Hum/Buzz When Power Transformer is Grounded to Chassis

[defaced]

Here's a fun one that I'm stumped on. I'm helping a guy on another forum debug a build. He's working on a Soldano XR88 preamp clone with an internally mounted power transformer. The build is exhibiting significant hum/buzz with a peak at 180Hz. Through much trial and error, he's figured out that if the power transformer mounting screws are disconnected from the chassis - laminations are floating - the 180Hz component of the hum/buzz disappears. If a jumper wire is connected between the laminations and the chassis, the hum/buzz reappears. I've had him go through the PT troubleshooting on Geo and everything with the PT looks good. I can't think of a reason why this could be caused. The two thoughts now from myself and the community are either it's caused by eddy currents, or by EMF. Eddy currents seem suspect to me because the condition can be recreated with a single wire between the laminations and the chassis, which means there's no loop for eddys to flow. EMF seems suspect to me because it appears to be a continuity problem, not a proximity problem. So I'm at a loss as to why this is happening or what can be done to fix the issue sort of a) isolate the transformer mounting from the chassis (I have some safety reservations about that), or b) new PT.

Anyone run into anything like this before?

Pics:
http://millerguitar.com/forumpics/X88R/X88R8.jpg
http://millerguitar.com/forumpics/X88R/X88R21.jpg

PT Datasheet:
http://www.hammondmfg.com/pdf/EDB270CX.pdf

Original description from the guy describing the conditions, in case my summary skipped over a detail I don't' know is important:

So...when you get to where you'll try anything...you try anything. I unbolted the trafo and rotated it 90 degrees and left it unbolted from the chassis. That nasty buzz wasn't there. I clipped a wire onto the chassis and onto the trafo and the buzz showed back up. If you'll remember, the buzz also appeared if the gain knob was turned down...this issue still exists but the constant buzzing with the gain up is gone. Any thoughts as to why the trafo grounding plays into this? Induction or something?

I've gone through the geofex article and the trafo seems ok. However I do have a question with regard to the greater than 1M resistance. In the instances where the article calls for resistance greater than 1M, my meter shows no continuity at all, which I assume is resistance greater than 1M. Am I correct in my thinking here. Probably a dumb question, but wanna make sure. If that is the case, the trafo is ok. Everything else checks out.

Is there a reason not to isolate the transformer from the chassis in this situation? if it shorts, wouldn't the fuse blow?

Hammond 270CX. I installed an IEC jack with a line filter built into it today...no change. I have rotated it every which way you can. There are two things that are impacting the noise.
1) Transformer bolted to the chassis...nice 180hz to 300hz buzz and a whole lot of other harmonics of that, and not much 60hz at all.
2)Transformer isolated from chassis. The 180hz to 300hz crap goes away, but then I get a wonderful 60hz hum that goes away if I touch the core of the transformer. I must note that the bells have been removed from the transformer. I'm not sure the impact this has though.

[dai h.]

(haven't taken the time to close examine the pics, but) shouldn't the pseudo heater center tap Rs be 100R0 (they are 100k BRN-BLK-YEL instead of BRN-BLK-BRN)? And wouldn't that be a hum source (sort of like no ground ref. since the Rs are too high)?

[Steve Conner]

Check the grounding. Input jack ring, chassis and mains plug ground prong should all show continuity to each other with an ohmmeter.

The PT has capacitance between windings and core, so the core leaks a few microamps of AC into whatever you bolt it to. A good ground system can absorb that no problem, but if the chassis was floating, a substantial AC voltage would appear on it which would then couple capacitively into the preamp.

Also check the heater centre tap resistors per Dai's suggestion. The heater windings pick up current capacitively too, and low value resistors are required to drain them to ground.

[Mick Bailey]

If the preamp is built to a tried and trusted design, and others have successfully built it, then there's a strong possibility that either;

1. A component is faulty
2. A component is the wrong value
3. The grounding scheme has a mistake somewhere.

A new Hammond transformer is probably the least likely suspect; repositioning and experimenting with isolating it is probably misleading in that you're directly linking cause and effect. If you turn the situation around you could also reason that because the noise disappears when the chassis is isolated from the transformer, everything else is faulty and the transformer is OK.

It also appears that the input isn't isolated from the chassis and this will often pick up hum. A preamp is going to be particularly sensitive to induced noise from any source. I always mount power transformers with three insulated bolts that don't touch the transformer, and a single grounded bolt that also carries the supply ground as well as the 1st star point for the main filter caps.

[Mick Bailey]

Just a further thought on this;

If you don't locate the problem, a workaround could be to isolate the transformer from the chassis, but ground it through a 15R or 22R 1w flameproof and pair of (say) 6A4 back-to-back diodes as per Fender/Marshall. It may work. you could test it just with the resistor - the diodes are just insurance.

There's a whole debate to be had about the safety aspect of grounding the laminations with this style of transformer if the mounting is isolated; they're insulated from each other and insulated via the bobbin and wire coating. Lots of domestic equipment isn't grounded at all, and most wall warts aren't. The resistor/diode combo would give reassurance - if it works in this case.