VR92 vs 6AL5W

[David Pannell]

I believe that some of the later VCM instruments used the 6AL5W military grade rectifying valve, presumably for deriving the negative grid volts supply.

As the VR92 is almost impossible to obtain, I was wondering if I might try out a 6AL5W, with possible adjustment, to see if it works on the Mk2. Heater current is 300mA as opposed to 150mA, but the transformer windings look more than man enough for the job. It might even be feasible to parallel up the anodes and cathodes.

Any thoughts, or am I ?

David.

[Alf]

I wouldn't have thought they were impossible to obtain, EA50 is an equivalent, there's currently four VR92s on ebay

[David Pannell]

Hi Alf,

Isn't it strange! Although I have 3 spare VR92s, I was worried that I hadn't seen any advertised for ages; then as soon as I mention it on the Forum, 4 come along - just like the proverbial bus!!

As I have 2 Avo VCMs which use this valve, and don't want to go solid state (I like to stay with thermionics if at all possible), I was getting a bit concerned. Anyway, thanks for pointing them out. No doubt others will snap them up before I get a chance.

Just a thought, - these are usually mounted horizontally, - I wonder if this causes filament sag, and whether it would be better to mount them vertically; though I'm sure the Avo engineers knew what they were doing. What do you think?

Thanks for the info,

Best,

Dave.

[WME_bill]

I dare hardly question the AVO Valve Characteristic Meter buffs, but surely with a bit of commercial test gear, with perhaps 10% accuracy, it is rather over enthusiastic to worry about the choice of diode, apart from historical veracity.
I assume the concern is the magnitude of the internal resistance when conducting.
These two valves are so different, that it must have been the availability of cheap surplus supplies that made Avo use them.
The EA50 (VR92) is a single diode, rated at 50v, 5Ma max. It can be mounted in any position.
The 6AL5 (and the special quality version 6AL5W), usual equivalent EB91, is a double diode, rated 330V, 9ma. I suppose that a single part of the 6AL5, and certainly the two parts in parallel, will have a lower resistance than the EA50.
How important this will be in a high impedance circuit that generates a uni-directional voltage for the grid I have no doubt the enthusiasts will inform me, to my great benefit.
My CT160 uses EB91 as the "full wave" rectifier. The other two diodes are used for anode current backing off (usually inappropriately called the "bridge circuit") and the screen reverse current protection, for which I would rather hesitate to use an EA50 when it could have 400 PIV across it. Bill

[David Pannell]

Thanks, Bill for your thoughts. I have in fact carried out the mod, albeit on a temporary lash-up basis, and I have to say that I am delighted with the result.

I have performed a self calibration on all parameters and voltage levels, and have found nothing wanting. I have tested a variety of known "test valves", from 6SN7s through to KT88s, incl. ECC82s, EL84s and EF86s. Without exception, all results are consistent with known values.

Although it has righty been pointed out that there are VR92s available, these will usually be used, rather than NIB or NOS, so the object of this exercise was to see if it were possible, and my evidence shows that it is. I have used a new higher grade 6AL5W! (CV4007)

I will now proceed to add a B7G ceramic valve base to the SRBP terminal board of the transformer, adjacent to the VR92 holder, and parallel up the connections. In this way there will be a choice of valves.

I hope others find this little exercise interesting.

Best to all,

Dave.

[PJL]

I am showing about a dozen EA50's here, I think more as I didn't bother counting them all. I am sure other forum members have them too so no chance of running out for a while yet!

[Dekatron]

David,

First of all thank you for publishing your test results on this!

There is an article in the Radiophile from Summer 2000 in 2 parts that you can buy a copy of from for instance CrowthorneTubes which describes modifications, restoration and repair of the AVO Mk II.

In that article the author (which I unfortunately don't remember at the moment) says that you can replace all selenium and valve rectifiers with silicon diodes without problem.

The author also mentions that if you want the Set Zero control to be a little bit less sharp after exchanging the selenium rectifier you can add a series resistor of approximately 150 Ohm between the connections of the 1/6A selenium rectifier and the R3 resistor.

You can exchange all the old rectifiers with for instance 1N4006-1N4007, or the diodes I recommend BYW96E for other AVO Valve Tester restorations/modifications.

