Orange or120 glow screen grid

[binito]

I have orange or120 that blew a power valve. the problem is that to the maximum volume screen grids glowing red , one side more than another.
I've changed Sockets, grid resistors, screen resistor (was 1k8 5w), new valves and I have adjusted the bias to 20mA (cold). I replaced resistors and capacitors of bias.
The plate voltage is 540v, the screen voltage is 540v and bias - 51v.
Allthing Ok but When I am increasing the volume grids begin glowing red.( no plates)
I think 540v is too much for screen grid. I've uploaded the screen resistor of 1k (schematic) to 1k5 7w and it remains the same.
that may be happening?
Thanks in advance.

[Jazz P Bass]

What are you doing ,test wise, to make the screens glow?

Playing the guitar is not the same as running a steady sine wave through an amp.
A sine wave is hard as heck on an amp.

You may be 'overtesting' the amp.

[binito]

What are you doing ,test wise, to make the screens glow?

Playing the guitar is not the same as running a steady sine wave through an amp.
A sine wave is hard as heck on an amp.

You may be 'overtesting' the amp.

I test playing guitar

[Chuck H]

There should be at least a few volts of drop at the screens. If they are the same as the plates there may be something miswired on the high voltage rail. Can you offer any history about the amp or the problem? Did this amp work for you or someone else before? Is this amp new to you with no knowledge of it's history?

[Mick Bailey]

That's a very high screen voltage for modern EL34s. The B+ isn't getting loaded when you bias too cold and you probably aren't getting representative voltages. After checking the wiring, I would see what happens to the voltage when the current draw is increased to 35ma.

What is the voltage drop across the screen resistors (a) static, and (b) under load? Maximum screen dissipation for an EL34 is 8W, so you'd expect to see around 110v drop across a 1.5k resistor just before the screens start to glow. You can't pull screen current without voltage drop, unless the amp has been seriously miswired.

Also check the voltage selector is set to your mains supply. Seems obvious, but I get lots of amps where it's set wrong.

[J M Fahey]

Agree on 540V being incredibly rough on screens, doubly so with modern ones.

I would have no fear about limiting their current and wouldn't hesitate to go as high as 4K7 (try intermediate values first) as long as you can stop them from glowing, which is unhealthy to put it mildly.

To worsen things, a red plate is not *that* dangerous, because plates are large and are physically metal plates, a red or orange spot does not instantly compromise stuctural integrity, they have a larger mass, etc. ... but screens are wound of minimal fine wire.

Also vibration while red hot is to be considered.

I don't care much that "if Orange/Jim/Leo did things one way or another then why mess with it?" because we do not have the same tubes they did in their time.

Personally I would build a small 400V supply (with a high voltage MosFet á la Power Scale) just for the screens .

[binito]

That's a very high screen voltage for modern EL34s. The B+ isn't getting loaded when you bias too cold and you probably aren't getting representative voltages. After checking the wiring, I would see what happens to the voltage when the current draw is increased to 35ma.

What is the voltage drop across the screen resistors (a) static, and (b) under load? Maximum screen dissipation for an EL34 is 8W, so you'd expect to see around 110v drop across a 1.5k resistor just before the screens start to glow. You can't pull screen current without voltage drop, unless the amp has been seriously miswired.

Also check the voltage selector is set to your mains supply. Seems obvious, but I get lots of amps where it's set wrong.

I will try to increase the current to 35mA but think I will fry the valves (540 x 0, 035 = 18, 9W = 75% dis.)
as soon as I can I look in screen resistor voltage drops.
The solución is increase more the value of screen resistor? so I will have less power
transformer (looks like original) only has an input voltage and no wiring voltage selector.
Thanks in advance

[Mick Bailey]

Your calculation assumes the B+ will still be 540v @ 35mA - it should drop as you increase the current from 20mA.

In my opinion, increasing the value of the screen resistors is something to be done once the fault has been baselined, and only then if necessary to accommodate new-production tubes.

If the screens are glowing more on one side, this suggests to me it isn't a resistor problem. Are the tubes a matched set? Switch over the tubes from one side to the other - still glowing on the same side, or has the problem moved with the tubes?

[g1]

Are the filter caps for the screen supply stock values?

[binito]

Your calculation assumes the B+ will still be 540v @ 35mA - it should drop as you increase the current from 20mA.

In my opinion, increasing the value of the screen resistors is something to be done once the fault has been baselined, and only then if necessary to accommodate new-production tubes.

If the screens are glowing more on one side, this suggests to me it isn't a resistor problem. Are the tubes a matched set? Switch over the tubes from one side to the other - still glowing on the same side, or has the problem moved with the tubes?

Already I tried to invert the power tubes and the glow screens are maintained in the same sockets. After I have tried to reverse the outputs of c23 and C24 (output inverter tube to input grid power tubes) and the glow screens moved to the other side of power tubes. More signal from the cathode of V2B that of the plate. ( I have tested with several inverters tube).Attached schematic photo where I pointed out in yellow everything that I have changed.
Attached video where you see a screen glowing.

https://dl.dropboxusercontent.com/u/...0/IMG_2495.JPG
https://dl.dropboxusercontent.com/u/...0/IMG_2497.MOV
https://dl.dropboxusercontent.com/u/...0/IMG_2487.JPG
https://dl.dropboxusercontent.com/u/...0/IMG_2489.JPG

The power tubes are perfect matthed.

I adjust the bias to 34.8mA and glowing screen grid continues (More on one side than the other)
528V dc in plates
522V dc in screen (pin 4 power tubes)
5,72V dc across of screen resistor without signal
65 V dc aprox. across of screen resistor with signal and more volume.

Are the filter caps for the screen supply stock values?

The amplifier does not know its previous history. The filter capacitors are somewhat larger C5-C6 are 100uF (33uF schematic) and c3-c4 are 22uF (16uF schematic).

[g1]

Can you post the schematic you are using?
Is this an old unit or re-issue?
Does it have a choke?
What is your AC line voltage, and what do you measure for heater voltage?

[binito]

Can you post the schematic you are using?
Is this an old unit or re-issue?
Does it have a choke?
What is your AC line voltage, and what do you measure for heater voltage?

in the previous message attachment schematic that i used.
Is a old unit and yes it has choke.
The AC line is 230-240V AC.
After i will measure and i tell You .
Thanks

[Mick Bailey]

Dropping 65v across the 1.5k screen resistor at volume is a screen current draw of 4.3mA, or 2.8W dissipation - well within limits and not enough in itself to cause the screen to glow.

[binito]

Dropping 65v across the 1.5k screen resistor at volume is a screen current draw of 4.3mA, or 2.8W dissipation - well within limits and not enough in itself to cause the screen to glow.

as you do the calculation?

[g1]

schematic here: http://www.orangefieldguide.com/OFG_...hem_post74.jpg