Buzz Kill Success w/ MuMetal

[stephenl]

I was kind of skeptical dropping $40.00 on an 8" x 12" sheet of this stuff - but it actually worked!

I used about 1/4 of the sheet. 3" square sheets above and below the output transducer, a 1" x 3" piece on the end, and I wrapped a little 1/4" wide strip 2/3 the way around the coil itself.

The buzz is barely detectable with everything on 10 and no input signal - huge improvement...worth the $40 .

[dorrisant]

Can you post a picture of your project?

[sonicmojo]

I wonder if MuMetal would reduce the hum between transformers on my TW Rocket clone where the transformers are fairly close but even at 180 degrees off axis I have slight hum. Right now I have an aluminum backplate from a pedal stuck between them and that helps about 50%. Has anyone tried this stuff for that type of application?

[TUBEDUDE]

Where did you get the Mu metal?

[sonicmojo]

Where did you get the Mu metal?

I see a lot of options in different sizes, most .004 thick, on eBay.

[stephenl]

I bought it off Amazon - .004 thickness.

I should have taken pictures pics, but I was in “ get it done” mode.

I installed it as described above. I cut the pieces with scissors and put them in place with silicone adhesive. The only piece I bent was the thin strip directly around the coil. I was warned by several people not to work it too much. Stressing it reduces its effectiveness.

I think it would be similarly effective at decoupling the PT/OT on a Wreck, as mentioned above. Four it to size and mount it vertically on brackets. You’d probably want to stick it to another piece of metal for rigidity - .004 is pretty thin.

[R.G.]

Mumetal is good for small external fields. It's common in the old CRT based oscilloscopes to wrap the CRT with a cone of mumetal to reduce beam wiggle from external fields. See https://en.wikipedia.org/wiki/Mu-metal

Magnetic fields are very difficult to shield. There is no such thing as a magnetic field "insulator" like there is for electrical fields. The "resistance" of free space to magnetic fields is too low. Ferromagnetic materials can provide some shielding because they have an even higher "conductivity" for magnetic fields than free space. The ratio of how much better they are at conducting magnetic fields than free space is termed their "permeability", and normal transformer iron can be up to maybe 10k to 20k times better than free space.

Mumetal has a permeability of 80K to 100K if everything is just right, so it can be much better than even transformer iron. If everything is right, that is. When any ferromagnetic material saturates, its permeability drops to the same as free space, so ferromagnetic materials are only good up to some maximum magnetic field. Above that field intensity, they may as well be cardboard. Mumetal has a very high permeability, but also extra low saturation density. So it goes free space on you at low field intensities. But that is enough in some instances.

Conductors can also shield a CHANGING magnetic field. Non-static M-fields cause eddy currents and other field effects in conductor sheets, and this either eats the energy in the M-field as heat or diverts it with field effects. The Faraday cage construction uses field effects to shunt changing fields around the outside of the cage.

Highly effective magnetic shielding uses both ferromagnetic and conductive layers to increase their effectiveness. Good mike input transformers may have a four or five layer shield of mumetal/copper/mumetal/copper/mumetal/copper... to do this.

If a sheet of mu-metal does good, that's great. It's quite expensive, as you found. Check your scissors, though. Mumetal is notoriously hard on cutting tools, as it work-hardens just ahead of the shearing zone. And as you may have found out, it cuts hands like a razor blade, especially in thin sheets.

I would be very interested in seeing how different your results are with a box of 0.080" soft aluminum sheet around your OT is from the mu-metal. I've always wanted to see this, but never would pony up the money for mumetal. That's not just cheapness - it used to only be available in $100 minimum orders, so having it available in an 8x12 sheet for $40 is new to me. It's a great opportunity to test the theories in the real world.

In shielding, you try to shield the source of the noise first, then shield the target. In this case, it would be useless I think to put mu-metal around the PT or choke, as the fields are big there. Putting thicker aluminum or even sheet steel around the noise sources would cut the radiated hum somewhat. Mumetal should be reserved for low-intensity areas, like the OT. It's probably most useful on PP OTs, as I suspect that the "DC" leakage field around an SE OT might saturate it.

