only do RF, etc.

Can anyone recommend an affordable unit for this purpose? Will a software signal generator work?

Very much appreciated. Thank you.
 
This isn't something I've ever heard of.

That said, it is fairly believable (more believable than directional speaker/power cable that the Hi-Fi boys are into for sure).

There's no reason why audacity or something couldn't do what you need. Only thing will be you'll have to do some sums to work out what load resistor to put in the secondary to present big enough a load to keep 1-2v across the primary (if that's what you're trying to achieve) while driving sufficient current through the Tx to affect any permanent magnetism the core may have "caught". You could even make a nice fade in Audacity right down from loud to silence over a comparatively loonnggg time.

Also, I assume you'll be taking these transformers out of circuit for this....wouldn't recommend putting 2v up a mic pre!

I suspect loud program audio rather than tone would have a smiliar (possibly superior) effect.

Just out of curiosity, what are the symptoms of a magnetised transformer?
 
I continue to call bullshit in regards to 99.9% of possible transformers.    As plenty of others have said, if you're real worried, reverse your test leads and measure again the other direction.  Done; you are demagged.   Nah, it's bullshit.  Just as bad as directional cables in my book.  
 
Someone will need to test this theory with small core transformers and typical currents before we really know I suppose.

It's a bit like the old phantom power / ribbon mic argument.
 
Geoff tanner told me over at Aurora Audio. I mentioned i was using a DMM to test the windings on these mic inputs. He said that was a no no with small transformers.

Part of his reply:
"First thoughts are that it is a bad idea to test small audio transformers with a multi-meter. You will magnetise the core and this may affect the sound. Nowadays you can buy multimeters with inductance ranges (that work on an alternating current). You can demagnetise the transformers by zapping them with a large audio signal but referring to the easily accessible EDO transformer information on the net negates any need to test them this way."

I'd like to believe that it shouldn't be a concern, but there's experienced guys out there saying otherwise. So i'm not sure what, who to believe. Hmm.
 
If you feed a small trafo a very big amount of DC and leave it there for a very long long time, yeah *probably* you could magnetize it by a small amount.

Will there be a change in the trafo's performance. Probably yes, probably no.

Will it make the trafo sound horrible? I don't think so. The difference may not even be audible. 

Experiment:
Wound some magnet wire on a piece of nail, apply a 9V battery.
The nail becomes a magnet, right? Science 101. Picking up paper clips.

Now remove the battery, is the nail still magnetized?

 
One thing you might try is to run a sine-wave signal from the output of your DAW into the mic input with the gain set to minimum, at the highest level you can feed it before the preamp starts to clip. Record the output on your DAW, pull up the spectrum analyzer and look for a strong 2nd harmonic in the recorded signal. If it's not there, your transformer's almost certainly not magnetized.

Peace,
Paul
 
it is easy to test,

get an inductance meter,

get a new transformer,

measure the Henries

then zap it with the DMM set to ohms,

measure the Henries again.


usually, over time, the core will lose that bias, especially if you hit it with a good audio signal.

just plug an overcompressed cd line output into the primary,

use Black Sabbath - Into The Void,

it is in D flat, one of the few Sabbath songs in D Flat.

actually, the whole album is in D flat, my mistake.

Children of ther Grave will work also.

lower the volume at the end to prevent a zero crossing re mag.


i have to de mag these steel coils at work, the come from China.

they are on the ocean for 2 months, under influence of the Earths mag field, that is what i suspect, at least.


so be sure to rotate your mic inputs from time to time.  :D
 
Anything that can be fixed by Into The Void or Children of the Grave is worth breaking....

.....so you can fix it!  ;D

Seriously, this is one of those things that needs a controlled test performed on the people who claim it's a real problem.  I don't thing they can hear any difference at all.  It seems to come from references to certain Neve transformers, and certain V-series Euro stuff.  Never heard a real reference to anything else.  Damnit, it's like a bad flashback, saying this every few months when it comes up again.  Eventually I'll stop caring, and superstition can rule.  The whole idea that certain Neve transformers need to be cap isolated from phantom power is among the most insane things I've ever heard; if so, they are a defective design.  Yet, it gets repeated....and the price for these 'defective' transformers goes up and up and up. 
 
I knew I already read that somewhere. And I just remembered, it was here. Of course no further explanation.
Also worth noting, they only state that on the pages of the microphone transformers, line level seems to be fine.
 
I got all worked up about this a couple of years ago, and set up to prove it's an issue.

Couldn't make it stick.

Of course I used my U87, my previously used JT-13K7's, the current Sowter units, and a couple
of others I've since forgotten.

For love or money I couldn't get any of them to show significant effects for more than a few seconds using
variations of shorted P48 power, 48 volts fed thru 6K8 resistors.

I wasn't hugely scientific, but enough to convince me not to worry about it. I didn't try any of the shitty little Beyers or
other low rent transformers, but I wouldn't use 'em anyway.

I basically used a mic level signal generated by my HP 8903 thru a pad to a "mic input" with normal associated phantom
circuitry, with the transformer feeding a "normal" load resistor, fed back into the 8903.

In the case of the U87 I fed it into a typical INA-103 mic in stage, using film caps for coupling (10uF). Used the 8903
to feed the input of the impedance converter, than measured the output of the 103 into the 8903.

I fed the analyzer output of the 8903 into my Stanford Research SR770 FFT analyzer. No, it's not an AP, but I can do
most everything I need with the combo.

After sorting out various ground loops and noise in such a setup I could get low measurements that would show, briefly, the
change caused by magnetizing the cores by shorting one line or the other to ground, forcing the 7ma of phantom current
through the transformer.

I tried leaving things shorted for days, no significant difference. After a few seconds the residual distortion would fade away to
the original value.

Possibly the test signal was demagging? Well my test signal was in the -45dbm range, so power-wise it was in the normal
operating range of each of the transformers.

Different frequencies gave similar results.

I still get the creeps thinking about potentially crippled input stages, but I no longer have night sweats or heebie-jeebies, take it
for what it's worth.

 
I've had experiences of permanent magnetization with very little transformers [Studer and OEPs in my case], but only because of huge mistakes [big DC], or shorted traces or caps.
And yes you can see/hear easily the saturation via 2nd harmonic evidence and bad freq response.
The above mentioned method for demag them worked well.

Never seen a transformer magnetized by a DMM though.
 
Well, thanks for all your responses everyone. FWIW, i'm pretty sure i zapped the first channel input transformer on my API's with 48v...significantly(meaning a few hours cumulative). Since then i've recorded a few songs using that channel alone, bass response seems full, overall it sounds ok to me. So maybe i won't worry so much until i can actually prove or hear something.

You know how it is? Your building your prize preamps and want everything just right.
 

Latest posts

Back
Top