Magnetoresistive Compressor

[Michael A]

Anybody ever heard of, thought of, seen, heard or tried a compressor based on US patent 5724434? It is a bridge circuit that can be balanced/unbalanced by a magnetic field. Seems moderately do able. Would it work?

Michael

 

[NewYorkDave]

I don't have time to look at the patent now. But the first thing ya always gotta ask yourself is, "what could be the advantage of doing it this way?" Many things are do-able but not all of them are worth doing.

 

[CJ]

here is the link: 1996, so fairly recent...
http://www.google.com/patents?id=xjwjAAAAEBAJ&dq=5724434

Great idea, wonder about the response time.
Must be a ni alloy, permalloy lam, they expand, the crystals expand , that is, by way of an external mag field, this ,is why you don't clamp Ni lammed xfmr's.

just use la2 sidechain to drive it, or prr side chain.

I guess Phillips (leave it to phillips for weird ****) makes a mag bridge, KMZ10B.

Phillips PDF:

http://vacuumbrain.com/The_Lab/Compressors/mag_restrict_b.pdf

Wonder how it differs from an FW Bell Hall sensor?

 

[Michael A]

Worthy question.

It looks like it could be capable of very fast limiting. The device that was suggested was the NXP (used to be Phillips) KMZ10B. I haven't quite figured that bit out yet.

A quick model of a generic bridge seems to say it doesn't care if the input is single ended or balanced, ie same output either way. I don't know that there would be an advantage, could be some inherent disadvantages. Attenuating with a simple resistive bridge does look like it could be clean though.

Michael

 

[CJ]

Well, we have the 670, the 536, the LA2, the SSL, the 1176, so a new sound is always welcome, thats the way I scramble it.

Looks like a wam temp keeps it more linear, thats always the deal with mag sensing, linearty vs temp, real pain in the rear.

And perm magnets change with temp, also.
But probably not enuff for audio, more of a concern for instrumentation freaks.

 

[Michael A]

CJ, you're just kickin my butt all over the place to day. I just uploaded the figures and found the link for the datasheet and here you've done it all already.

I'm still not real sure how that bridge would work, like where you would put the magnets and coils to affect one resistance in the bridge. On the other hand he mentions the large change in resistance of bismuth with field, so I suppose a person could make a simple bridge and insert a handmade bismuth resistor. Then only one leg could be affected. The DC to 1MHz range of Phillips part does seem interesting.

Michael

 

[CJ]

i think its like self contained on that chip?
.
i like this compressor because it is made out of a hi nickel transformer lamination! :thumb:
hey, i wonder if that chip has a B-H curve?
:?:
I can see it now, "I made the mag comp out of 70-28-2 percent cobalt-barium, for that laid back bass..."

 

[Michael A]

If you go down the page at NXP where the datasheet, here You'll find a couple of technical documents that explain the chip more fully. IIRC it is 81/19 Fe/Ni. It also shows something of the physical construction. I'm just wondering how you concentrate the field on one resistor in the bridge. The laid back sound of that special alloy does sound like a good selling point. Too bad it's already patented.

Michael

 

[BradAvenson]

The way I read the datasheet is that only one resistor in the sensor is varied by the magnetic field, so maybe you could just put it in a field coil of sorts to try it out.

 

[emrr]

[quote author="CJ"]Well, we have the 670, the 536, the LA2, the SSL, the 1176, so a new sound is always welcome, thats the way I scramble it.
[/quote]

Not magnetic control, but Collins 26C and 26W are bridge circuits with transformer coils and variable tube impedances for unbalancing. Very nice sounding, and almost impossible to make them pump.

 

[walter]

I met a tech who sold patent rights to a magnet wah, too bad it went into a file cabinet instead of going into production. He showed me how a magnet can be used as a master volume on a tube amp.

 

[burdij]

Apparently, one of the resistors in the bridge is made of nickel wire or nickel film that exhibits a change in resistance dependent on a change in temperature or a change in longitudinally applied magnetic field based on the Thompson effect, the longitudinal thermoelectric effect (William Thompson - First Baron Kelvin - major contributor to the identification of the Second Law of Thermodynamics). I find it amazing that the patent office will grant a patent based on a physical effect that was described in 1851. This isn't even the first application of this process. Aren't patents supposed to be based on new or unique principles or processes?

 

[bcarso]

Look at the claims to see what was really granted---they may be fairly restricted.

Having said that, the patent office is in crisis pretty much---an awful lot of prior art gets patented and most don't have the money and time to sue to get them invalidated (Bose is one of the major offenders).

Lord Kelvin and Peter Guthrie Tait anticipated some of the recent developments in loop quantum gravity and string/brane theory with their vortex atom theory. In fact as part of this, Tait invented knot theory, which has since become an important maths discipline in its own right.

 

[gyraf]

Also consider, that as the magneto-resistive element seemingly needs to be used in a resistor bridge, then this is probably an indication of very-little change happening vs. mag-field - i.e. probably high noise when scaled-up to gainchanges that we would need in real-world audio applications.

Jakob E.

 

[clintrubber]

[quote author="gyraf"]Also consider, that as the magneto-resistive element seemingly needs to be used in a resistor bridge, then this is probably an indication of very-little change happening vs. mag-field[/quote]
Most likely it is indeed , but it doesn't need to be.
I mean, IIRIC the 'EL-patent' mentions a bridge-config as well.

Regards,

Peter

 

[mikep]

[quote author="gyraf"]Also consider, that as the magneto-resistive element seemingly needs to be used in a resistor bridge, then this is probably an indication of very-little change happening vs. mag-field - i.e. probably high noise when scaled-up to gainchanges that we would need in real-world audio applications.

Jakob E.[/quote]

exactly! which is why I was thinking: gonna need a BIG coil.

 

[walter]

Or a permanent magnet on a mechanical adjustment. put a magnet next to a tube or transistor and find out what happens.

 

[gyraf]

yes, we've discussed beam-tubes and magnetic deflection before...

 

[Michael A]

So I put on my gumboots last night and waded into the datasheets and auxiliary materials for a few back of the envelope calculations. Ok, so I didn't actually use an envelope, or a calculator, so with a grain of salt. Noise could be a problem, mainly because at the maximum linear imbalance the output seems to be around 8mV/V. That would be the zero compression point. So it would seem that 50dB of makeup gain would be necessary just to deal with the device. Also, there might be some problems with larger ac signals flipping the polarity of the field and giving frequency doubling. That might require either a dc bias voltage/current or that might be what the permanent magnet would take care of. The field necessary to unbalance the bridge is pretty minute though. It seems like 50 turns on a 1cm slug of iron with a 10mA current would put you at the maximum linear imbalance. Might be fun to play around with on a this is something different basis.

Michael