odd? ground loop with the lid on, no loop when lid is off.

[pucho812]

ever see a rack mount unit be completely quiet until you put the lid on? unit is on the bench and completely quit until i put the lid on.  with the lid i get hum. the hum is pretty quit and only noticeable when there is no audio present.

I've been through the usual of rotate the toroidal and have done all the possible I can do to shield the mains transformer from the unit.  havint the psu transformer be external is not an option otherwise i would do that.
suggestions? 

 

[JohnRoberts]

Yes I've seen that phenomenon, if not that exact scenario. The internal magnetic field is distorted by any metal in the vicinity.

This changes the strength of the magnetic field that you are picking up... if you have a known loop, reduce the loop area of that loop.

Have you tried rotating the transformer with the lid on?

Add internal shielding to transformer (source) or shield the active circuitry (destination).

Not rocket science but it can be annoying.

JR

 

[pucho812]

I tried all kinds of things in rotating the transformer, shielding but was unable to get the hum out of you will.  which is why I prompted to post in the first place.  I ended up replacing the mains transformer and the hum went away.  can't see why that would cause change but it has done so.

 

[ruffrecords]

I'll bet the case is made of steel. I never use steel cases now simply because of all the trouble I had in the past with mains transformer hum being magnetically coupled through the case to sensitive mic transformers and such like. I recommend using all aluminium cases. They are more expensive but more than worth it for the reduction is aggravation.

Cheers

Ian

 

[SSLtech]

GAH! ...it's a per peeve of mine when people write "ground loop" or "ground noise" when they mean "hum".

But yes... if you put a transformer in a steel case, you often have field issues. One thing I've done before now is to build a dividing 'wall' 'isolating' the transformer or power supply section from any audio. -In fact even any DC power lines can also pick up modulation from current induced by the magnetic field.

I'd have to dig up a couple of old mic pre builds where I built the 'wall' owing to the case being steel. -I've used only alloy cases since then.

 

[JohnRoberts]

GAH! ...it's a per peeve of mine when people write "ground loop" or "ground noise" when they mean "hum".

Agreed, we share that peeve, but magnetic loops (inadvertent one-turn windings) are a common source of hum inside circuitry.

But yes... if you put a transformer in a steel case, you often have field issues. One thing I've done before now is to build a dividing 'wall' 'isolating' the transformer or power supply section from any audio. -In fact even any DC power lines can also pick up modulation from current induced by the magnetic field.

Yes, you must always loop at loop area when sussing out hum pickup... I used to like to keep split supply rail distribution as two parallel traces running together (often with a ground trace in the middle).

I'd have to dig up a couple of old mic pre builds where I built the 'wall' owing to the case being steel. -I've used only alloy cases since then.

When chasing hum out of low noise designs it is often a "raise the bridge" or "lower the water decision".  In other words address the magnetic field or reduce the tendency to convert that magnetic field to unwanted voltage corruption. Often doing some of both is advised.

JR

[edit]  while the cover can literally complete a magnetic loop, that should not be a concern since chassis should be on the sewer end of current distribution. Dominant mechanisms is generally altering the shape of the internal magnetic field.  [/edit]

 

[pucho812]

thanks for the tips guys.

 

[bruce0]

One other idea that wasn't mentioned.

PCB mount height off of the case is actually a substantial opportunity.  Especially if they have magnetically sensitive components (input trafo) on them. 

Since the coupling decreases as the square of the distance, and the distance is only a quarter inch, then coupling to the case can be reduced by raising the height of the PCB ( add even 20%, a couple of washers, to a 1/4 standoff and you get a 44% reduction in coupling).