AVO 7 MkII - AC readings problem

[Superscope]

Hi WD,

The information came from the "Model 7 & 7x Recommended Spares List" Publication.

I have a Hard Copy in front of me at the moment, and I have double checked and
the Spec is as I wrote previously.
I have no means to attach a Copy at the moment.

I think it is posted on the Forum somewhere. I will try to find it.


Ian

[WaveyDipole]

Ian, I am not suggesting that the information you posted is not accurate and I am grateful that you posted it. I just wondered where it had come from as I couldn't find that detail online.

[ms660]

Originally Posted by WaveyDipole

Ian, I am not suggesting that the information you posted is not accurate and I am grateful that you posted it. I just wondered where it had come from as I couldn't find that detail online.

https://www.vintage-radio.net/forum/...80&postcount=3

Lawrence.

[WaveyDipole]

Lawrence, thank you. I have saved a copy.

I spent a little time with the meter this evening, including re-checking my measurements in the table I posted earlier. Incidentally, I didn't mention it previously, but I have been leaving the DMM on for 30 minutes or so to warm up and settle before taking measurements.

I decided that I would not mess with the existing wire-wound resistors and bobbins, but, rather, would calculate a resistor to place in parallel to tweak the value. I focused on adjusting the 100Ω and the top left 20Ω as they were definitely out of spec, and the swamp, as swamp + movement was actually more than 1Ω adrift. I had to correct my earlier measurement in the table.

It worked out rather well.

The 100.37Ω with a 27kΩ in parallel now measures 100.02Ω, so is now within ±0.1Ω.

The 20.192Ω with 2.2kΩ in parallel now measures 20.001Ω, so is now within ±0.05Ω

I couldn't ask for closer than that. A 910Ω was added in parrallel to the swamp which brought the resulting combined value down to 49.987Ω. This was the tweak that made the most difference. The meter now shows exactly a full 1V deflection for a 1V input on the 1.0 range and exactly 10V for a 10V input on the 10V range. Measurements on the 100V range are just shy of their expected values. I tried 20V, 30V, 50V and 60V and all were a tiny fraction low, but I haven't changed the out of spec 800Ω chain resistor yet which does affect the 100V range. Nevertheless, I am pleased with the result so far.

AC mains measures around 235Vac.

Many thanks to those who helped me appreciate how critical these resistor values are.

Next, I will test the effect of swapping out the out of spec 800Ω resistor on the 100V range and probably look at swapping out the one on the 400V range, although technically that one is still in spec.

[The Philpott]

It sounds from the results that you're getting that the movement is fairly well mechanically balanced. If the pointer were significantly nose heavy or tail heavy it would likely have shown up in erratic readings....so hopefully that's one thing that you don't have to work on..

I'm generally impressed with the relative linearity of these PMMC movements, if i applied 10vDC to get FSD, i'd be surprised to get a reading of less than 4.95 to 4.97 on applying 5v.

(Ironic that a lot of modern 'pocket' DMMs have auto power off so you can never reach the 'assured' stability after 30 minutes! My old Farnell bench-top manual asks for 30 mins, and in practice the first 10 minutes are important for accuracy.)

Dave

[WaveyDipole]

Well, I thought I was doing OK, but this morning I decided to swap out one of those 800Ω resistors in the chain that read over 860Ω. Technically, it is still in tolerance (10%) but a long way off 800Ω. Unfortunately the reading I got was around 20% over. 1V read 1.2V and 10V read 12V. I re-instated the original resistor, but it still reads the same?

It did not seem to me to make sense that changing the 800Ω resistor would make a difference to the 1V and 10V DC ranges as it is further up the chain and should have affected the 100V range and upwards perhaps. Swapping the resistor back should have restored things back to the way they were but is hasn't. I can't see any obvious problem, but clearly I must have disturbed something.

Investigations continue.

[WaveyDipole]

I found the problem. It was an intermittent joint on the 60Ω bobbin at the point where the wire emerges from the bobbin to the contact. This also explained why it occasionally reverted to normal only read high again on the next reading. After re-flowing the joint, all readings were steady and normal again, the same as when I left it last night.

I tried temporarily tacking on 800Ω resistors in place of the existing parts, but the reading on the 100V range actually dropped (which does make sense), making it less accurate, so I ended up leaving the existing resistors in place.

