[bikerhifinut]

I was wondering the very same Joe. I figured to lift the all the heaters by about 40V, the EL34 run at 30V on the cathodes, bang on the circuit spec. I feel that the LM338 regulator on the EF86 and ECC81 which has its negative end at signal earth potential should sort out any heater induced hum. I did get an awful hum on first power up which was quickly diagnosed as a broken wire to the earth refernce on the DC heaters, really hard to see as the wire had parted just at the solder joint on the tagboard. I think lifting the heaters on the power valvres may help also do you think the wire runs to the outside EL34 might be compromising things as they cross at right angles to the feedback components which i mounted on the spare tags on the EL34 cathode resistor/capacitor board. I could easily mount a small tag board nearer each preamp board and run a cntinuous length of screened cable back to there. I also wonder if the EL34 heater passing 1.5Amps and running very close to the output transformer primary taps is the culprit, inducing hum directly into the transformer. This would not be cancelled out by the push pull output pair. Thats an easy fix once i get in the garage and twist some more wire up. its the way the hum cancels out when a shortin plug is inserted in the input sockets that suggest a hum loop on the source but why that should also be on a simple unpowered pot in a box too beats me.
As you say I need to scope it and see what that throws up if anything. It is a fairly low level of hum and this is the perfectionist in me wanting the hum down below the noise of the valves and resistors. It's already reduced sensitivity/gain by triode wiring the EF86, theres no advantage in fiddling about with triodes there. I initially mulled over using a 12B4 there but decided against it in the end. Around half a volt for full output is in the ball park for me.
by the way Joe the slight mechanical noise from the toroid stops as soon as the amp is right way up and the transformer is supported be the top plate and not hanging from the bolt so to speak, I also think the metal cover was coupling noise into the table top like a soundboard.