AVO Two Part Tester

[Chrispy57]

Hi Wavey - I don't know which version of the User Manual you have, but the instructions in the original Avo version seem fairly comprehensive, but require a bit of pondering to fully understand for the new user, pages 4 to 8.

I don't really have enough experience with valves to validly comment on those two anomalies you report, but I have read other similar accounts that turned out OK.

Re. the EBC33 double diode triode, since there is no screen in a triode (or diode) the position of the Screen Volts switch is irrelevant - I would generally set it to the lowest value for safety reasons.

To test the triode section the Anode Selector is set to normal.

To test the diode sections; the Set Zero control is set to Off, the Anode Volts control is set to the D position, and the mA/V control set to the mA/V position, then each of the two diodes individually selected as D1 and D2 by the Anode Selector switch. The scale is 10mA FSD and although no figure is quoted, each diode should be greater than 0.5mA, typically between 1 and 5mA.

There are some strange heater voltages quoted in the manual, as you have mentioned in the PCx series - 7.4, 8.4, 8.5, 9, 9.4, 9.5, 10V etc that I suspect will be rarely encountered as hobbyists, unless you specialise in equipment that happens to use a particular valve. Your series resistor solution (perhaps switched?) and also combining "normal" and Shottky power diodes in series (dropping by 0.7V and 0.3V increments), or using an adjustable DC PSU seem like possible, if messy, solutions should it become a practical necessity.

Good luck with your testing - practice makes perfect!
Cheers
Chris

[WaveyDipole]

Regarding the manual, I have found several documents around the Internet including a couple of versions of the user manual, various parts of the service manual, as well as original and re-drawn circuit diagrams and some information on modifications. In order to put all of this information together I have been re-typing some of this stuff into a single document of working instructions so as to have it all in one place. I am aware of the section you mention and had read it previously with a view to understanding the general test procedure. However, I am very much a newbie when it comes to testing valves and still have much to learn.

For this reason, I decided to go through that section again, taking each type of valve in turn, reading the applicable instruction, then setting up the test and measuring the valve. This has been a very instructive and given me a better understanding of each valve type and the procedure for testing it. Since I don't have any to hand, I was unable to test a Q.P.P valve or a magic eye, but I was able to test the types of valves found in a typical radio - rectifier, pentode, frequency changer and double triode/diode. I have yet to test either of the 7-pin battery valves I have to hand.

I also found the cause of the problem with the NOS EBC33. It has a loose fit and some pins were making poor contact with the base. It needed to pressed gently to one side for the heater pins to make proper contact. It then came to life. I opened up the panel section and used long nose pliers to tighten up the pin sockets on the octal base a bit. It now sits firmly in the socket and can be measured properly. The ma/v of the triode does seem rather low, but the status was 'good' and a reading of around 2mA was obtained on both of the diodes.

Unfortunately I then arrived at the 'Check meter sensitivity' section where things went rather downhill. I followed the instructions to "de-solder the lead from the wiping contact of the SELECT ANODE switch to the rectifier" and take the two connections as described. A resistor decade box, an AVO 7 and PSU set at 10V were then connected in series as described. As a starting point, resistances of 10kΩ, 1kΩ and 100Ω were dialled in and then adjusted as required to obtain a current of 1mA, 10mA and 100mA as shown on the AVO 7. The readings on the valve tester were way off:

1mA (positiion 1) - AVO7=1mA - scale reading = 5.8
10mA (position MA/V) - AVO7=10mA - scale reading = 6.4
100mA (position 100) - AVO7=100mA - scale reading = 6.2

Significantly higher currents were required to get F.S.D. at each setting. Even with the 40Ω resistor shorted out the movement could not achieve F.S.D. It would seem that the movement sensitivity is far short of what it ought to be. I am unsure of what to do about that (if anything)?

Incidentally, part of the instructions require "an Avometer which has been standardized against a sub-standard meter". What does "sub-standard" mean in this context?