After re-checking everything once more, the meter was returned to its case. Now with batteries in place, the three resistance ranges were checked using a random variety of resistors and all readings were accurate.

Fortunately the movement on this one does flow without sticking anywhere.

Dave, I tried your test. At 10V it was spot on 10V. When reduced to 5V, the pointer landed exactly on 5V. When reduced further to 2.5V, the pointer landed exactly on 2.5V. So yes, the linearity appears t be spot on. Voltages were within 0.01V.

[Superscope]

Excellent news,

Another Model 7 saved. It was clear to see you weren't going to give up.

Your Model 7 looks in pretty good cosmetic shape as well and has the much better Rotary Knobs
found on later 7 models, so will for sure be a Joy to use, as and when the Time calls.

They can certainly be a pain to fix sometimes though, but worth it in the end.
The Model 7 is often overlooked, but is a very useful instrument.


Ian

[WaveyDipole]

Thank you. Since I like a challenge, I do tend to persevere. That way one might learn something and in this case I learned quite a bit about the intricacies of AVO movement circuitry.

The case on this unit was a bit rough and had to be sanded down and re-sprayed. The battery compartment had to be thoroughly cleaned, but the front panel was clean and did not need much attention other than a wipe down. I have attached a couple of photos to show what I did with the battery compartment. The batteries are AAA, easily obtainable from a thrift shop. The original copper tabs are still there, just insulated and hidden behind the foam.

This is kind of weird, Ian, but I have a further post from you today @17:53 for which a notification has arrived in my e-mail, but the post has not appeared on this thread?

In any case, thank you for the additional data.

One item I have not tested is capacitance, but I need to read up on that first.

[Superscope]

I did a reply to the Thread when you mentioned you still had a problem, but when I posted you had clearly posted again saying you had solved it.

Since my Post was now irrelevant I deleted it. So you probably got notification of my Post before it was
deleted. It was only up for about 60 seconds.

Now the Capacitance is a coincidence! I was going to ask if you had or were planing to test that?

Being totally up front here, I've been planning to test this function myself, but as yet have not got round
to testing it.
I am also planning to test out the Power Factor functions (pf sockets) at some point. I have the adapter, but again, I've not yet got
round to it.

I'm assuming, the 1uF Capacitor inside the Model 7 will probably need reforming as it looks like an electrolytic and probably never been used in it's entire life.



Ian

[WaveyDipole]

Ah, so that was it! Post was deleted. I made a note of the additional info I got by email anyway.

You also just reminded me that I was going to test that capacitor with the LCR meter before returning the meter to its case and forgot, so now I have to take it out again! I don’t have the adaptor for the Power Factor function so am unable to test that, but I will test that cap before attepting to do anything further with the capacitance function. If I recall correctly, the AVO Test Bridge had one and surprisingly it tested and worked fine on that particular range.

Incidentally, I was going to say that I prefer the knobs on the AVO 7 to the ones on the 8 and later. They seem easier to use somehow.

[The Philpott]

...Likewise i was going to mention capacitance range ..! (No, i haven't tested it either..yet..both of mine are 80-odd years old.)

The packing in the battery compartment is similar to the mods. that i've carried out, but i use expanded polythene as it seems (so far) to be quite unreactive. Thankfully (along with cardboard) it's gradually replacing expanded polystyrene in consumer packaging.
Recyc. carpet underlay does have all sorts of different fragments in it, and tends to break up easily.

Thinking back, both of my Model 7 meters took a lot of work to get them serviceable, although some of the faults were due to abuse.

Dave

[Leon Crampin]

The internal 1 uF capacitor for the capacitance range is a paper block type which is usually OK for its intended application. Slightly annoyingly, there is not enough adjustment to get a zero on an average 240 V mains supply.

A PP3 is fine for powering the highest Ohms range and doesn't usually leak. I pack AVO and some other batteries in cling film to contain leakage.

Leon.

[The Philpott]

Note the factory D cell converter in the picture- simple & effective.

[WaveyDipole]

I had a look at the capacitor today. There are actually two in parallel, the 1μF block capacitor plus a 0.1μF in a metal can. It has a black band at one end, presumably denoting the screened end. I tested them with the DE5000, but the question was at what frequency to measure them at? The default 1kHz or since these are being used at mains AC whether 100Hz would suffice? There is a difference but it is fairly small.

