Magnetic shielding of trafo inside Leben amp

[ardilla]

Hi
 
I have a Leben CS300XS. It's tube push pull. There is a slight (but annoying) hum in the right channel with some high impedance headphones (sensitivity doesn't seem to matter). Leben tells me it is because the input trafo is too close the output trafo. This is only a problem with some headphones, and probably not with speakers. 
 

 
The input trafo is on the right. 
 
So I am thinking - how to fix this? 
 
My thought is maybea steel cage would shield the magnetic field from the input trafo? Is that a good iedea? What should I use- how thick must the steel be to have an effect?

 

[kevin gilmore]

what you want to do is cancel the magnetic leakage from the EI core.
you do this by wrapping the entire core with a wide strip of copper
sheet (about 1.5 inches wide in this case)  around the transformer.
And solder the ends of the sheet. This acts as a shorted turn
at the fringe of the magnetic field and is very effective.

 

[Armaegis]

Solder the ends of the sheet to make a closed loop, or solder to the case? 

 

[tomb]

Solder to make a closed loop, then ground it - hookup wire, spent lead, whatever.
 
Here's an example - on this one, the tape is not copper (the gray stuff), but it's the same effect.  There's a spent resistor lead in the solder joint (underneath) going to one of the pins that would be connected to a PCB/Chassis ground.

(pic from Edcor's website)

 

[Armaegis]

Thanks. The picture helps clear that up. On transformers with a bell cover should we wrap that copper underneath the bell? 

 

[tomb]

  Thanks. The picture helps clear that up. On transformers with a bell cover should we wrap that copper underneath the bell?

Yes, but I'm not sure - the bell may do the same thing as the tape.  Maybe KG can answer that one.  I'm positive it's not going to do any good outside the bell.

   The question is, does the tape do any good at all when there's an endbell?  I don't think so.

 

[jcx]

you can do some cheap diagnostics with even PC motherboard sound chipsets and free Visual Analyzer or similar sound card oscilloscope software - just need the connectors for noise (no signal) measurements, maybe some protective series R, attenuators, clamping diodes if you want to see waveforms
 
Hum is power line fundamental 50/60 Hz depending on country - often ground loops and xmfr leakage magnetic coupling
 
the "belly band" mod may help - additionally cheap low carbon steel is a good magnetic conductor - can be bent in a U shape placed to intercept mag field leakage from the power line transformer
 
 
 
Buzz is power line frequency harmonics 2x and higher integer multiples of your line frequency - from cap input ps rectified current spikes - can be magnetically coupled from poor wire routing or may be "ground contamination" from poor ground wiring topology
 
sometimes twisting the right pairs, rerouting wire runs can reduce hum/buzz from power supply currents

 

[ardilla]

  what you want to do is cancel the magnetic leakage from the EI core.
you do this by wrapping the entire core with a wide strip of copper
sheet (about 1.5 inches wide in this case)  around the transformer.
And solder the ends of the sheet. This acts as a shorted turn
at the fringe of the magnetic field and is very effective.

 
Thanks
 
1) Why 1.5 inches wide?
 
2) Wrap it which way - A, B, both or did I get nothing right?
 

 

[ardilla]

  you can do some cheap diagnostics with even PC motherboard sound chipsets and free Visual Analyzer or similar sound card oscilloscope software - just need the connectors for noise (no signal) measurements, maybe some protective series R, attenuators, clamping diodes if you want to see waveforms
 
Hum is power line fundamental 50/60 Hz depending on country - often ground loops and xmfr leakage magnetic coupling
 
the "belly band" mod may help - additionally cheap low carbon steel is a good magnetic conductor - can be bent in a U shape placed to intercept mag field leakage from the power line transformer
 
 
 
Buzz is power line frequency harmonics 2x and higher integer multiples of your line frequency - from cap input ps rectified current spikes - can be magnetically coupled from poor wire routing or may be "ground contamination" from poor ground wiring topology
 
sometimes twisting the right pairs, rerouting wire runs can reduce hum/buzz from power supply currents

 
Thanks
 
Pretty shure this is a 50Hz hum. How/where would you place that U shaped steel? Where can one get hold of such a think? 

 

[kevin gilmore]

B is the solution you want. Make a bit of overlap and solder
The entire strip.

If you can get mu metal then it goes as solution A.
But you are going to silver solder or arc weld the loop.

 

[ardilla]

B is the solution you want. Make a bit of overlap and solder
The entire strip.

If you can get mu metal then it goes as solution A.
But you are going to silver solder or arc weld the loop.

OK - Then I have to take out the transformer first, and wrap 1.5 inch of copper foil/strip around it and solder. Do I need silver for the solder? I am a solder novise, but might have handy helpers
 

 If you can get mu metal then it goes as solution A.

 
mu metal?

 

[Armaegis]

If you can get mu metal then it goes as solution A.
But you are going to silver solder or arc weld the loop.

 
Why would mu-metal warrant the different configuration?
 
 

  OK - Then I have to take out the transformer first, and wrap 1.5 inch of copper foil/strip around it and solder. Do I need silver for the solder? I am a solder novise, but might have handy helpers
 
 
mu metal?

 
http://en.wikipedia.org/wiki/Mu-metal

 

[kevin gilmore]

The shorted copper turn reduces the stray flux and is in the same
direction as the windings.
 
A mu metal box (well 5 sides of it anyway) prevents leakage out
of the E/I laminations.
 
Some people do both. The second one is much harder.
 
yes you have to take the transformer out to wrap it in the
copper. Any kind of solder will do. What you want is a single
low impedance shorted turn.

 

[Steve Eddy]

If you can get mu metal then it goes as solution A.
But you are going to silver solder or arc weld the loop.

Mu metal can saturate at pretty low levels. Don't know that I'd use it for shielding a power transformer. At least not without some steel underneath it.

se

 

[jcx]

don't destroy the xfmr's bobin, insulation soldering the belly band - may even want a wrap of kapton tape to maintain creepage/clearance distances if it is a pcb pin mount style
 
 
mu metal is not needed/useful anywhere near a line power transformer - mu metal saturates easily - only used to case very sensitive components - knock the mag field down by shorting with cheap iron 1st
 
while a box with more sides is better for shorting the leakage mag field - just a strap bent in a U as much taller than the xfmr as fits should help a lot