[Chrispy57]

Hi Wavey,
Ah yes the old "sub-standard" meter causing confusion again - I used to wonder as a teenager reading adverts in old Practical Wireless mags in the Library why the sub-standard meters were dearer than new ones, if they were "sub-standard"? Many years later I realised that this meant that they were actually perfectly good meters that had been compared/calibrated against a "Standard" meter in a high quality test laboratory, which itself had been calibrated against physical standards in an even better lab.

I recently came across the term "Q.P.P." when restoring an old battery-radio mains power supply unit, which after a little research I found to mean Quiescent Push Pull, and it seems to refer to the circuit configuration rather than a particular type of valve. Any old pair of triodes or pentodes can be wired in push-pull configuration, although in practice they would be of the higher output power type in a power amplifier stage driving the loudspeaker, or transmitter aerial. There are double pentode valves which have the same characteristics for each section, as well as double pentode valves which have a low gain section and a high gain section (eg PFL200). Similarly for dual triode valves.

Well done for sorting out the loose valve holder issue, it's the sort of intermittent fault that can drive you crazy!

The sensitivity check procedure you describe may seem rather baffling, but it just shows up the same sort of 600 vs 1000 situation that you've always had, doesn't it? I don't have any real answers, but wonder how much you had to adjust the starting point resistances of 10kΩ, 1kΩ and 100Ω to obtain those current values from your 10V supply?

Cheers
Chris

[WaveyDipole]

Thank you for confirming the explanation. I did wonder whether it meant calibrated against a standard and therefore one might say it has been (sub)jected to comparison against that standard and deemed to be accurate. The modern-day connotation tends to imply something rather different!

Yes, the sensitivity check procedure does seem to show the same 600 vs 1000 situation. I had left everything set up on the bench last night, so went through the process again this morning and noted the resistances:

1mA - 9800Ω*
10mA - 970Ω
100mA - 80Ω

I also noted the resistances and currents required to get F.S.D.:

1mA - 5228Ω - 1.85mA**
10mA - 590Ω - 16mA
100mA - 44Ω - 160mA

*The first 1mA reading baffled me because it retunred a value of 10,200Ω so I repeated it on the 0.002 scale of the AVO7 and the result was 9800Ω which seemed more consistent. Just to be sure, I also ran the tests using the Fluke 87 and the results were comparable: 9800Just to be sure, I also ran the test using the Fluke 87 and the results were comparable: 9800Ω, 970Ω, 79Ω. I dare say that the 0.01 scale on the AVO7 is bit off.

** On the 0.002 scale of the AVO7 this comes out at 5110Ω and 1.92mA. Tests with the Fluke showed 5180Ω - 1.908mA.

[Chrispy57]

Hi Wavey,
so it looks like there's a minor problem with one range on the Avo 7, which is perhaps to be expected, but the overall results confirm the meter transplant theory - so you're never going to replicate the original results as per the manual. If the previous repairer did indeed recalibrate the tester for the new meter and left it as a useable tester, then you should be able to benefit from owning it without worrying too much about the fine detail.

Cheers
Chris

[WaveyDipole]

Yes, the movement transplant theory is now looking much more likely and as you point out, the instrument seems to be usable and working and is capable of at least giving an indication of the condition of a valve.

However, has it been calibrated/adjusted for the meter transplant? I did one final test using an AZ31 rectifier valve. This was the easiest to work with as it has the fewest connections. I hooked the two heater connections and the two anode connections directly up to the panel socket using croc clip leads and inserted the AVO 7 in series with one of the connected anodes. The AVO 7 was set to read DC current. The valve tester indicated a current of around 10mA, but the AVO 7 indicated around 18mA. I repeated the test with the Fluke DMM and got a very similar reading (17.488mA).

The valve tester is definitely giving a much lower indication and I think one can assume that all readings will be around 60% of the actual value. It therefore does not appear to have been adjusted for the meter transplant, nor does it seem possible to make such a dramatic adjustment.

There is a two-panel tester currently listed on eBay (Search "vintage avo valve tester") with just the tester itself and no panel included. It has an old 2 pin plug so might be unmolested. I am wondering whether it might be worth bidding on? It could be restored for comparison, or else if the movement is OK, could be used as a parts donor. Unfortunately the starting price is not that cheap, but is cheaper than the two comparable listings.