[Malcolm Irving]

With some amps you can hear the hum magnetically induced into the OT, through the speaker, when the power tubes are removed. (All the other sources of hum are eliminated as there is no output stage.)
But, with the power tubes removed, there is much less current in the PT and choke, so their magnetic fields are low.
Haven’t tried it (yet) but it would be interesting to create restive loads to draw the correct idle currents, emulating the (removed) power tubes - plate currents and screen currents if you want be fancy. The primaries of the OT would need to be disconnected for the test. The currents through the PT and choke (and hence their magnetic fields) would then be the same as if the amp were idling normally.
The hum you hear from the speaker would all be down to magnetic induction and you could test any screening methods.
I think there is a slight challenge to build the high powered and fairly high resistance loads, which might not be available ‘off-the-shelf’.

[stephenl]

The PT is a toroid. I bought a pretty thick/heavy steel toroid cover from Antek as a first
pass - it lowered the tank noise quite a bit.
I also added a steel cover to the open side of the tank.
The MuMetal was the final tweak.

[R.G.]

Toroids are good at having low external field if they are well made. Their construction forces the smallest practical air gaps, and a high air gap area running along the field lines of the core field, so there is both low field leakage and a wide area to conduct it, hence a low field intensity in the air gap and again less leakage. The total copper length is low, and the "window area" can be large, leading to low copper resistance. Leakage inductacnce, the lumped-constant measurement of leaked M-field is low.

But there are other issues. The close coupling of the copper windings and wide winding areas mean that capacitive coupling from primary to secondary is near-maximized. The super-continuous magnetic field circuit with nearly no air gap also means that tiny DC offsets push the core into saturation very easily. A volt or two of DC offset on the incoming AC power line can cause major issues with saturation in a toroid that was not conservatively designed. There are stories about a hifi enthusiast whose amps started producing major hum, but only at night. He had installed the "life extender" diode pellets in his outdoor garage lights and when they were on at night, the offset caused by the half-cycle loading of the lights let his hifi toroids walk into saturation.

If you were using a toroid, in a steel enclosure, then you were probably right to concentrate on the OT. The source field was made much smaller that way. I'm still quite interested to hear what an aluminum box around the toroid would do.

[stephenl]

Thanks for the education RG, much appreciated. In my case, I was shielding the output transducer on a reverb tank.

[dorrisant]

I've tried an aluminum enlosure around a toroidal PT... Only because the customer paid for the time. Non ferrous, no good. Tried mu metal... It cut out 70% maybe... Upsized the PT for more mA. Now that did the trick.

[R.G.]

Thanks for the education RG, much appreciated. In my case, I was shielding the output transducer on a reverb tank.

Doooh! I got myself going and didn't stay on the "reverb transducer" thread. I moved right into the PT vs OT channel. Sorry.

I've tried an aluminum enlosure around a toroidal PT... Only because the customer paid for the time. Non ferrous, no good. Tried mu metal... It cut out 70% maybe... Upsized the PT for more mA. Now that did the trick.

OK. I was interested in how well it worked for the real world. Good data point. Thanks.

Another thought hit me about reverb hum. Reverb is usually set up to not let too much bass get into the reverb tank, as it makes for muddy, flabby reverb. If there's a hum problem in reverb, even if it's from hum getting into the signal sent to the reverb tank input, you could do a high pass filter in the reverb return circuit to cut 60Hz by a huge amount. This isn't a good generic solution to hum, but it ought to work for reverb. In fact, I suspect the small input caps in reverb return amplifiers are for this purpose.

[TUBEDUDE]

As general advice with reverb tanks it's a good idea to inspect the tanks connectors for grounding. Reverb tanks ground the output, input, or both. Insure your design doesn't result in a ground loop, or an unreferenced signal. This isn't your issue, as the Mu metal had such an impact, but may help someone. Several sites have the Hammond guide showing the codes for specific tanks, if it's a Hammond.