Reverb Tanks -insulated jacks?

[Nathan]

Hi all, I have a 8FB3C1C which has connectors "input insulated, output grounded". Can anyone explain to me what that means? Do I need to do anything differently?
Thanks!

[tubeswell]

all it means is that the pan input is insulated from the pan chassis, and the pan output is grounded to the pan chassis (so you don't get a ground loop when you connect the two reverb cables).

[Nathan]

Ok cool, so there's no concern about how I wire my jacks on the chassis?

[tubeswell]

Ok cool, so there's no concern about how I wire my jacks on the chassis?

All logic screams at me that it 'shouldn't make any difference' because if the jack sleeves (on the chassis jacks) are bolted directly to the chassis, they are getting their potential from that point only. However the 'verb circuit is particularly sensitive to noise, and the cable shields are theoretically a potential source of unwanted capacitative EMF coupling (if the microwobble in the ground return path at that point is strong enough to capacitatively couple into the cable signal wire).  But this for this to occur it would require a pretty strong (micro)voltage wobble in the ground return.

In this regard I think the chassis material influences the outcome. I recently did a BFPR clone type build with a steel chassis, where I couldn't get the noise floor in the 'verb low enough without isolating the sleeves and running separate ground returns for them.  Whereas with other amps with verb, I have used ally chassis and it doesn't seem to make any difference (whether the jacks are isolated from the chassis with separate ground returns or not).  I guess it is possible for microvoltage wobbles to be higher within different points in the ground return plane of a steel chassis, because steel has higher resistivity than ally (but as I said earlier, they would have to be fairly strong microwobbles in order to affect the signal).  If anyone else has a view on this I'd be keen to hear about it.

[tubesornothing]

I would treat it like an other sensitive sheilding situation.  Only one end gets grounded to avoid loops. Attach the ground to the sensitive side (the output of the tank).

Now upon looking at the tanks I purchase, they just so happen to be:  output grounded, input insulated - so there ya go!

[PRR]

The case has to be grounded if there is AC power anywhere in the area (of course there usually is).

If both ends run to the same box (the usual way), then tying both connectors to ground is not necessary and invites ground loops.

Given a choice of input or output, we usually prefer the one with the lowest signal.

Floating-in bonded-out is usually best.

Other schemes usually work in most cases. If not, you can usually snip something.