AVO 163 amp board ~ transformers

[Dekatron]

It looks like Mullard Technical Handbook Vol 6 should have data on the LA2401 ferrite pot core, I can't find that on the Internet as most scans only contain valve data - does anyone here have that volume to check if this core is found there?

[David Simpson]

As Bazz said yesterday, there should be a heap of my standardised project valves(6AQ5's & 6AU5's) floating around out there amongst the Forum fraternity. Must be well over 20. Right enough, his VCM163, Pamphonica's, Dekatron's, & several others used by careful owners all revealed excellent Gm's, as I recall.
Sadly, my last remaining spare st./valve was phooked by HPW.
I might do some more, if I look out some spare similar "middle of the road" pentodes. They'll again be free, but with the caveat that, as before, recipients pass them on to other Tester/VCM owners. And, as before, the project isn't a "my thingy is bigger than your thingy" competition. Just a way of enabling & helping valve testing enthusiasts to achieve confidence in their equipment. Some folk even joined in with their Sussex, and Martin kindly included a go with his RoeTest.
All good fun by coconscious valve testing enthusiasts.

Regards, David

[Dekatron]

Roy at Electrojumble kindly offered to scan the data on the LA1224 which he will put on his webpage, I've included it here too.

[Radio Wrangler]

Great!

It looks like mid adjuster position gives +8% on the no adjuster inductance.

58.16 turns for 1mH

so 1,000,000/(56.16)^2 = 317 nH/(root turns) The usual modern definition of Al value.

Next line down they say Al is 295.7nH for one turn. Close enough as they say.

Modern pot cores are made with Al values in the 300 vicinity, but marked figures are without adjuster, so for a new design the numbers of turns will need slight adjustment.

TDK/Epcos do an RM6 core in N48 material (suitable stuff) ground to Al 315nH/1turn
and the adjuster will increase this about 7%.

But RM6 is a bit too small. It'd be difficult getting all the turns in.

There's Al =315 in RM8 which is 19mm across flats outside.

RM10 size is 24mm across flats and has an Al=400 choice. So turns need reducung a little

https://www.tdk-electronics.tdk.com/.../fer/rm_10.pdf

To make a core up in N48 material:

B65813+0400A048-D is the upper Al400 core half with adjuster hole (that's the -D)
B65813+0400A048-N is the lower Al400 core half with a threaded plastic insert for the adjuster.

One of each of the above is needed per transformer.

Bobbins, there are 2-section and one section models

B65814N1008D001 is a 1 section bobbin with 8 pins
B65814N1008D002 is a 2 section bobbin with 8 pins

I'm not sure of the adjuster they quote two diameters in the data sheet

clips to hold it all together are B65814B2203X000, two are needed.

I'm not sure who stocks these things nowadays, it used to be Electrovalue.

Anyway, these are clues to a modern replacement. Going up in core size allows space for better primary/secondary insulation.

David

[Craig Sawyers]

Mouser seems to stock gapped RM10 cores in N48. Various Al's. The 315 variety is non-stocked though, but others north and south of that value are.

Craig

[Dekatron]

I've looked into cores somewhat and the problem is that there are no mu-shields available from any of the manufacturers that I've looked at. Some talk about their better performance where a mu-shield is not necessary and that may be so but I hink it has to be tested to see if that works as AVO specified one.

I've also seen a few different P-cores very similar to the original ones, also from TDK, like the B65661+0250A048 which can be bought with the same values as the original cores with µe of 100 and with all other parts except for the mu-shield, the number is B65661+0250A048. I've included the datasheet from TDK.

This core has a yoke/bracket that I think would be possible to use in combination with home made a mu-metal shield kept in place by the bracket.

There is one other possibility to have a mu-shield on any of the cores and that is to make one that fits over the hole core like a box, which you then solder in place on a small PCB that holds the core and the mu-shield while at the same time having the same pinout as the old core so that it could be soldered in place without having to modify the original circuit board - just like an adapter.

[Radio Wrangler]

That's just a mounting clip, not a shield that I see in that datasheet. It doesn't do any shielding, the cores do that themselves by having ferrite round the outside, and the vulnerable part, the gap, buried tight in the middle.

This application is tightly tuned to 15kHz to reject mains frequency components, which it is getting hit with anyway via the waveform of the anode current. The field pick-up should be trivial.

I've not had to mumetal screen pot cores yet in my career. I suspect the VCM163 should be OK without one.

David

[Dekatron]

*******, I just realised that I had mixed up the transformer for the TT-537 with the one in the VCM163, there is no mention of a mu-shield in the VCM163 it was the TT-537 that had it specified in the text.

So, we can use any ferrite core transformer we want!

I do have a few of the RM6, RM8 and perhaps one or two of the RM10, but I might not have the correct former and certainly no slug/trimmer. With my parts and the datasheet for the AL1224 I could at least make the small adapter PCB.

[Radio Wrangler]

Remembering the need for high voltage isolation, the larger core will make it easier. I don't think there is any need to be terribly space-efficient, and does this case itself constitute evidence that the original transformer design can fail?