I am rather tempted by the CT160, but for the relatively few valves that I come across it seems a lot of money to spend.

[Chrispy57]

Hi Wavey,
another round of testing that seems to solidify the meter transplant theory, I understand your dilemma over what to do next - valve testers can become addictive!

I guess the options, in ascending cost order are:- buy a calculator (£1.99), make a meter amplifier (£10), buy another Avo 2 Panel tester donor (£100), buy a later valve tester (£500+) - all to eventually test a few valves, of which you will only ever use/sell a handful at best. I can imagine the look on Mrs. Dipole's face if you mention that you are going for the latter option!

I believe there are designs for an Op-Amp based meter amplifier on the Forum somewhere, and I knocked up an amplifier for my faulty Avo EM272 recently based on an Instrumentation Amplifier just because I like working with them, given my Medical Electronics background. So that is a viable option that may be worth trying before committing further expense and risk to the project? Don't be rushed into bidding for one of those currently advertised, they come up all the time.

Cheers
Chris

[WaveyDipole]

I agree, the one up for £99 is one to avoid. The £50 one is the unit I was considering. On mine, when you stand the meter up vertically, the pointer moves to the right a little, not the left. That's an valid point you make about it being hard against the end stop. Not a good sign. The sellers description also seems rather steampunk oriented.

The CT160 I am seeing is up for £320, which is a lot of money, but cheaper than they often seem to go for and from the description and photos, the movement seems to be working. Its a lot of money to risk on something in an unknown state and a bit overkill for testing the odd valve occasionally.

I am seriously considering the amplifier though. I found the circuit online some time ago and am just in the process of copying the hand drawn diagram into Kicad as this will allow me to create and order a PCB. I think that is probably the most sensible way to go forward other than the calculator!

[Chrispy57]

Hi Wavey,
good luck with the Kicad PCB, I tend to work with stripboard for prototyping as I still have plenty in stock, then find I'm too lazy to make another one up on nicely designed PCB if I only need the one unit. It would be interesting to see the circuit that you found on-line.

That £320 CT160 looks good at 4:20pm but let's see what it goes for at 6:30 when all the Lurkers jump on it. Sounds like an honest description, and the demo valve falling between the Good and Bad doesn't mean that it's Ugly, as he seems to imply - the valve is what it is.

Cheers
Chris

[WaveyDipole]

Here is the amp circuit I found.

BTW, I have also worked with stripboard in the past and Mouser have the IC in a DIL package so that's a possibility. I need to figure out how I am going to mount the PCB.

[Radio Wrangler]

The amplifier circuit usually isn't too much trouble, it's arranging floating power supplies for it which is the nuisance.

It's also possible to use a better grade of opamp and to not need the offset adjustment.

Anothe thought is that with an amplifier on the go, you no longer get any benefit from particularly sensitive meter movements, so wht not go for a 1mA or 10mA movement? They are usually more robust than more sensitive types.

David

[Chrispy57]

So, the £320 Avo CT160 actually went for £545 with economy delivery!

[WaveyDipole]

Originally Posted by Radio Wrangler

The amplifier circuit usually isn't too much trouble, it's arranging floating power supplies for it which is the nuisance.

I did wonder about that but hadn't studied the problem in detail yet. I am also uncertain whether its meant to be 9-15v total or -9-0-9 to -15-0-15.

Originally Posted by Radio Wrangler

It's also possible to use a better grade of opamp and to not need the offset adjustment.

I am open to suggestions, but what would be considered a "better grade of op amp"?
I have a number of common ones to hand (e.g. LM358, NE5532, MC1455) but not sure if they constitute a better grade...

There is also a discrepancy with the OP amp as drawn. On the datasheet for the TL071, pin 8 is n/c. It is connected to VCC for the TL072, but then pin 7 is out, not pin 6. The symbol for the TL071 in Kicad shows pin 7 connected to VCC but one also has to take care to choose the correct package (D, P or SP) as pins 1 and 4 are n/c on the H variant.

