GB514729 patent on output transformer tertiary winding from 1938

[trobbins]

Does anyone know a method to get a copy of a UK patent from 1938 ? The gov.uk website indicates it can be done online and with email return for 5 quid plus 1.2 quid per page.
https://www.gov.uk/get-copies-of-int...tent-documents
I tried that, but can't get past the "Add Request" step 3 of 4, as it doesn't recognise the patent number or application number that I have details on from either WIPO IP portal (https://patentscope.wipo.int/search/...K2CK4T-55631-1), Espacenet and Google patents.

So I can see an OCR of abstract/description/claims, but that includes OCR errors and no diagrams.

The patent relates to using a tertiary winding on an output transformer for global feedback, and came from Mayo of the BBC. Baxandall refers to it in his WW Jan 1948 'high quality amplifier article' (https://www.americanradiohistory.com...ld-1948-01.pdf), and there are subsequent references in WW to the patent royalty and probably construction complexity bringing the price of an output transformer up quite a bit. So far I haven't identified any commercial output transformer, or amplifier, that has used this technique, so it would be interesting to know if that was the outcome

Ciao, Tim

[Uncle Bulgaria]

It tells me that you can't apply between 1 - 3a.m. UK time. From the time of your post, this might be your problem.

Try again!

[trobbins]

Yes that was an interesting application proviso for 1-3am. And yes an appropriate number does parse the on-line engine now - so it seems like there was a downtime for the search system.

PS. They've taken my money, so here's hoping.

[rontech]

The circuit diagram for the QUAD II power amp seems to show a tertiary winding on the output transformer. This feeds back via a resistor to the EF86 stages of the amplifier.

[GrimJosef]

The Quad II output transformer does have three windings (strictly speaking four, although two make up the two halves of the speaker secondary and are always meant to be wired together). The 'global' feedback to the EF86s actually comes from this speaker secondary winding. The third winding, if you like, is the centre-tapped one that is connected to the output valve cathodes. This only provides local, not global, feedback so is presumably not covered by the patent.

Cheers,

GJ

[emeritus]

You can download GB patents for free from a number of official sources. I usually use the German patent office site DEPATISNET , which has an English language option, or the European patent office's Espace. When I worked for GEC, in the pre-internet days, we used to get copies of UK and US patents from Leeds library, fast and cheap.

[Ed_Dinning]

Hi Tim, Menno van der Veen, Jan Didden, Rudolf Moers and Johan Bergqvist are all Elektor authors who have written extensively on feedback in audio amps, so that may give you a link, as well as the "Linear Audio" series of books they produced.
I do recall a PW article in the 50's about tertiary windings in the design of an output transformer, but being PW it was quite low tech and from memory did not mention anything about interleaved or bi-filar windings
The Australian book "Transformers for Tube Amps" has some detail and also references to the US book "Practical transformer design Handbook" that covers a lot of design theory.
Plenty to research there.
There may also be some BBC monographs or research papers available.

I'd appreciate any other references you may come up with.

One problem with a tertiary winding may be self capacitance to other windings leading to frequency variable coupling into the feedback circuit. This may be overcome on a practical basis by the use of copper screening foil (earthed) on either side of the tertiary; a practical but bulky solution.

Cheers, Ed

[trobbins]

Thanx guys. The depatisnet recommendation by emeritus has access to a scan of the required patent, see link:
https://depatisnet.dpma.de/DepatisNe...729A&xxxfull=1

I couldn't access a scan from the other patent sites, and have now donated to the UK service (which should email me the scan when they've accessed their database).

A cursory read of the patent has a few relevant claims, but no practical examples/results or diagrams. One claim is to use electrostatic screens connected to primary winding ends (plate ends). A separate claim is to use tertiary windings on either side of primary winding sections, so as to couple all flux from the primary windings, where primary winding sections interleave on either side of secondary winding(s).

Ellis was an engineer in the designs dept at BBC in 1955, and pretty much at the same time (1938) proposed the phase inverter / push pull format in:
https://depatisnet.dpma.de/DepatisNe...697A&xxxfull=1

Tanner started at BBC, with more of an acoustics focus.

[Radio Wrangler]

I wonder if it was an attempt to use capacitive coupling in the transformer to improve its bandwidth at the upper end in order to reduce phase lag to the feedback path and so ease the limitation on the maximum amount of feedback imposed by the need for stability? A move towards something like a transmission line transformer.

David

[trobbins]

Originally Posted by Radio Wrangler

I wonder if it was an attempt to use capacitive coupling in the transformer to improve its bandwidth at the upper end in order to reduce phase lag to the feedback path and so ease the limitation on the maximum amount of feedback imposed by the need for stability?

Yes, that was exactly the intent, and what Baxandall was targeting, although he was compromising on output transformer cost by just using one section of interleaving (P-S-P) with secondary wedged between a split primary. To minimise the phase shift of the feedback winding voltage relative to anode voltage, as patented, Baxandall just used a simple single layer on the outside of each primary section (whereas the patent indicates the feedback winding would be on each side of each primary section).

I can see that a multi-leaved design would get quite intricate and possibly bulky if each primary section required a feedback layer on each side, especially with working voltage insulation.

It's a pity nothing else has come to light on any more detailed design, as alluded to by Baxandall, or on what the patent was trying to do with respect to the screening claim.

I searched the BBC reports listing, but nothing applicable.