AVO 163 amp board ~ transformers

[Ed_Dinning]

As suggested by the mods I'll start a new thread in this section about this piece of kit.
It may help if someone can post the circuit and does anyone have an idea of the spec of the transformers in question?

Ed

[Dekatron]

Specs can be found here: https://www.vintage-radio.net/forum/...5&postcount=48

Diagram here: https://www.vintage-radio.net/forum/...65&postcount=2 on page 70 (in third file from top: AVO_VCMs_Circuit_Comparisons_P61-72_CH6.pdf)

I don't know exactly what ferrite core and trimmer that was used but maybe HBWOODY might do some measurements while his amplifier board is being checked by him.

[Radio Wrangler]

It looks like a Mullard Vinkor pot core, though which material and Al gapping, who can say. If it's been varnish impregnated, any attempt to open it will crack it.

However, it's tuned with the adjacent capacitors to the same frequency as the oscillator board. Maybe better put as the oscillator board is tuned to the frequency the meter board peaks at.

The turns ratio will be a bit of a guess, but within reason the gain of the following amplifier can be fiddled to suit.

David

[Dekatron]

You are lucky!

I just remembered that I had one of my VCMs in my storage close to home so I went and fetched it and de-soldered the ferrite transformer and took some photos. I might unsolder the wires some day to take some measurements, but probably not soon as I still don't have a proper hobby room after my renovations stalled due to Covid-19.

The core is indeed a Mullard core, type number LA24 with full specification LA2401-B6!

But here they are!

[Radio Wrangler]

So, it has an adjuster slug fitted, and that means the centre post of the core must be gapped to set the inductance factor Al (Al is often given in nH for a single turn) Multiply the Al value by the number of turns squared to get the inductance of the winding. This is for the core WITHOUT the adjuster slug. Fitting the slug, can only increase the inductance so the centre of the slug adjustment range is something like 5 or 7% greater than the qoted Al would give.

LA2401, I think gives the size of the core in the databook, and I think the -56 part relates to the gap ground onto the centre post (The same core comes ground to several choices of Al value. The 56 may ecen be the Al value itself, but Mullard might have used different units. I'm familiar with later cores from Siemens and Philips/Ferroxcube that used nH/(turns squared)

David

[Dekatron]

Some data can be found here on the Mullard Vinkor series but I didn't see anything special on LA2401 data there, it all seems to start at LA2403 in those documents : http://www.electrojumble.org/techdata.htm

Googling some more didn't turn up anythin on the LA2401, everything seems to start at LA2403 that can be found as PDF-files.

[Dekatron]

I made some more measurements of the ferrite core. The top core has the gap which is 0.25mm, outer height is some 6.75mm and center height is some 6.5mm, the bottom core has the same overall height.

[Craig Sawyers]

In old money 0.25mm is very close to 10 thou, which is probably what the gap was specified to be.

Craig

[Dekatron]

This document: http://electrojumble.org.uk/Mullard_PotCore_Equivs.pdf says that the LA2401 core is obsolete and should be substituted with LA1224 cores. Some data on LA1224 can be found here: https://www.tvsat.com.pl/pdf/K/kat_mullard.pdf which says:

LA1224 Pot core 21mm A13 (3h1) µe = 100 with adjuster at mid range position, Adjuster LA1502

[Dekatron]

Did some quick measurements with my caliper and it shows 0.15mm wire on the high ohm side and 0.3mm on the low ohm side, the thicker wire is wound closest to the center.

The bobbin is a double section type where it looks like the low ohms winding is only wound on the bottom section and the high ohms winding on both sections - it could be that the low ohms section stretches to the top but I can't see it there, except for one end exiting at the top.

The yellow and black sleeves are at the top side.

[Radio Wrangler]

It will have to stand several hundred volts between the Anode/Low Ohms/Primary winding and the Meter/High Ohms/secondary side. So just how they've got the windings separated will be quite important.

Re-designing wouldn't be too difficult if the material can be found and that LA1224 reference should tie it down. A currently available core of about the same material and centre post cross-sectional area, and a similar Al value, then just calculate the turns to get the right inductance for resonance.

As the thing is run at 1kHz, most L-C-R bridges run at 10^4 redians/sec = 1.59kHz, so they would give a trustworthy inductance measurement. Note the stress on bridges... the multimeters and hand held LCR meters are not bridges. These work with time domain pulse techniques and can give rather confusing results.