Originally Posted by Radio Wrangler

Another thought is that with an amplifier on the go, you no longer get any benefit from particularly sensitive meter movements, so wht not go for a 1mA or 10mA movement? They are usually more robust than more sensitive types.

David

A valid point and tests already carried out seem to indicate that the consistentlyu low readings are the result of a 1mA movement having been previously substituted for the original 600µA one. Since there is insufficient adjustment available within the instrument to cope with a 40% reduction in sensitivity, the amplifier has been considered as one possible cost-effective solution (the other being a calculator!). I wasn't aware that thetre were 10mA AVO movements around but the schematic of the amplifier seems to indicate a 2.5mA or 1.5mA output, although I imagine this could be adjusted to drive as 10mA movement if needed.

Originally Posted by Chrispy57

So, the £320 Avo CT160 actually went for £545 with economy delivery!

Yup! Interesting. No bids until the last few minutes. Seen that happen before. eBay can be a bit weird like that. Sometimes there are no bids at all and the item doesn't sell. Other times you get nothing happen until the last minutes when suddenly there is a pile of bids. I also see a lot of items with one bid placed. If you don't bid on it, it will stay with one bid. If you do bid on it, you constantly get outbid. If you leave it, it stays at 1 bid until it sells. I guess that was no real surprise.

[Radio Wrangler]

The primary limitation of that opamp in this circuit is the offset voltage. Better grades of that part have Vos specs of 1mV, the more common grades are 5mV.

The manufacturer uses 1mV as their headline figure in the bold passage at the front of their data sheet. You really have to dig into the data table to see what you really get. It's borderline fraud as the headline figures are each the best of the varianrs in that respect, you may not fing there is a part giving all these bests at once. You have to read and understand the guaeanteed spec tables or you'll ve conned.

TL071 and TL072 are different opamps. Some difference in pinouts is quite likely.

TL071 has very low bias currents (because of FET input) substantially lower than you need. Ut has rather ho-hum offset voltage spec. You need better and then you don't need that offset trim pot. Pots cost money and take appreciable time to find out how to set them.

NE5532 and NE5534 are special fast, low noise low distortion really good audio parts. They have terribly large input bias currents, which their target applications don't care about. Your application rules them out.

You need to get up the learning curve on opamps. The standard recommendation is to have a good long read on them in Horowitz and Hill "The art of electronics" Get one via your library, but it's worth buying. This is a fabulous book intended to educate physicists about how to make electronics work. It's not really an electronics for electronics students book, but it turns out to be one of the best how-to-do elctronic hardware books ever.

You won't understand it all at once. You just pick up a little more on each visit. It's painless this way

Change of subject:

Be very careful how you mention ebay or allude to it. Stick to 'I bought it at auction' or just 'I bought....' Is mention of any specific auction site necessary to what you'rre trying to achieve?

There was big trouble on the forum well before my time where people offering things for auction found them being discussed here. There were arguments, threats of legal action etc and some profound nastiness. Let's just say that the moderators are strongly averse to any replay of that experience. As this would seem to strongly discourage any critical comments, which would give an unfair imbalance to the good/bad remarks, we want to be balanced and impartial in our action, so praise also gets banned, just as diligently.

There has recently been a lot of ebay rules transgressions, sometimes with disguise attempts ("A well known auction site"... or "You know where") - Which just make it look more intentionally guilty.

So it's likely from the consensus amongst the moderators that whole posts will be deleted rather than just editing the ebay references out. The rules are on the forum. Kat Manton has put a good explanation accompanying them. We've kept bumping the sticky reminder about them, but it doesn't seem to be working.

[WaveyDipole]

Some time ago acquired an AVO two part tester from a member, sold as spares and repairs but complete and in a nice condition and already re-capped. This is the first time I have owned a valve tester so I downloaded and studied the available material online in addition to that already supplied on paper.

This example has a trip switch indicating that it is the admiralty version. There were problems zeroing it and to cut a long story short, after re-soldering a joint on the MAV pot I first had to climb a bit of a learning curve to understand how to use the meter. Then, with a steer from other members as well as other threads on this forum, the movement was checked for debris between magnet and coil and adjusted so that it now works smoothly. Unfortunately it was also discovered, that the movement has an F.S.D of 1mA.