David

[Dekatron]

I've only seen a few occasions where the transformer has been open circuit, in some cases the 1.99uF has been either shorted or open circuit at the same time and the user haven't been able to give me any explanation on what has happened. My guess is what I wrote earlier about oscillations that could damage the transforme, or perhaps someone accidentally shorting the circuit so that the transformer would see the full swing of the Anode voltage.

I'll make an adapter PCB where either the RM8 or the RM10 can be used.

[Bazz4CQJ]

Originally Posted by Radio Wrangler

Remembering the need for high voltage isolation, the larger core will make it easier. I don't think there is any need to be terribly space-efficient, and does this case itself constitute evidence that the original transformer design can fail?

David

I guess that they have lasted for 50 years, and it may be that the way each tester has been stored or used is showing up with a few transformers failing. Is it the thin end of the wedge?

B

[Radio Wrangler]

I think that also a lot of VCMs are getting into the hands of people who don't know how to use them, but most importantly don't have anyone to learn from. No-one is born with fully formed knowledge. Not knowing things is natural.

Back in their day, these were very expensive pieces of equipment and in a workshop anyone who was allowed to use one had been trained to the workshop manager's satisfaction.

David

[Bazz4CQJ]

If ever we were to design a modern replacement board for the osc/amp, then very high on "list of requirements" would have to be some for of protection for the original 50uA Gm meters. They really are intrinsic to the original aesthetic of the 163, and must be retained, but we need to very kind to them, many have already been injured .

B

[Radio Wrangler]

Certainly, the ability to drive original meters is a must, but adding the ability to drive less sensitive and more robust meters will give people with dead meters a wider scope of meters they can choose from, to be able to find a good visual match and use it even if the sensitivity is different.

david

[Bazz4CQJ]

yes, that would be essential. Dare I suggest that drive for a panel digital voltmeter might be nice, or is that a step too far?

B

[Radio Wrangler]

There would be people writing letters to the times, and accusations that it spoiled the sound of valves tested on it, making them sound more 'digital'

I'd do it, of course, but not tell anyone.

David

[Craig Sawyers]

3:47am David? Do you ever sleep?

And Bazz too - 2.25am!

Which is off-topic of course, other than to agree with both dark watches of the night posts.

Craig

[Dekatron]

If I would like to wind an amplifier transformer myself and test it per the test specifications, what instruments would I need instead of using the recommended old Marconi TF1313?

Could it be done in a simple way with say an oscillator and an oscilloscope or do I need a proper LCR-bridge?

[Radio Wrangler]

You need to measure the inductance quite accurately and in a realistic circumstance. That rules out scope and oscillator. Modern digital LCR meters are likely to be misleading, their pulsed waveforms occupy a lot of bandwidth and will contain frequencies where a perfectly suitable inductor for the end use, may be more influenced by strays than is good.

The ideal would be an LCR bridge running at 15kHz. Some bridges allow you to run them with an external oscillator input.

If this is sounding awkward, there is "Plan B" use the oscillator and oscilloscope with the transformer wired up in its resonating circuit and check for resonance at the correct frequency and roll-off where expected. It'll be fairly broad with that 10 Ohm resistor. After all, in circuit functionality is the most important thing.

David

I tend to sleep in patches, and need something as a diversion until I feel ready to doze off again. My thinking might not be at its best, though. Maybe it's a side effect of growing up in a family with a business on 24 hour call?

[David Simpson]

My two pennoth :- I fully agree with David's point of view. But, to get back to basics, whilst avoiding the merits & seemingly complicated theory of ferrite cored wee transformers, AVO's range of VCM's are just that - Valve Characteristic(Gm) Meters. The "Gm" being regarded as "Gm(static)", but in the 163's they introduce an element of "Gm(dynamic)" measurement. I.e. - the 15KHz signal superimposed on the 50Hz sourced 10mS pulses. Following on from Barkhausen's Law - - Gm(d) = Gm(s) x ra/ ra+ Ra. Ra being the Anode load. As long as this anode loading remains in the region of just a handful of hundreds of ohms, then Gm (d) will be just a few % lower than Gm(s). Simple enough to compensate for in the calibration procedure, so that 163 users can refer to the AVO VDM's "mA/V" used by all other AVO VCM owners. In the real world - valves in radios have a wide variety of ra's & recommended Ra's, hence nearly all valve testers just test for Gm(s) & just have a test anode load of 1 or 2 hundred ohms.
In other words, "if in doubt - shout", & don't fanny about. Complicated VCM's, particularly 163's - send them off to Mike Barker, if he's still offering a professional service.

Originally Posted by Radio Wrangler

I think that also a lot of VCMs are getting into the hands of people who don't know how to use them, but most importantly don't have anyone to learn from. No-one is born with fully formed knowledge. Not knowing things is natural.

Back in their day, these were very expensive pieces of equipment and in a workshop anyone who was allowed to use one had been trained to the workshop manager's satisfaction.

David

Regards, David(Simpson)