The reason for picking the same (or larger) cross sectional area and a similar ferrite is tha the primary winding carries the valve's full anode current, and we don't want it to go anywhere near saturation, which would distort the mA/V reading with increased Ia.

David

[Dekatron]

There is some kind of yellow tape between the windings, at least partly as it looks like the different windings are touching at both top and bottom.

I'm not keen on unwinding these windings as it comes from a working VCM163, but if someone else has one with broken windings you are welcome to send it to me so that I can unwind it and document its build.

Its run at approximately 15kHz in the tester and the test procedure is done at 10kHz.

[dazzlevision]

Perhaps Forum member AndyGilham can help with the original specifications, as he works for Megger Instruments (previously known as Avo) in Dover?

[Radio Wrangler]

15kHz.... I wonder where I got 1kHz from? The Sussex maybe?
That reduces the Al it would need to be, and helps with saturation.

Don't under any circumstances unwind a known good one.

Once made, these transformers ought to be very reliable. But then, the metering board is very basic. The transformer and the board itself ought to be fine. There's only a small population of transistors, diodes, resistors and capacitors on it that could be faulty. My mouth fell open when I saw thoughts of buying another whole VCM163 just to get another working metering board instead of fixing the original one. We all know what prices those things change hands for.

A megger test between primary and secondary will check the isolation, an inductance test with D or Q will check for shorted turns. So transformers are quite testable.

David

[Bazz4CQJ]

It's a while now since I first had to give attention to these boards. Maybe 3 or 4 years ago, my 163 was not responding to the gm cal check. I took out the osc board and quickly found that a number of the electrolytics didn't check out at all well, especially the small 1uF ones. So, I bought some quality replacements for both boards and installed them and re-installed the boards but then I seem to recall that the 163 still wasn't working and had to do additional work to sort it out. I will see if I can find the notes I made at the time to find out exactly what the problem was.

I do recall that although the boards are really quite simple builds, and very sparsely populated, I just had this sense that I was working on units that either were not the most fantastic quality to begin with, and/or, had not aged well. So the fact that some others are having problems doesn't come as a surprise; I wonder if we are at the end of the bath tub curve?

B

[Dekatron]

Yes, these are getting old now and most seals on the electrolytic capacitors have dried out slowly evaporating all of the electrolyte leaving them dry. I regularly replace all electrolytic capacitors just to save time checking which are good or bad and I usually use RIFA long life electrolytics instead of the old ones. Then I replace most of the black Welvyn resistors with new ones, the silvery looking Welvyn resistors are usually ok, but I sometimes replace them to. Other types of resistors are usually a lot better and seldom need replacing. I have seldom had to replace any of the transistors or diodes but if I do I use BC549C and 1N4148.

Since the oscillator is a Wien-bridge type it sometimes have problems starting to oscillate if the component values are out of tolerance, if it happens I replace all components that I can. The trim potentiometers sometimes work very poorly, noisy contact, so I replace them with Bourns 10-turn types like the 3296W.

[Radio Wrangler]

Wien bridge oscillators have a relatively narrow range of gain where oscillation will stabilise, set by the range the thermistor will slew over. Go a bit below the useable gain and it either never starts or stops if you even look at it. Go a bit above the useable range and the amplifier bangs against its end-stops giving a very distorted waveform.

David

[Ed_Dinning]

Many thanks for all the research gents, I'll have a look in my old Mullard ref books and see what I can come up with.
The yellow tape is type 56 polyester transformer tape, now available in black as well!

I may even have some of these cores in my collection

Ed

[Bazz4CQJ]

It might be helpful to try and define what's being discussed here. Are we involved in a discussion about whether one or more of us would like to acquire new/spare transformers for the 163?

If Ed was in a position to offer to make these, I'd certainly be happy to join a group buy.

B

[Ed_Dinning]

Hi Folks, it all started as a bit of a mind exercise, but I believe I may have some suitable vinkors (SWMBO is always interested in stock reductions !) So I'm quite happy to try and wind something based on Martin's figures. Did you have the turns ratio Martin?

I believe to OP was also talking about another (output ?) transformer; anyone any ideas about that spec?



Cheers, Ed