There is one more modification which you will have to do if you exchange valve/tube rectifiers to silicon diodes in an AVO Valve Tester on the Screen or Anode voltage path, it is necessary to fit a 100kΩ resistor to earth, after the diode, in order to avoid any spurious voltage readings, presumably due to charge stored in the reverse capacitance of the diode. In the case of the AVO Mk II this will affect the replacement of the EA50/VR92 valve with a silicon diode as the rectifier is put in parallell with a 250 KOhm resistor, which will then result in a voltage divider with the 100 KOhm resistor to ground. My guess, and this is just a guess since I can't test this as I do not have an AVO Mk II Valve Tester, is that you can omitt the 250 Kohm resistor in the situation where you replace the valve rectifier with a silicon diode (this is not mentioned in the article in Radiophile, but the necessity for the resistor to ground is not discussed either).

If you have the time to spare and can test the replacement of the valve rectifier EA50 with a silicon diode and a 100 KOhm resistor to ground and also removal of the 250 KOhm resistor and write about your findings here in the forum it would be very kind of you!

If you go for your solution with a 6AL5W valve I suggest that you check the heater voltage so it does not fall to much, the difference in heater current is quite large.

[David Pannell]

Hi Martin,

First, let me thank you for your encouraging words. I will possibly experiment with the silicon diode arrangement at some point - but the object of our interest in this Forum is surely for all things thermionic!!!

Nevertheless, I quite understand those of us who might like to feel they have the "latest technology" in terms of their test equipment. I'm afraid I'm quite the opposite.

Regarding the heater filament volts on the 6AL5W, there is no difference at all between this valve and the VR92. An earlier inspection of the 6.3v transformer winding indicated that it would be more than man enough for the job. In fact, with the correct line voltage setting on the VCM, both valves drop the voltage imperceptably, showing 6.4v on load! Might even be ok for an EZ80 / EZ81

Cheers,

Dave.

[G8HQP Dave]

There is a valve which is essentially an EB91 with half missing, so presumably half the heater current. I think it was made by Cossor for one of the test equipment manufacturers, but I can't remember the valve name. It was something like 6D1, but that is another subminiature.

[Dekatron]

Originally Posted by David Pannell

Hi Martin,

First, let me thank you for your encouraging words. I will possibly experiment with the silicon diode arrangement at some point - but the object of our interest in this Forum is surely for all things thermionic!!!

Nevertheless, I quite understand those of us who might like to feel they have the "latest technology" in terms of their test equipment. I'm afraid I'm quite the opposite.

I fully understand your, and the rest of the forum members, thermionic interrest, but even AVO replaced the valve rectifiers in their later models with silicon diodes. In some cases it will also improve the measurements as the measurement circuit gets rid of the case of degrading valves that affect the readings.

If you read what I and my friend Euan Mackenzie have written about the AVO CT160 here on this forum you can see that AVO did use the 6AL5W outside of the specifications, https://www.vintage-radio.net/forum/...ad.php?t=58088. We found that a current outside of the specifications is flowing through the 6AL5W which could possibly lead to shortened service life.

Originally Posted by David Pannell

Regarding the heater filament volts on the 6AL5W, there is no difference at all between this valve and the VR92. An earlier inspection of the 6.3v transformer winding indicated that it would be more than man enough for the job. In fact, with the correct line voltage setting on the VCM, both valves drop the voltage imperceptably, showing 6.4v on load! Might even be ok for an EZ80 / EZ81

Cheers,

Dave.

I meant that you should check the voltage since the EA50/VR92 only draws 150mA heater current whereas the 6AL5W draws 300mA current and it could possibly have dipped if the winding was not able to deliver that much current.

Very kind of you to report your findings on this as it gives a solution for people who want to stay with valves rather than use silicon diodes.

[Dekatron]

Originally Posted by G8HQP Dave

There is a valve which is essentially an EB91 with half missing, so presumably half the heater current. I think it was made by Cossor for one of the test equipment manufacturers, but I can't remember the valve name. It was something like 6D1, but that is another subminiature.

You can find equivalents for this tube here: http://www.r-type.org/exhib/aaa0414.htm

[David Pannell]

Fully understand your point, Martin, but as I said, the transformer IS able to deliver the 300mA easily, with the voltage dip being no more than with the 150mA valve.

My point is that the transformer in the Mk 2 has excellent regulation.

Dave.