AVO 8 lubrication

[Craig Sawyers]

I've done a pretty comprehensive search of the forum and AVO8 survey threads, and can not find any reference to this.

I've just opened up a Test Set No 1 version of the AVO 8. Not surprisingly in great condition. But the leaf spring main rotary contacts have got a black grease on them.

Now there isn't much in the Test Set 1 manual on this. But the 8 service instructions says "Under no circumstances should lubricating oil be applied to cut-out, switch rings or switch contacts"

I'm unsure what that implies. Does it suggest that lubrication should not be applied to the leaf springs and the contacts on which they ride?

Any collective wisdom folks?

Craig

[pmmunro]

I've always followed the advice not to apply any form of lubricant to either the rotaty switch contacts (range selection) nor to the leaf spring contacts (function selection). The cut-out contacts are in pairs - the large outer ones to handle the breaking of heavy currents and the small inner ones to ensure a low impedance contact for accurate measurement. The small contacts break first and make last leaving the large ones to do the heavy work. Neither of these should nor the "Reverse Moving Coil" switch contacts should be lubricated.

For cleaning the contacts, it is important to avoid spraying anything on any of the switches because this risks droplets coalescing between the turns of the hairsprings, sticking the coils together. For any extensive work it is better to remove the movement from the front panel, avoiding loosening the magnet clamping screws and any contact between the agnet assembly and a steel screwdriver as this could weaken the magnet. (Use a brass or non-magnetic stainless steel screwdriver or interpose a screen of thick card or plastic.

To clean the rotary switch contacts, I have found the head of a cotton bud impaled on the hook end of an engineer's scriber can be got under the side pointing contacts with a bit of manipulation. The tops are more accessible. For the leaf springs, thick paper dipped in isopropyl alcohol is is good - methylated spirit is a second choice. The spring contacts can need adjustment and a sequence diagram is useful to check that they are operating at the right pointin the switch rotation. (Send me a private message if you would like more detail).

Experience has shown that the retaining screws of the face cams which operate the leaf springs cam loosen, allowing the cams to be rotationally displaced, so it is worth checking that these screws are tight.

The one point about the switches which can benefit from lubrication is the detent balls (phosphor bronze) and te indentations in which the are retained. I use petroleum jelly here.

If the meter is being serviced, it is a good idea to clean the cut-out plunger rod and the cut-out table. This calls for delicate handling and adjustment. Please ask if you want any guidance.

PMM

[Craig Sawyers]

Another minor quirk you might be able to help with. All DC voltage reads uniformly 1.3%-2% low.

AC voltage reads 2% low. The same at 100Hz and 1kHz.

This for DC is using my Fluke 332B voltage calibrator with a maximum of 1kV. And for AC my HP 745A (to 100V) and frightening 746A amplifier to 1kV at a lethal 63mA (hands in pockets time). Heaven knows why anyone would need 1kV rms at 63mA.

Anyway, DCV is out of spec and low. AC is in principle withing spec (+/-2.25%). But my experience with other AVO 8's is that they are ridiculously accurate on DCV and ACV.

So I need to find around 1.5%....

I don't have an DC or AC current calibrator, so I cannot check those ranges.

Craig

[The Philpott]

The shunt for adjustment of the magnetic flux is in the shape of an old-school razor blade, and is retained by a cheesehead screw and spring washer....if your under-read is uniform across several ranges it's likely that you can adjust the shunt to compensate for this. (Comment about keeping ferrous tools away from the magnet proper applies) Get the meter accurate on a low DCV range and this gives you a baseline to work from, then look at the ACV ranges and see what's happening. If the voltage ranges are happy and the cam timing/leaf switches are happy (per Peter's comments) it's likely that the current ranges will take care of themselves. The 100mA DC shunt seems to come in for abuse more than others on these meters for some reason, i've seen a few that are cooked.

I have seen very small amounts of black (moly?) grease on the RADIAL sliding contacts of one or two meters but never anything on the leaf switches. Leaf switches should without doubt operate 'dry'.

