I am delving into the bowels of my Paton Valve tester; early model .... no Vibrator.

I have isolated the "Line Check" circuitry because I am not getting full deflection of the Meter. There is not much to this circuit ....
secondary coil is providing the correct 300vac as measured with Multimeter when Line Selector switch corresponds to our local ac mains voltage .....
6X5 rectifier ....
a smallish capacitor which is not enough to smooth out all the ac post rectifier valve ....
and a couple of resistors.
Components have been changed except for a wire wound resistor .... but it measures okay; values align with circuit diagrams etc..

I am measuring voltage across a couple of the resisitors at the 'take-off' points to a current limiting resistor and Meter. These voltage measurements are proportional to Meter deflection.

In search for a fault, I have noticed that tapping on the 6X5 glass envelop 'encourages' the voltage to rise. This is not a sudden increase as one would expect with a short circuit .... pins are clean ... etc..

Continual, gentle tapping on the glass has caused the voltage to rise in stages from approx 65 volts to about 170volts ... holding 165volts at the moment ... which is equivalent to a meter deflection increase of 1/3 to 3/4.

Am I somehow 'rejuvenating' this valve? Is there a better method?

Despite being allegedly being made for car radios, 6X5 in some mechanical manifestations, has proven to be an extremely unreliable valve, notorious for internal shorting. The fact that it may be faulty has an extremely high probability. The best solution if you cannot test it in something is replace. If it is gassy it will often have a pastel blue halo around the plates (Valve is no Angel), or the Getter will have gone from silver to a bright white.

VCT later types only use half of it. The early ones like mine (1938) had a 1V (UX-4), that could fall out. Jumping voltage could be a bad connection, or a short after a resistor: Do check. If that block cap comes out of circuit the volts will also fall.

In mine I suspect that it is paper in oil (PCB) and 2μF. I leakage tested the one in mine at rated voltage & it was OK so I left it.

There are other Wax paper caps in them: But not after you fix it & get rid of them. They leak & they upset things.There is no place for a leaking non polarised cap, in a test instrument.

The author of the site linked to below is quite a fan of the 6X5. That aside, he provides some useful photos of various plate design changes to the 6X5 over the years:

http://vinylsavor.blogspot.com.au/2014/01/tube-of-month-6x5.html

The earlier 6X5G (shouldered envelope) seems to have a worse reputation that its GT cousin, although the G version also had offset style plates:

http://www.r-type.org/pics/aaa0807a.jpg

http://www.radioold.com/images/6x5g_.jpg

Hi Marc,

Thanks for the response ...

No blue halo ...... no getter bright white .... All joints resoldered ... new cap in place of 2μF can. No amount of tapping anywhere other than the glass envelop produces a voltage change. More 'tapping' has produced 204volts, albeit temporarily, holding 190 volts.

Strange .... but must be some explanation .... I'm wondering which 'tap' will be the last ....

The result of your tapping does seem to suggest some kind of an internal short. Some Americans warn that 6X5s have been known to go short slowly, taking out the power transformer by stealth, so to speak.

I haven't experienced any such problems with 6X5GTs myself.

Hi GTC,

Mine is RCA .... Primary side fused which may provide protection if it acts quickly enough. Will be interesting to see what voltages are when it is restarted from cold.

Now if you tap the valve and the socket is not dodgy, or there is a dry joint on the socket or the valves pins (has happened), some wires actually corrode off, or break off when the base & envelope come loose.

If its falling apart inside and the heater hits the cathode, things could get expensive. It may be moving?

That over & under design is one I have seen do the heater cathode short trick. Zenith US sets had a bit of trouble with them as a lot of their stuff seemed to sail with the wind.

AWA was one that had 100 ohm resistors in the plate circuit of those & the wattage was kept down. It was not only to reduce surge, they fused before the transformer. Never replace with a bigger wattage.

I've found the 6X5 to be mostly reliable..

My only example of a 6X5G has a heater that acts more like a flashing christmas light, it starts to heat up, then goes off, then the cycle repeats. Unfortunately it can never produce any useful output.

The only failures of 6X5GT are from shorted input caps, which causes them to die from the usual blue flash which cuts the cathode in half, or can short out one of the diodes.

For the OP's problem, the only thing I can think of is a bad joint inside one of the pins. Obviously it would be quicker to just get another replacement valve instead.

If the tube emission gradually increases (making for increases of voltage in your circuit) after you tap it, it might be a poor heater connection.

Thanks for all the responses ......

This morning the voltage at Meter 'take-off' point is slowly fluctuating between 190 and 195 volts ..... giving 80+% needle deflection at my local mains volts. It is easy to increase this, but best to leave as is until I sort out its other problems. Do I get 'ex-communicated' if I convert it to solid state ..... ????

I am leaning towards a heater problem also as my first impression this morning was that the filament glow is easier to see under my normal lighting conditions. Subjective ... and unfortunately, no problem is evident.

I've had the heater from a 6X5GT spring out of the top of the cathode and bend over like an umbrella handle in one example. Apart from that no issues though most of my radios are fitted with either an 80 or 5Y3G/GT.

Do I get 'ex-communicated' if I convert it to solid state ..... ????

Not by me, however there are what I regard as good and bad ways of doing that. The good way puts a solid state solution into the socket thereby allowing retrofitting of a 6X5.

Alternatively, you could simply try a replacement 6X5.

I agree with GTC's second option. Where at all possible, valve radios should never be altered from their original specification. When valve radios are modified their market value and historic value decreases markedly. The same applies to vintage test equipment.

Installing a new or used, known working, valve is the best option.

There actually has to be a risk assessment, when changing from a Selenium, or a valve rectifier, silly as that may sound.

There firstly is the risk of adding a fault.

Silicon diodes are more efficient than valves. The tappings on the transformer may be enough to compensate for this. In a radio you need more resistance and have to know the current draw of the valves. You do not get that luxury with this. So you have to rely on the transformer.

Where it goes pear shaped in a radio, is that the Silicon diode, behaves like a filament rectifier without a cathode sleeve. If there is no substantial current draw (cold heaters) the "B" voltage can rise to twice the Voltage of the loaded "B" rail, prior to the heaters bringing the valves into conduction. The components have to be able to withstand that surge.

You are likely best to use a 1N4007. But I would prefer a 6X5 that is not an over & under style.