Avo Valve Tester Rotary Switch and GM Pot Issues

[woodchips]

This is a collection of extracts from a wanted post for an AVO rotary switch.

"I have never seen those particular switches anywhere else. They are prone to breaking so the only hope (very slim) is to find someone who is scrapping one.
A tip I was just given for mine (to avoid possible breakages) is to make sure you only rotate them so that the numbers go up in value (1,2,3 etc) not the other way."


"Due to the way the switch segments are made, you should always rotate the switches "towards you" and do it slowly, treating them with care."

OK, remembered this post and finally dug my Mk1 out to make a note,

but,

we have contradictory advice. One says only rotate so the numbers go up, the other only rotate towards you, needless to say the numbers go down.

Any votes as to which it is?

On a different topic, has anyone tried replacing the gm pot, which isn't linear, with a 24 or 30 way rotary switch using resistors? Obviouly remove the indexing mechanism but this way can set any taper you want. Or is the quantization error too large?

[peter_scott]

Originally Posted by woodchips

On a different topic, has anyone tried replacing the gm pot, which isn't linear, with a 24 or 30 way rotary switch using resistors? Obviouly remove the indexing mechanism but this way can set any taper you want. Or is the quantization error too large?

The quantisation error is probably less than "hysteresis" effect of the wiper dragging the resistive windings from side to side.

Peter

[Dekatron]

I'd say that the roller selectors used from the Two Panel Tester up to the Mk IV and CT160 can be turned either forwards or backwards without any problem at all as they have a robust construction that doesn't obstruct the wheels in any way.

But when it comes to the roller selectors used in the VCM163 they are a problem no matter which way you turn them as their construction in itself is the problem. In the early model there is nothing that can obstruct the rotating parts at all an they are made from bakelite which is very sturdy, but in the VCM163 AVO switched to plastic which does deteriorate over the years as the softener evaporates and the construction then becomes the obstruction. The VCM16 roller selectors have two main areas where the roller selectors are obstructed, one is the finger guard that runs on top of all roller selectors, the one with all slits where the cogs and numbers are displayed and also the metal contact that exits each one of the roller selectors. The cogs can (and in many cases) will snag on the finger guard from beneath and also on the metal contact, especially after someone who isn't aware of this has replaced one or more of the roller selectors and haven't aligned them properly afterwards, they do snag naturally too as they move and wiggle while being used.

I did put PTFE shrink tubing on the metal contacts so that the cogs would slide over them more easily and I also carefully trimmed of the sharp edges on the finger guard where the cogs travel - I did this on a few VCM163's that I've replaced broken roller selectors on.

So no matter which way you turn the roller selectors on a VCM163 you risk to break them but on the earlier models I've never seen a broken roller selectors that has just been turned either way - people stacking stuff on top of the AVO VCM's have however broken the earlier models as the lid is only aluminum and can't take much weight until it buckles down and crushes the finger guard and roller selectors.

It does seem that AVO became aware of the poor grey plastic in the early VCM163's as they changed from grey to black plastic in later models, the later black models also have a very thin clear band over the numbers protecting them from wear, this band is not present on the grey models.

On the gm potentiometer: you can replace it with a rotary switch in combination with a potentiometer. There are some rotary switches where you can have a centre shaft extending to connect for instance a potentiometer. You can then use the switch to select a range and then the potentiometer to adjust within that range. However I do not know how the resistance values could be calculated nor how the switch and potentiometer should have its ranges printed on the front panel nor how they should be chosen. I've seen similar arrangements done to the grid volts potentiometer on a CT160 and it works very well.

[Freya]

I serviced the switch assemble on both my two panel testers; the switches are designed to rotate both ways.
The biggest problem is the lubricant they used, in that its gets absorbed over time into the Bakelite leaving only a hard residue. Even using a modern suitable grease leave them fairly stiff to turn after a period of no use. I tend to use a warm hairdryer prior to any extensive use which softens everything and makes the wheels glide freely.