Dave

[Craig Sawyers]

It looks like the source of ACV and DCV in accuracy is drift in the value of the series and shunt resistors.

Looking at the DC chain:

40M is 3.7% high
14M is 0.7% low
4M is 1.25% high
1.4M is is 1.3% high
400k is is 1.13% high
140k is 1.64% high
57.5k (an encapsulated bobbin) is 0.7% high

All of which is enough to explain the low readings.

Turning to the AC section.

2M is 0.04% high (so basically bang on)
700k is 3.7% high
200k is bang on
99k is 2.5% high
9.8k is 0.7% high
973 ohms is 6% high
96.5 ohms is 0.3% high

So before I point the finger at the meter, and go to the trouble of setting up a sensitivity measurement, I need to correct the resistor values. Many of them have got carbon resistors in series with the Welwyn Panclimatic film resistors.

In other equipment I have with these high Welwyn film resistors suggest they drift high over many decades. So perhaps it is not surprising that these seem to be somewhat high.

Craig

[The Philpott]

That's interesting, would those Welwyns be different from the ones used in the non-military versions? I don't think i've ever changed Welwyns unless they had visible physical damage. I hope you can get sufficiently close to accuracy by chopping out some carbon trimmers.

Dave

[pmmunro]

Avo used carbon resistors to trim either resistor wire windings and, later, high stability resistors. (At least they were sold as high stability components). The philosophy seemed to be that the much lower stability of the carbon resistors was insignificant as the carbon resistor had a much lower value than the one it was trimming.

In the case of this meter, it would seem that the high stability resistors have drifted over their 60 years, so it would be justifiable to replace them with modern components as close to the original specification as possible. They could still be trimmed with lower value padding resistors and, it is to be hoped, the new combination would be more stable than the original.

The value of the encapsulated bobbin is probably acceptable as the instrument would still be incalibration with this deviation, but there is no reason - other than originality why it could not be trimmed down to near the exact theoretical value. Is it possible that there is already a trimmingbresistor in parallel? Generally these wire windings are very stable, which is of course why they were used. Eventually they were replaced by high stability component resistors to reduce costs and in the hope the components would prove accurate and stable enough.

PMM

[Craig Sawyers]

It turned out to be an easy solution.

I simply adjusted the meter sensitivity by moving the magnetic shunt - the thing with the slotted hole - until I got full scale on the 10VDC range. Using a non magnetic screwdriver of course.

Then all the DCV and ACV ranges were bang on full scale and tracked nicely from lower than 100%.

I've buttoned it up.

The only cosmetic thing to do is refill the arrow pointers on the knobs.

Phew

Craig

[The Philpott]

Tired magnets..broadly speaking a 30 year service item but sometimes they stay stable for even longer. With the older horseshoe magnets you have to get inventive as there's precious little adjustment provided....these later ones are better.

The legends are ever so slightly off-white (and more ivory colour the older the meter)
I've used more than one method but eventually settled on white humbrol enamel dulled with a trace of ochre and olive drab (that's not what the colours were called but you get the gist) Wiping excess off with thinners before it's completely dry.

Dave

[Craig Sawyers]

I did my first ever AVO8 Mk2 with tippex, back when it used trichloroethane as a solvent. So you could clean the overspill with trich.

Before it was banned both for health reasons and it was an ozone depleter.

The AVO service manuals say cellulose, which is easily cleaned. They don't mention the colour though.

Looking at the lettering, you're right - is distinctly off white. There must be a RAL code for a close to off whiteness.

1013, 1014 and 1015 might be in with a chance

https://www.ralcolorchart.com/ral-classic/yellow-hues

Craig

[Refugee]

I used a mixture of tippex and yacht varnish applied with a cocktail stick.
The varnish comes in various shades.

[pmmunro]

I've always favoured the cellulose method. The excess easily rubs off with a cloth/paper soaked in methylated spirit.

For me the colour is not critical as any time the meter was serviced, the colour to hand, usually a white touch up paint for a common make/model of car would have been used.

