• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Isolation Transformer: How to decouple chassis from circuit?

I'm installing an isolation transformer in an old pepco tube amp (this one's labeled Marlin), and i have a question about general approach.

it's my understanding that i want to ground the amp's chassis using only the ground connection on the 3-prong side of the isolation transformer (IX), so it makes sense to me that i no longer want the amplifier circuit to be connected to the chassis anywhere. everything that was previously connected as a ground or neutral plane should be together, but floating from the chassis.

thankfully, there were only a couple of components that were visibly soldered to the chassis, so i used a spare tab on the terminal strip to connect them together.

now i've realized that the 4 inputs are still grounded through the chassis, but if the chassis is only connected to the ground on the 3-prong side of the IX, they won't be connected to the ground plane of the rest of the circuit. i can't find any mention of this separation in the discussions i've read online.

is it ok to have the chassis connected to the neutral on the 2-prong side of the IX as well as the ground on the 3-prong side? do i need to isolate the inputs from the chassis, and run separate grounds to the neutral on the 2-prong side?
The amp is a chassis 801, so identical to the circuit here:


what i've done is try to take the chassis out of the circuit by removing the 2 grounds (both labelled 39v) from the chassis, and connecting them together. what's omitted from that drawing is the ground sleeve on the 3 inputs, which are grounded through the chassis, just by contact.

i'm thinking i need to replace the input sockets with isolated sockets, and them run ground wires to the same place i connected the other 2 grounds. that way the entire circuit is isolated from the chassis, and then i can ground the chassis by itself. am i overthinking this?


Joined 2003
Paid Member
That is a very tangled drawing of a very troubled amplifier.

They "separated" 1st-stage common from amplifier common to "reduce shock hazard". This is an intermediate step from true Death Traps to the later fully isolated cheapo amps. (Was this form actually sold in Canada? Or did it leak over the border?)

It needs some significant re-wiring to make a happy little SAFE amp. All the internal commons tied to each other (losing the R-C network which does not even look safe). The chassis ties to plug Green. Wall-power white and black ONLY to isolation transformer.


  • milothicus-42.gif
    50.2 KB · Views: 442
Thanks, all.

PRR, That's exactly my plan.

What I've found strange is that in all of the discussions of these amps that I've seen, there's lots of discussion about what the connections at the transformer, but not the other connections, like the chassis.

It made me wonder if for some reason it's ok to have the ground connected to the circuits on both sides of the transformer. It didn't seem right, so I decided to ask for clarification.
So I installed the transformer, and now the amp just makes noise...

The transformer seems to be stepping up the voltage rather than just isolating and I wonder if the higher voltage on the heaters is the issue.

My understanding is that the heater circuit has a large resistor in series so that the circuit voltage drop matches the supply, so could I add more resistance to account for the new higher supply voltage? I could calculate the resistance I need, but I don't know how much current is running through the circuit.

I have some high wattage resistors here somewhere...

Any suggestions?
Stepping up the voltage? Make sure it's an isolation transformer. A lot of "transformers" for sale are actually autoformers. You should have no continuity between input and output if it's truly an isolating transformer, and it should give you 120V to 120V out. If it's stepping it up to 240V you will blow the amp up if you leave it like that.
I assume to have a DMM?

That sounds a ground loop though if it's 120Hz and varies with volume.
Last edited:
I installed a Triad N-68X isolation transformer.

I measure around 125VAC on the primary side, and the secondary side is around 135VAC.

i'm focused on the change in voltage because the amp worked well with its old 2-prong power cord, and I changed so little of the circuitry when i installed the transformer. i disconnected 3 of the inputs completely, and installed one new input jack that is now isolated from the chassis, and i disconnected a couple of other components that were grounded through the chassis, but are now just connected to each other, and to the rest of the ground plane.

i've now connected the chassis to the grounded third prong of the new power cable, but nothing is grounded through the chassis anymore, so there shouldn't be a loop there... maybe i need to recheck my work.
Did you measure 135 with the load connected or unloaded? It's normal for the voltage to be higher without a load.

Connect the chassis and house ground to the signal ground at the jack and see if the hum goes away?

If you short the input jack the hum goes away, yes?
Ok, shorting the input silenced the hum.

i mistakenly first shorted the input to the chassis/house ground, and that resulted in some strange oscillation. i could change the speed of the clicking with the volume dial.

then i reread your message, and grounded the input ground to the chassis, and it went quiet... and now after removing the jumper, the hum that was there originally is gone, regardless of volume level. could a static build-up in the floating circuit do that?

i haven't connected a guitar yet. i'll try that this evening.


Joined 2003
Paid Member
.... it should give you 120V to 120V ...

As you go on to discuss: The "120:120" is typically at FULL load, like most transformer ratings. All transformers sag; conversely they "un-sag" at zero load. This is typically less for large iron and more for small iron. The N-68X actually has a number: "15%". So their "115:115" full-load becomes 132.2V at no load (138V for 120V input; 143V at my 125V house).

"Over 140V no-load" does suggest a rather high wall voltage. Like mine. (I sag to 108V when laundry runs.)

"Full load" is 50VA, or >0.4A. The amplifier is 0.15A AC plus 0.05A DC, or about 0.23A equivalent AC. Just over half rated power. So we expect just about halfway from 132 to 115 (143 to 125). Say 123V (134V). It seems to fit milothicus' observations.


Joined 2003
Paid Member
> connect the house ground to the floating side of the isolation transformer.

Correct for any modern amp.

Keep a clear distinction between "Ground" and "Neutral"!!

In the attached snip-shots you see that a 1964 Champ ha NO ground pin (cuz most outlets had no ground). However even by 1990 most outlets had the 3rd hole and most-all stage amplifiers had the Ground pin tied to chassis AND to the circuit common (the part of the jack-cord you touch).


  • Amp-ground-42.gif
    18.7 KB · Views: 106
Just a thought, measure DCR of primary and secondary.
Also, depending on which country you live in, the earthing arrangement can be different. For example in Oz, neutral is 'usually' taken from the star point of the wye (Y) which is distribution side of the delta (triangle) transmission via transformer, ie isolated, and connected direct to the 1.2M electrode inserted 1.1M into hard 'physical earth/ground' mains earth at the switchboard. I've worked on some electrical installations in the US (Nth California) and its clearly different, in that they don't use a mains earth at all, just active and neutral. Also in rural Oz they can use SWER which is just about the opposite, single wire active (one conductor) and use earth as the return. Class A Lic PGE 209366.
Last edited: