Shed some light for me

[iknowjohnny]

Heres an odd one for you, and if possible i'd like to hear the theory behind it. When i built my 1st amp, a 18 watt, i was told the OT ground should not be grounded but should just go to the ground tab on an isolated output jack. So i did that. Also did it on the amps i built since, tho on at least one i eventually grounded it tho i can't recall why. But maybe because of the same reason i grounded the OT wire on my new amp yesterday. heres what happened. i did it the same way the 18 watt guys said and everything was fine. Amp was quiet and sounded fine. But then i had to pull the board to get to something and after i did i was getting output hum. I figured i move some wire and needed to do some chopsticking. But no amount of wire manipulation seemed to stop it, tho it did get much better and seemed to go away completely after a short time. So i didn't think much about it after that till i had to lift the board again, and this time it began to hum horribly. And also this time no amount of wire manipulation helped at all. then i had myn hand on the speaker plug and at the same time touched the chassis at the front and the hum subsided about 25%. Thats when it hit me that the OT ground must need to be grounded to the chassis, and when i did the hum was totally gone. So now i'm curious why this happened and also what the theory is behind leaving the OT ground at the ground tab of an isolated jack only or grounding it to chassis. Man, electronics are bizarre sometimes !

[Andy Le Blanc]

There's two kinds of coupling that happen between circuits components
that are not physically connected. There's inductive coupling and capacitive coupling.
For what ever reason you setting up the conditions for some sort of coupling.

Providing a ground reference alleviates the condition.

If the amp has feedback the only ground reference your provided to the OPT sec.,
might have been left to where ever the global arrangement was tied into
circuit, floating the whole FB wire, making the sec. and load stand on one leg.

Bad juju. Just a theory.

[iknowjohnny]

Whats so bizzare tho is that no amount od wire manipulation changed it a bit the second time it happened till i grounded the OT wire to chassis. just flat out weird...

[Andy Le Blanc]

I had a similar issue but a different cause in one amp.
It developed a coupling issue around the opt and plate leads.
I finally had to shield with metallic tape, the symptom became worse
over time.

I think there might be a transformer core magnetization issue.
but its a guess, funny where it pops up. Capacitive coupling gets critical
in high impedance/high gain stage, wire type/proximity issues, chop stick, etc
Capacitive issues are very present just with general considerations.

you might be getting it from a loose winding, the opt might have been layered
with one to many layers of tape. Could be that flexing the wire dress made
the insulation loose and created a general capacitance that contributed to the condition.

but you found a fix, I'm surprised that the opt wasn't ground referenced
at least in some way, If its an aluminum chassis the different metals might
be reacting and the opt might be floating on a resistance above ground potential.

[dave g]

It's not coupling. It's a ground loop.

The reason you don't ground the OT secondary just wherever you want is because you have several amps flowing through that winding.

Courtesy of Aiken:

The amount of current flowing in the secondary is huge compared to the signal currents in the rest of the amp, so the utmost care should be taken here. The secondary current goes only to the speaker; it isn't used anywhere else in the amp, unless there is a negative feedback loop, in which case a small portion of the secondary voltage is fed back, usually through a large resistor, so the current in that path is small. In either case, feedback or not, the secondary should be wired directly to the output jack. Do not ground the output transformer common to the chassis and then ground the output jack to the chassis. This will create a heavy ground current path through the chassis, which may run through a preamp section, depending upon the location of the output jacks and the output transformer. Do not ground the output jacks to the chassis at all, they should be isolated from the chassis. Also, do not route these output transformer secondary wires anywhere near the preamp stages, they should be routed as far away as possible, around the edges of the chassis to the output jacks.

[Andy Le Blanc]

but the issue described is the opposite, I haven't been inside an amp made
in the last 50 years that didn't ground the sec..

This one will be fun..... How do we test either approaches ?

[iknowjohnny]

Oh geez.....well if thats the case my new amp is trash because i tried everything imaginable and the hum would not go away. And i took great care when laying it out too !

I'd like to see a poll.....who grounds the OT's secondary common and who sends it only to a isolated speaker jack?

[KellyBass]

cast my vote for grounding the sec.

But I'm a novice hack and just do what everyone else says to do.

Great subject, BTW!

[tubeswell]

I'd like to see a poll.....who grounds the OT's secondary common and who sends it only to a isolated speaker jack?

FWIW I always ground the OT sec to the speaker jack ground (not isolated), and I never have any hum problems that way.

[Structo]

Yes, I'm guilty of just running the ground wire to the jack, non isolated.

[Jana]

You have to ground the OT secondary someplace, especially if you have a NFB network. Sometimes you can get away with using grounded output jacks. Sometimes not. I use isolated output jacks and bring a ground wire from the output jacks and tie it to the same grounding point that the PI is grounding.

In any case the OT secondary should be referenced to ground.

[Phil_S]

I have run the secondary ground directly to the HT ground point, and I have run it to the jack and then to the HT ground. I've always chasis grounded it. By the HT ground, I mean where the HT CT, power tube cathodes, and main filter caps are grounded. It always seems to work. I use isolated jacks. No hum. <shrug>

Aiken makes an interesting point. I'm going to try it his way next time.

[iknowjohnny]

Ok, so everyone grounds it with no problems such as aiken describes. Wonder what his deal is with that article ?

[dave g]

Ok, so everyone grounds it with no problems such as aiken describes. Wonder what his deal is with that article ?

I didn't say you don't ground it. I said you just don't ground it wherever you want. As in, don't put that ground anywhere near the preamp.

[Cygnus X1]

I isolate it.
No grounding.

No NFB either.