Phase splitters and hum

[Refugee]

I am shortly going to do some trials with the single triode phase splitter in an amplifier.

The problem is that it produces a bit of background hum that disappears if the HT to the phase splitter is disconnected. The other section of the ECC33 is used as a voltage amplifier with the wiper of the volume pot on the grid. The hum remains with the volume turned down.
The HT is about 380V and the valve runs with a bit over 100V on the cathode in order to get the second output grid drive.
This looks like the source of the hum as after looking at the data sheet for the ECC33 I have spotted that the maximum rating for heater cathode volts is only 100V.
My conclusion is that they were pushing it a bit with the original design.
I have done the kitting for a plug in mod and will shortly start testing.
Testing will start with an NOS 6/30L2 that I already have as the data sheet shows it as having a maximum heater cathode voltage of 150V.
If I see an improvement it is most likely that I will buy some of those RS ECC83s that are on sale at a reasonable price.
I will then roll the mod out over several amplifiers if I am successful.
Has anyone else had this problem?

[pwdrive]

Any conclusion yet?

Lawrence.

[Refugee]

I will start testing shortly.
I have a big scope on the bench with one more minor problem to sort out and then I will be starting work on this one.

[Alan Douglas]

Biasing the heater positive with respect to the cathode will reduce some hum. With the cathode sleeve positive, you have essentially a diode and some conduction since the heater insulation isn't perfect. Depending on what else is fed from the heater winding, you might rig a voltage divider across the HT and connect the heaters there.

[Refugee]

It will not work with the heaters biased positive as the other triode section is a conventional voltage amplifier. This actually is of benefit as the overall gain of the complete amplifier is enough to use an MP3 player without an additional external gain stage.
As I don't want to modify the original 1950s chassis I have gathered all the parts to make an adapter.

[Alan Douglas]

You could still bias the heater at +100V? Even that will reduce the conduction through the "diode" as there will be no anode voltage.

[Refugee]

Putting 100V on the heater would make the cathode of the other section strongly negative and the data sheet shows this to be very bad.

[Alan Douglas]

Well if the data sheet says so...but negative with respect to the heater rather than positive would seem to reduce leakage currents. It could no longer act as a thermionic diode.
Easy enough to try: a 250k pot across the HT with the wiper tied to the heater winding; move the pot off ground and see if the hum level changes.

[Refugee]

Biasing the heater will surely only work if single triodes are used but with the other triode section running as a conventional phase splitter with the cathode only a few volts above chassis it would be impractical.

[Alan Douglas]

If the other triode heater is biased at +100V it's still within its heater-to-cathode rating. I don't see what's "impractical."