Quad ESL 57: hum on one treble panel
 

Hello,
I have completed refurbishing a pair of ESL57, which means all panels except one: a treble panel, which appeared to be good and besides already refurbished (don't know when).
This treble panel makes some background noise, some kind of hum. If un-connected from power supply, this noise disappears. But output power reduces significantly almost immediately and more and more as time passes (timescale in seconds). I changed power supply (the full EHT block) but no improvement.
On the treble panel I refurbished, this is not the case: no hum and even after removing power supply connection, output power still stays high for a long time.
Any idea ? I have one but I would like to get confirmation.

I tried this: add a layer of insulating tape on the outside of the stators (I had to remove the dust cover for this). But I did not test yet if it improves behaviour.

Thank you.

Hi,

sounds like a seriously leaking panel, that draws too much current from the HV-supply. Try to measure the current or test on leakage with a blinker circuit.
If it prooves to be leaky, inspect the panel closely and try to find out if there´s some dust, humid grease or a particle at or within the panel that may cause the leakage.

jauu
Calvin

I think, it is very important to avoid leakage current not to destroy the HT-psu tranny.

My local dealer warned me, that he has seen a lot of burned trannies. His conclusion was, that the trannies are made for 220VAC and not 230VAC.

I think, this conclusion is not correct.

The tranny is marked with 0.5mA (at about 610VAC). And the fuse is a 100mA type.

So, every leakage current could imho be dangerous for the tranny, specially when feeded with 230VAC.

I strongly recommend the blinking neon bulbs to control leakage.

Franz

Any reference for this blinking neon bulb to suggest ?

Yes, here:

http://www.diyaudio.com/forums/plana...ml#post2343083

I used very "normal" NE-2 bulbs.

100V types without series resistor.

And I connected them simply in the HT-line. One for each bass panel and one for the treble panel.

Here you can see the bulbs:

http://image.n0t.de/f-fe1505495248bd...50c5028749.jpg

Franz

I removed dust covers to re-insulate the external side of the stators.
No improvement: the panel is certainly leaking. Knowing that there is no rivet at the peripheral (fixed by some strong tape) there remains one option from what I read on the net: coating is too conductive and might be in fact graphite based.
It looks like I have to re-build this panel completely. Anyway I will open it and see if I find some dust inside.

Hi,

the blinker consists of a small neon flash bulb, in parallel with a small film cap and a high ohmic resistor in connected in series to the bulb||cap.
The bulb may have a flashover treshold of less than 100V, the cap may be 100nF and >100V, the resistor may be 10MOhm. If the HV-supply already features a high ohmic resistor You may use that one instead.
Since the bulb is rather a shortcut when flashing You shouldn´t omit with the resistor in any case, because it reduces the current to a small and safe value and reduces the stress on the HV-supplies components. The small cap then supplies for the charging current through the bulb and it guarantees a clearly visible flash.

jauu
Calvin

I opened the panel. No obvious sign of dust or whatever could create some leakage.

The coating itself is rather dark; it looks like graphite. I measured its conductivity: 1e7 Ohm/square, which is much lower than what can be achieved with original coating, soluble nylon: from 2 x 10^9 to 1 x 10^12 ohms per unit square according to quadesl.org.

See here.

This panel is really noisy, in the sense this hum is audible at a distance of 1m. In my opinion this is due to the chosen coating.

Hi,

this could be the case indeed. AFAIR the caps in the HV-supply are of rather low value (<<100nF) and a low ohmic coating might be a too heavy load for this.

jauu
Calvin

The caps I use are 10nF (ceramic, 3kV).