The block capacitor reads 988.1nF @ 100Hz and 980.6nF @ 1kHz. Tolerance is stated as +2% -10%, so still in spec.

The parallel polarised Dublier capacitor reads as 113.63nF @ 100Hz and 106.4nF @ 1kHz. Tolerance is stated as ±20%, so again, in spec.

In either case their combined value exceeds 1μF by around 8.6% - 10%. I haven't tested either specifically for leakage yet, but I was considering removing the Dublier and replacing it with a high voltage metalyzed polyester 22nF which would bring it much closer to the target 1μF. Their combined value should then be adrift by only a handful of nF. On the other hand, I imagine that the Dublier cap will have been matched at the factory to provide as close as possible to 1μF so it seems likely that both have at least some leakage and should therefore be replaced.

Given that mains voltages are involved, I was also wondering whether a safety capacitor should be used as a replacement, but I guess the cut-out would activate before any serious damage to the transformer could occur? Of course, in the day, safety capacitors did not exist.

[WaveyDipole]

After giving it some thought, I decided not to retain the use of either of the existing capacitors inside the meter. The block capacitor has been left in place, but the tubular capacitor has been removed. Three capacitors were used to replace the existing ones: a 1μF metallised polyester cap that measured 986nF, and a 10nF plus a 4.7nF to trim the total value to 1000.4μF on the LCR meter. These were installed in the space vacated by the removed tubular cap, supported at one end by the terminal of the block capacitor and at the other by the existing connection point. The other end of the block capacitor was disconnected and insulated.

Now for the actual test. Most on here who routinely work with valve equipment will know how to safely work with high voltages, but it seems only appropriate by means of a disclaimer, to highlight that working at mains voltage is dangerous and carries risks to health so if you try this with your own AVO 7 then please be careful!

The manual states that capacitance tests can be carried out using between 60Vac and mains 240Vac. With a view to testing at both extremes of that range, a variac was used. The first step was to set the meter function knobs to 'AC' and 'Capacity', and then pull out the 'Q' knob. The meter was then to be connected to the AC supply. At 60Vac, the Q control could not quite adjust the pointer to INF, so the AV voltage was increased to 65Vac which then worked fine. At the other end of the scale, it was a relief to find that when full AC mains was applied, the meter did not let out the smoke! The pointer swung to the far right and by adjusting the 'Q' knob to the opposite end of its travel, it was just about possible be set to 'INF'. So far so good.

The next step was to connect a capacitor in series and read the result. I tried a number of capacitors. First up was a 1μF metallized polyester rated @ 630V. The pointer landed bang on the 1. Various modern film type capacitors rated for high voltage as would be typically found in valve radios of the era were tested. Most read accurately at least down to the 100nF mark. The 10nF (which is the lowest value than can be read) read a bit low, as did the 22nF. Possibly the left side of the scale is not quite linear but it is only slightly off. A 4.7μF electrolytic cap was also tried and measured exactly the expected value. Finally, the original paper 0.1μF cap was tested. At 60Vac it read around 120nF. At mains voltage it read a bit higher, closer to 1.5nF. Presumably this would be an indication of leakage?

The conclusion is that the capacitor range does work and seems to measure quite accurately within the specified range although values seem to be slightly less accurate towards the low end of the scale. The function would no doubt would have allowed an engineer to measure capacitors that were typically found in radios and equipment of the day. According to the manual, electrolytic capacitors down to 12Vdc working can be measured. I am curious whether this function was likely to have seen much use in the day? It doesn't seem practical unless some sort of adaptor to drop the voltage and facilitate connecting the capacitor was available?

I guess its over to Superscope, who has the adaptor, for the Power Factor test....

[Superscope]

Originally Posted by WaveyDipole

I guess its over to Superscope, who has the adaptor, for the Power Factor test....

Well, I'm happy to be prompted to "Pull my Finger out" as they say and get on with Testing
the Power Factor function, as sometimes a little encouragement to get on with something
is definitely needed.

But it will have to wait for a bit as I'm tied up Restoring Phil's old Type D he gave me with the
little Avo Time I have at the moment, and that's turning out to be far more involved than I
thought it would be.

Thanks for being the Guinea pig on the Capacitance testing. I will certainly follow your lead
and try the same.

I will report back when I get round to it





Ian