Hum in global feedback circuit?

[greiswig]

Hi, all,

Hoping one of you might have one of those "Oh, that's easy...check the _______" moments.

I just finished making a bunch of changes to my D'Lite, and everything seems to work fine except for the global feedback part of it. Voltages all look good everywhere. PI is a Bluesmaster style that worked fine before with a 100k feedback resistor. The presence pot/resistor/cap arrangement is the same one that Scott et al recommend as changed from the D'Lite style.

Here's what I changed, in addition to cleaning up the solder joints on :
- OT (used one of Brandon's)
- put a feedback resistor trimmer on the global circuit, which has a switch to completely disconnect it, or variable resistance from 10k - 500k.
- isolated the speaker jacks to allow a polarity switch

The hum is present when there is continuity between the 4 ohm tap and the board, whether the resistance is at the high end or low end of the trimmer. The trimmer does make a difference (high resistance = quieter), but not much.

If I turn the presence pot up, the highs taper off from the hum.

Bypassing the entire trimmer and using the fixed 100k resistor doesn't have any effect on the hum.

Could this be caused by the fact that the secondary of the OT is now isolated from ground/chassis? Or do I perhaps need to change the grounding scheme for my filter section because of the isolated speaker jacks?

Thanks in advance!

[ampdork]

Did you remake a ground for the OT secondary ground? If it's floating that would be my first guess.

[greiswig]

Did you remake a ground for the OT secondary ground? If it's floating that would be my first guess.

Okay, that's what I figured the problem must be...if I'm understanding you right. There isn't a ground for the OT secondary to the chassis right now.

When I isolated the speaker jacks, that got rid of the chassis ground connection for the black wire of the OT secondary. Is there a way to have a polarity switch for the speaker outs if the black wire always has to be connected to ground? I thought you could not have a polarity switch without disconnecting the ground from the chassis.

[heisthl]

With no ground on the OT secondary you have no global feedback. That is why you cannot easily reverse speaker polarity in an amp with global negative feedback. You have to do it external to the amp or if you insist on doing it internally your switching arrangement has to switch 3 things: the GNFB connection and both OT secondary leads to the jack (it must be grounded for GNFB to work). The real nightmare would be trying to do this for multiple OT secondary taps....

[greiswig]

You know, the older I get, the more common the slap myself in the forehead moments become. And then what one moment seems crystal clear is the next moment as clear as a Bush press conference. I must be missing something, but I'm having a hard time wrapping my head around this one...

If the phase switch is after the impedance selector, then shouldn't I be able to tie the common to ground at the entrance to the phase switch? When the phase switch is in the reverse position, the tip of the jack would become grounded, but since it is reversed, why should that matter?

Or maybe a different way to ask the question is this: why does it make any difference which end of the speaker wire the phase switch is on...the speaker end or the chassis end? In my own feeble brain, if it works at the speaker end it should be able to work at the chassis end.

Pardon the crude drawing, but why would something like this not work as long as the speaker jacks are isolated from the chassis?

[heisthl]

When you flip the wires outside the amp the relationship between the feedback line and ground doesn't change - in your drawing the feedback line would have to be connected to the Black OT wire when the switch is in reversed position. Even if you did this you would be changing the amount of feedback depending on which tap was selected.

[greiswig]

in your drawing the feedback line would have to be connected to the Black OT wire when the switch is in reversed position.

Heistl, I've learned as much from you as from anyone else on this board, so I have sincere respect for your opinion. Bear with this foolish viejo for a moment:

Why should the system care which side of the speaker the feedback is tapped from? To me, the fact that the jacks are completely isolated from the chassis and the only thing grounded is the OT secondary common means that is the effect of switching the phase.

[heisthl]

My bad - I didn't look at your drawing closely - you're right all you're doing is flipping the polarity after the chassis relationship has been established (just like if you swapped the connections on the speaker) so this should work just fine. My earlier post about the GNFB line needing a switch position was if you chose to leave the jack grounded at the chassis. The coffee has finally kicked in this morning...

[greiswig]

My bad - I didn't look at your drawing closely - you're right all you're doing is flipping the polarity after the chassis relationship has been established (just like if you swapped the connections on the speaker) so this should work just fine. My earlier post about the GNFB line needing a switch position was if you chose to leave the jack grounded at the chassis. The coffee has finally kicked in this morning...

Yup...proof is in the pudding. I was confident enough to try it even before I got your reply. Works fine. Changing the speaker phase doesn't seem to change the tone much, but it does change what notes feed back and how easy it is to get that to happen.

Thanks, all!