[David Simpson]

Folks,
Servicing-wise I can only speak of my RAF experiences back in the 60's/70's on just CT160's.
Yes - the rotary pin selector switches can move in both ways.
No - dont use oil-based lubricants, even WD40.
Yes - its safe to use silicone spray or grease.
Yes - I've actually tried using " Hellerine" lubricant-as supplied by Hellerman
for sleeving - a vegatable oil much like baby oil but with a lovely
smell.
Just a wee dab with a baby bud. Works a treat.
Yes - On recently aquired Mk1 & Mk3 - cleaned with "Vapourmatic" VLB4227 Electrical cleaner, then a wee spray with Silicone Spray. Works a treat.
But remember - these switch assemblies can be 50 or 60 years old, and need treating with care. If they're really really stiff, even after using say silicone spray, then there probably will be a foreign body stuck in there somewhere & the whole caboodle will need careful removal to delve further.

Regards, David

[AC/HL]

I was advised to slacken the clamp screws slightly to ease the rotation. Only just enough, not too much.

[murphymad]

Apart from what Dekatron says;

The quote -

"Due to the way the switch segments are made, you should always rotate the switches "towards you" and do it slowly, treating them with care."

Certainly does apply, but only for the VCM163 !!!

The switch construction is of a single pivoted finger that glides over the contact rods within each of the roller switch assemblies.

Because it is only held from one point it can be damaged by rough/fast movement of the rollers in a direction that places the turning force pushing against the end of the finger.
Where as if you draw the rollers towards you, the finger inside the switch roller is affectively dragged over the contact rods in a direction that cannot bend this thin metal finger.

A diagram would make everything clear, but I don't have one. I could take a picture of the inside of a roller switch which would reveal all.

With the early versions of AVO roller switches used on everything other than the VCM163, you may well find the tightening of the rollers is due to the Bakelite actually deforming or growing around the centre which is fixed and does not move.
A complete strip down (careful, the bits spring out everywhere) and gentle rubbing of the inside of the roller with fine wet and dry before re-assembly will make everything run much easier, especially with a tiny bit of silicon grease.

Mike...

[Dekatron]

I've included some photos of the insides of a roller selector from the VCM163 which shows the one-way design of the contact finger - I didn't find the rods right now so I couldn't put them in the appropriate holes, so I made one picture with some arrows showing where stuff goes.

There are three small balls that have three finger springs that press them down against the outer ring which has indentations for each position.

The inner ring also has indentations (cogs) which lifts the contact finger just before it should connect to the next rod, this makes the movement smoother and miminizes the risk of the contact finger bending.

I have two different versions of this contact finger, see the last photo, which shows that AVO might have been aware of the problem with turning the roller selector the "wrong" way, but I have never seen a bent contact finger in any of the broken ones that I have replaced.

I know for sure that if you clean and lubricate the three small balls and the track the roller selector turns like it is new, however that is very tedious work and I don't recommend that you do it as you might expose the plastic to cleaning solutions or lubricants that might make it deteriorate, I tried it on one that was already broken and it was a very smooth movement afterwards - I compared the movement against a roller selector on an unused VCM163 which had been in military storage since it was bought.

There is another reason why cleaning and lubricating the small balls and that is that the springs keeping the pressur onto them sticks as the lubricant used dries and they then stick to the plastic and don't flex as intended. This also means that the pressure on the plastic is raised which might also be a reason for these roller selectors breaking.

[woodchips]

Thanks to all, gathered some useful information in one place.

Don't lubricate with Hellerine oil, it has a completely different use. It is used to lubricate sleeving and similar for installation over cables, BUT it is designed to evaporate so that the sleeving doesn't then slip off again! If you have never used Hellerine oil then how have you survived? It is an essential in the toolkit.

Found my instructions and the gm pot only seems to be used for go/nogo testing so the errors from using individual resistors to get any resistance law you want would seem to work. Just now need to know what the original law is, I haven't found it mentioned in the service data.

[Bazz4CQJ]

I've not needed to replace the gm pot, and I note the service manual says nothing about the taper. Critically, is it the case that replacing the original with a standard log or lin pot makes adjustment difficult/impossible? Doesn't the use of a multi-turn pre-set help out?

