UL84 all hot & bothered?

[egerton]

Seeking advice on the o/p stage of my Philips G85U (see attached) where V5 UL84 is radiating a large amount of heat and making a slight mechanical rattling sound. The valve museum states that it has a 100mA heater and is designed to operate at lower anode voltages and increased anode current - so perhaps it will be unusually hot, but I was shocked at how hot it is and a momentary touch burnt my finger! The set is working ok and I've taken some measurements of V5 :-
g: 0.023 k: 14.5 a: 224 s: 209 (all w.r.t to chassis)
g to k: -11.6V
I'm getting 241v on HT rail and compared to trader figures I'm running high on anode volts by about 20v is this a concern?? With the 14.5v across R31 gives about 48mA cathode current. Any advice would be greatly appreciated.

[Station X]

A UL84 should be good for 75mA of combined anode and screen current. Have you checked (by measuring) that R31 is the correct value and doesn't have a leaky capacitor in parallel with it?

[Herald1360]

If you've got 11.6v g-k and a cathode voltage of 14.5v you have a leak somewhere in the UL84 grid circuit. Coupling cap is prime suspect, but it could be the valve itself or its holder.

Do Station X's checks as well......

They do get hot as all (most) small output valves do.

[Techman]

I thought he stated -11.6v ?

Those valves do run hot - fancy touching it with your finger!

You SHOULD expect a nasty burn!

I doubt you'll touch an output valve again after that lesson - lucky you didn't crack the glass with the dampness from your finger - fsssst......crack!

[paulsherwin]

Yes, they do run hot in normal operation so that in itself isn't a cause for concern. They sometimes rattle a bit too. Check the other things as advised though, particularly the UL84 grid coupling cap.

The voltages do look reasonable. You shouldn't expect them to be exactly the same as the service sheet values, as mains variation, component tolerances and component aging will all affect them.

[G8HQP Dave]

A UL84 should be good for 75mA of combined anode and screen current.

Yes, but only with an anode/screen voltage of 170V. At 240V you would need to keep anode current down to 50mA.

If the valve is unusually hot then double-check anode current.

[G6Tanuki]

Yes they do run a bit hot.

Check the voltage across R28: if you see anything, you have a problem, either a leaky C49, R28 gone way-high, or the valve is leaky internally.

Also check R31 and C55: If C55's gone leaky or R31 has gone low it will increase dissipation.

In the past I've been known to significantly increase the value of cathode-resistors to reduce the amount of heat dissipation in these output stages and give the valve an easier life (more important with the likes of the EL/UL41 which is known for not liking to be sweated too hard).

[ms660]

If everything was healthy the voltages reported wouldn't compute for max or sub max Pa dissipation. Maybe low emission plus leak to g1 (external/internal)

Lawrence.

[paulsherwin]

If the valve is getting tired that's not a reason in itself to replace it though. They can soldier on for years like that.

[Biggles]

You soon learn which valves run hot. Once bitten twice shy. Saying that, I still get the odd burn, but normally from an unintentional brush with one, or not waiting long enough after switching off for the things to cool down. As a VERY general rule of thumb, signal valves hot, rectifiers and output valves, skin removingly hot! I dare say most of us learned the hard way.
Alan.

[egerton]

It was just a bit painful even to touch momentarily whereas all others including the rectifier were not that hot so it surprised me. When I lent over the set the heat from above the output valve was significant so that should have been my warning - just I couldn't believe it was running that hot. The heater is lit up in the core like orangey red bar much more than the other valves so I guess this a big part of the heat. The valve has quite a lot of "staining" - dark brown marks in a ring around the top, is that a sign of anything?

[flyingtech55]

My advice for what its worth is to wait until it's dark outside (or as dark as it gets at the moment), close the curtains and view the valve with the back of the set off and with the lights out. If the anode is glowing a dull red then it needs investigating. If the anode isn't glowing then probably all is well.


What you can also do is disconnect C49 and see if the valve runs any cooler. Obviously there will be no sound but it is a useful test.


HTH


Tim

[Herald1360]

Originally Posted by Techman

I thought he stated -11.6v ?

Those valves do run hot - fancy touching it with your finger!

You SHOULD expect a nasty burn!

I doubt you'll touch an output valve again after that lesson - lucky you didn't crack the glass with the dampness from your finger - fsssst......crack!

Yes. But Vgk = -11.6v and Vk = +14.5v equals about +3v on the grid ref chassis......

[Chris55000]

Hi!

AC mains–supply voltages have been creeping up a bit over the last few years, measurements on my local supply have consistently been 245–255V for at least the last five years, which will give 250–260V rectified DC from a UY85 in good condition with a 50uF reservoir capacitor.

What I recommend you do is to reduce the smoothed h.t. for the screen–grid of the UL84 output valve and the main h.t. line to about 205V (to chassis) by adjustments to the main smoothing resistor – this will give an effective UL84 G2 supply of about 190V, which is about right for the cathode–bias resistor you have!

An example UL84 circuit from a UL84 record–player, with component–values and working voltages, to use as a guide, can be downloaded from the American Radio History website – Download the Jan 1965 issue of "Radio Constructor" and turn to Dick & Smithy's "In Your Workshop".

In that article, Va is given as 225V, Vg2 as 205V and Vk as 18V, which are the figures you should aim for – most manufacturers tended to run these valves on the conservative side (8.0 to 8.5V GB for an EL84, 17.5–18.5V for a UL84, to keep heat down and economise on h.t. current – my experience has been that there is still more than adequate volume/power output under these lower–running conditions.

Note, because of the relative "constant–current" nature of a pentode with respect to it's anode voltage, 20–30V more on the anode than what the service–sheet says doesn't matter too much, it's the G2 voltage and cathode–resistor value (and of course "that!!!" cap when there's a previous R.C. coupled stage!) that's important!

Chris Williams

[julie_m]

Some quick back-of-the-cigarette-packet calculations:

The heater of a UL84 drops 45V at 0.1A, so it is dissipating 4.5 watts. A cathode current of 0.048A at 224V gives another 10.8W. Some of the cathode current will be coming via G2, but the voltage is close enough (209V as opposed to 224V) not to make much difference. There will still be a bit more than 15 watts of heat coming out of that valve.

The UY85 will be the next hottest (3.8W for the heater, plus the HT current times the forward voltage drop) and the small signal valves will run the coolest, with the majority of their dissipation from their heaters.

[Sideband]

Most output valves are too hot to touch but you must check the points already mentioned just to make sure it isn't too hot.

[egerton]

Out of interest I measured the temp. of UL84 using a thermocouple and the hottest spot I could find was about midway up the valve at around 160 C. I then measured the EL84 in my Ferranti and it is at 140C but that has a metal retaining clip that may act as a bit of a heat sink.

[paulsherwin]

Those temperatures are plausible. The exact temperature will depend on both the valve construction and operating conditions.

A simple way to reduce heat dissipation in the output valve is to increase the value of the cathode resistor somewhat. This will reduce the maximum undistorted volume but that won't be noticeable in normal use. The valve will run cooler and last longer. Make sure you have identified and corrected any faults first though.