The instrument usually requires a movement with an F.S.D of 600uA and although the instrument functions work, the mis-match causes readings to be considerably lower than they should be.

Unfortunately I was a little careless with a comment and after a gentle reminder, my original thread on the subject was closed by the mods to discourage the conversation from drifting in an undesired direction.

I thought that would be the end of the story but for a PM a couple of days later from a fellow member named Electronpusher0. He had experimented with the same movement, initially using a re-magnetiser and then replacing the original magnet with a Neodymium magnet. The result was an F.S.D of around 300uA.

https://www.vintage-radio.net/forum/...2&postcount=48

Since the movement was now surplus to requirements I was kindly offered it and it arrived a couple of days ago. Whereas I am very grateful for the consideration, unfortunately there is a drawback. The Neodymium magnet is quite large and chunky and its magnetic field is so strong that it has a rather heavy damping effect on the pointer, resulting in it moving rather slowly. It takes maybe 20 seconds or more to get from F.S.D back to zero and requires a tap at times.

Nevertheless, it has already been useful. I first confirmed the F.S.D to indeed be somewhere around 300uA. The movement coil measures 36.23ohm, compared with 39.11ohm for the existing one so a 33ohm resistor acting as a shunt nicely doubles the F.S.D to around 600uA.

I temporarily un-soldered the wires to the existing movement in the TPT and connected up the shunted 600uA one instead and performed the calibration tests. At the MAV setting of 1 and a current of 1mA, the meter read very close to F.S.D as it is supposed to. At the other settings it is slightly off but close enough to show that the instrument is calibrated for a 600uA. The same test with the 1mA movement showed a reduction of all readings of around 40%.

The question is what to do next. The TPT would work with the shunted Neodymium magnet powered movement, but the pointer travel is very slow.

I have a couple of thoughts:

1. Replace the rather chunky Neodymium magnet with a much smaller one to see whether it reduces the damping and improves response. This will almost certainly affect the F.S.D as well.

2. Transplant the movement block from the Neodymium magnet into the existing horseshoe magnet, replacing the original 1mA movement. My reasoning is that if the 1mA movement coil works fine in the original horseshoe magnet, then presumably so should the 600uA one? I still have the mini Noedymium magnets to boost it a little should it be required. The process should be just a matter of removing the 2 screws on either side of the block and slipping it out one magnet and into the other. Presumably the process should also be possible to reverse should it not work out. Anyone tried that?

BTW, I notice that the forum has suddenly stopped accepting the omega and mu symbol characters?

[WaveyDipole]

The movement sent me by Electronpusher0, for which I am very grateful, is now installed in the Two Part Tester and working. As previously mentioned, the movement had a very strong Neodymium magnet installed, which resulted in an F.S.D of around 300uA, but damped the movement excessively, making it rather slow to respond to input. I removed the magnet and experimented with a combination of smaller Neodymium magnets and eventually found a combination (18 pieces of 20mm x 10mm x 2mm magnets arranged in two rows of 9) that gave me almost exactly 600uA F.S.D. The damping factor is also considerably reduced and the movement now behaves much the same way as the original.

The valve tester now works as expected and gives me the correct readings for my NOS EBC33, a mutual conductance of 2ma/V and a status reading that 90% into the GOOD range. I also tested a PCF80 pentode-triode. The mutual conductance of the triode seemed a bit high at nearly 6ma/V, but the triode was again spot on at 5ma/V.

This valve tester can now be considered repaired and working and my investment in it saved! I intend to add a B9D base. I have already discovered that the two "blank" spaces on the left cannot be used since that is where the heater div/7 transformer is located so I am hoping to squeeze the B7 base in the centre between the four bottom positions. This also means that I can finally get on with sorting through the small collection of valves I have here and sift out the weak and iffy ones. I know there are at least some.

I just wanted to say thanks to those who replied with helpful comments and to Electronpusher0 for supplying the spare movement that he had experimented on and re-built with a Neodymium magnet.