[David G4EBT]

If I'm not mistaken, the pot is wirewound and uses three different gauges of resistance wires in series, with three different Ohms per Metre. This is wound onto a circular former which is like three 'sawtooths' in shape on the rear edge but is straight along the edge on which the wiper wipes. I believe that the finer wire is quite frail and prone to failure.

It seems that the reason the pot uses three different gauges of wire wound on the sawtooth shaped former is to attempt to emulate a logarithmic law track. Apart from the three steps in resistance, the sawtooth shape of the former means that each turn of wire will be progressively longer than the adjacent one, so each turn is slightly higher in resistance. Ingenious, but fragile.

Clearly the reason a wirewound pot (normally linear of course) was used is to cope with the current flowing through the pot. Normal carbon track pots are I think only rated at 0.5W, so wouldn't be up to the task.

I would have thought that the suggestion of a mutli-way rotary switch, using a series of modern metal film 2 or 3 Watt (or for that matter, wire-wound ceramic resistors, which are compact these days), would be a more robust arrangement. It would be necessary to plot the track of a working pot to draw a graph at the number of intervals of the switch to ascertain the value of each resistor. Obviously the more contacts on the switch the better, given that it would need to emulate the same total degrees of rotation of the original.

I profess no knowledge of Avo testers - I'm only 'thinking out loud' based on a friend's unsuccessful attempts at recreating a new track identical to the original. (Colin Wood - 'Retired', who is still a forum member, but no longer into radio restoration). When I was in correspondence with Colin about this topic three or four years ago, I got the impression that as with the meters on Avo valve testers, these pots are problematical and a real headache when they fail. Colin spent a great deal of time experimenting, made a replacement sawtooth former from Tufnol sheet, an oven in which to coat the wire with special varnish obtained from the US, but in the end, despite his very skilled efforts and perseverance, he abandoned the project.

Maybe someone has managed to surmount what appears to me to be an unsurmountable problem, (though not one that troubles me as I don't have an Avo valve tester and don't want one!).

[woodchips]

My understanding is as David's. Sure there was a series of posts about fixing these pots, must be his friend?

My understanding of what the pot is used for is when the go/nogo testing of gm is done. You set this pot to indicate the normal gm of the valve, this is compared with the tested valve and the meter displays the result. Sort of a bridge arrangement. This is why the non-linear taper is used, expand the low gm values because a 20% variation in 1.6mA/V is the same importance as 20% of 16mA/V. This is why the strange taper is used, it has to match the gm values engraved on the front panel otherwise the gm comparison will not be accurate, hence can't use any linear, or log, pot. To determine the resistor values just use the panel markings. Mine is still working so not needed to do it, yet.

[David Simpson]

Yep, Woodchips is right - Hellerine does evaporate, so it makes a decent initial cleanser, but you'll still need to apply a wee squirt if silicone spray for lubricating the moving parts.
Reading Mike Phelan's detailed article, back in 2005, on RV2(-ve Vg)(10K over 270 deg.) is a must for anyone with a CT160 !!!
Some folk believe that the whole pot is non-linear - wrong. It's three linear segments, 0 - 5V, 5V - 15V, and 15V to approx 40V. As displayed on each of the three linear dial sections of 90 degrees.
RV1(Set mA/V)(Gm), is supposed to be a 2.5K linear over approx 270 deg., but has a non-linear scale with approx 10deg. set aside for operating SW3.
There is another excellent article by Martin Scobie on CT160's. And on the "Duncan's Amps " Website.
Not to mention reams & reams of info just on this Forum.(See "Search").
There is an old RAF saying - - "If in doubt, keep it(your finger) out!". Particularly RV2, and the RV1 assembly, are now as rare(replacement-wise) as rocking horse manure, & hence frightningly expensive. Take extreme care, or contact one of just two or three CT160 experts down south. eg. Mike or Jeremy.

Regards, David

[Bazz4CQJ]

Perhaps I am the only one who is confused by this thread, but I think that I've now figured out that the OP is not specific, but actually refers to to a CT160.

Subsequent discussion of the 163 refers only to the selector wheels, which have a degree of commonality, but not the "gm pot" which has no commonality!