The "dressing" of the front panel with thin machine oil (3 in 1 in the UK) is recommended;

'Place a little thin mineral oil on a piece of clean rag, and thoroughly wipe over the whole of the black mouldings. This general "brightening up" of an instrument will usually have a most profound psychological effect on the owner of rthe instrument and immediately conveys to him the correct impresion that his meter has received careful and pains-taking attention'

Model 8 Mk I & II Servive manual.

I can't imagine any manufacturer writing that, certainly not in those words, today!

PMM

[Craig Sawyers]

Having bought a number of enamel options and thinners - my brain clicked into gear. At one stage I had a car painted in Tudor White. And I had half a tin of cellulose Tudor White.

Now because I have a brain that works spatially, it took me about three minutes to put my hand on the tin - last used 20 years ago.

So now I am in a lettering restoration mode.

Peter, thanks for the hint that meths works perfectly as a clean up solvent. Works a treat.

Craig

[The Philpott]

I've found that ACWEECO were also modest about limits of accuracy.

I'm glad it reached you in one piece, watch that leather strap though as they can be a real killer- if the nickel plating's worn off the rivets the brass reacts with the leather and gives localised hardening.

Dave

[Craig Sawyers]

Here's another detailed question - how are the ohms adjust knobs attached to the control shafts?

Reason is that that the ohms control has a couple of mm end play (ie in and out of the meter front) and and maybe 5 degrees or maybe more of radial play. The ohms/100 and ohmsx100 controls are fine and dandy. Just the ohms one is weird and I'd like to sort it out.

Thanks for any help on this.

Craig

[pmmunro]

If not a thousand, I hope the picture will save a few words.

For the ohms and ohms x 100, the knob has a very short stubby spindle which is attached to a short SRBP tube by a brass split pin, one of two. A phosphor-bronze crinkle washer is fitted under the panel end split pin allowing a very slight axial movement.

The second split pin engages with the slot in the end of the pot. which is mounted on a phenolic moulded sub-panel which fits on the three bushes moulded into the front panel.

The ohms divided by 100 pot is assembled onto the front panel. The knob is moulded onto a hollow brass spindle against which the phosphor bronze arm rubs. The brass spindle is drilled and tapped axially and a cheese head screw holds together an assembly consisting of - a large plain phosphor bronze washer, a pressed wiper arm, a crinkle washer, and a thick plain washer.

PMM

[Craig Sawyers]

Peter - many thanks!

Wow - that is more complex than I was expecting. Wonder why AVO decided on such a cryptic method of attaching a wirewound pot to a front panel knob. I was naively thinking of a grub screw somewhere that I needed to tighten. No such luck it seems ;-)

looking at the subpanel, it seems to be held by three screws - and from witness marks on the screws, it has been out before (why on earth do people not use the correct screwdriver for the screwhead!)

Craig

[Electronpusher0]

For infilling the pointers and text on AVO meters I use Sharpie "Peel off China Marker", this is a wax type marker and does the job very well.
Peter

[Craig Sawyers]

Originally Posted by Craig Sawyers

Peter - many thanks!

Wow - that is more complex than I was expecting. Wonder why AVO decided on such a cryptic method of attaching a wirewound pot to a front panel knob. I was naively thinking of a grub screw somewhere that I needed to tighten. No such luck it seems ;-)

looking at the subpanel, it seems to be held by three screws - and from witness marks on the screws, it has been out before (why on earth do people not use the correct screwdriver for the screwhead!)

Craig

Well that was easier than I though. Took the precaution of taking the meter out first, and that gave me the opportunity to clean the scale, the mirror, and the inside of the window.

After desoldering the connections, the pot plate comes right out with three equal length 2BA screws.

The cause of the woe was the crinkle washer, that pre-loads the knob, had lost most of its crinkle (no idea why). I took the split pin that holds the knob in place out, re-crinkled the washer and reassembled.

All is now as it should be.

I had to adjust the magnet shunt a bit to get it back on calibration, but that is all.

Craig