Transformer HT secondary windings.

[Nanozeugma]

I've noticed recently that the centre tapped secondary windings that feed a full wave rectifier in several radio receiver designs are not "balanced" in the sense that the dc resistances of the "arms" are not equal.
This struck me as curious, in the sense that I can see no reason why this should be necessary and I'm quite sure that there is no reason why the winding resistances could not be identical - if the maker wanted them to be.

The attached circuit extract is from a Murphy A146CM and is typical - the disparity is only ten ohms.

[ms660]

Think of wire length for the same number of turns on an increasing dimension.

Lawrence.

[Refugee]

That is exactly right.
The lower resistance winding is the one that went on the bobbin first. The second one has more wire in it to get around the outside if the first winding.

[turretslug]

As they say. Bifilar winding would result in equal resistance halves, but would be excessively risky as regards insulation stress and isn't worth it (though those more knowledgeable than me can probably quote exceptions). Considering the effect of reflected primary impedance, plus the forward impedance of valve rectifiers, the difference in secondary resistance becomes less significant. In theory, the result will be a tiny 50Hz imbalance component in the much larger 100Hz ripple but, as in life generally, it's a case of maintaining perspective!

Colin

[GrimJosef]

If we feel it matters then we can fix the effect of the lower resistance in one half of the winding by simply adding a suitably high wattage resistor in series with it to make up the missing ohms.

Cheers,

GJ

[kalee20]

Originally Posted by Nanozeugma

I'm quite sure that there is no reason why the winding resistances could not be identical - if the maker wanted them to be.

How would you suggest the maker gets them equal? While keeping the voltages equal too?

[Jon_G4MDC]

Different wire gauge could compensate.

[barrymagrec]

Why does it matter?

[GrimJosef]

It doesn't much. Usually. But as turretslug has pointed out it will add some 50Hz ripple to the normally present 100Hz component. The subsequent low-pass HT filters will be less effective at filtering this out. The lower resistance half of the winding will also be doing more than its fair share of the reservoir cap recharging work, and this will cause it to suffer more Ohmic heating. This will worsen any already existing overheating problem caused by the fact that it's buried deeper inside the windings. It'll also load the rectifier asymmetrically. In the end everything comes down to the numbers. Depending on these it either will matter or it won't.

Cheers,

GJ

[M0FYA Andy]

Whether it matters or not probably depends on whether you are a scientist or an engineer...………..

Andy

[broadgage]

Some HT transformers have some extra turns on that half of the secondary that would otherwise have the lower on load voltage.

Off load, the outer winding with the longer turns will have a higher voltage due to the extra turns added thereto.
At full load, the outer winding will have the lower voltage due to the extra voltage drop in the longer turns. The extra turns only partially compensate for this voltage drop.

At about half load or at about two thirds load, depending on design, then the extra voltage drop due to the longer turns, and the higher voltage from the extra turns, balance out and the two halves of the winding produce equal voltages.

In the case of large power transformers used in the electricity supply industry, then great care is taken to ensure that "equal" windings are truly equal in both number of turns and mean length per turn. That would add excessive cost and complexity to small transformers in wireless sets or other equipment.

[G6Tanuki]

It's quite common to find the same different-resistance thing in push-pull output-transformers/modulation-transformers.

As has been noted, it's because the different windings go onto the core in different layers. The same number-of-turns can involve different lengths-of wire-depending on the diameter of the core.

Provided the turns-ratio is right, the transformer should work.

HiFi 'golden ears' types will split the windings of push-pull output transformers into numerous layers and wire these layers up in ways aimed to defeat the different-wire-length bugaboos. I've never been anally-retentive or obsessive-compulsive enough to fret about this sort of nonsense (but I'm from the school of '"If you can't hear the distortion it's not loud enough!"' amplification).

[GrimJosef]

Sectioning output transformer windings is very much more scientifically-based than that and doesn't have anything to do with different wire lengths. Among other things it reduces leakage inductance.

Williamson goes into some detail about the sectioning needed for his output transformer in the Appendix here http://www.r-type.org/articles/art-107b.htm. That suggests the techniques were known to professional transformer engineers as early as the 1940's.

Peter Walker's Quad II output transformer, over which he took a good deal of trouble, is highly sectioned too. Despite that one half of the primary has a DC resistance of ~120ohms and the other half is ~180ohms (very possibly by design !) https://www.keith-snook.info/quad-ii...amplifier.html.

Cheers,

GJ

[kalee20]

Originally Posted by Jon_G4MDC

Different wire gauge could compensate.

It would - but the difference is generally less than a gauge change. The next larger gauge for the outer half-secondary would actually give slightly LESS resistance. Plus, stopping and changing gauges will add to production costs.

Originally Posted by GrimJosef

The lower resistance half of the winding will also be doing more than its fair share of the reservoir cap recharging work, and this will cause it to suffer more Ohmic heating. This will worsen any already existing overheating problem caused by the fact that it's buried deeper inside the windings. It'll also load the rectifier asymmetrically. In the end everything comes down to the numbers.

True, though the slightly greater charge passed by the inner half, will be countered by it having a lower resistance. The difference in internal heating will be negligible.

[John_BS]

Whether it matters or not probably depends on whether you are a scientist or an engineer...………..

Upvote from me......

[GrimJosef]

The resistance will indeed be lower and the charge (integral of the current over the charging pulse length) will be higher. The heating goes down with the resistance. But it goes up with the current squared of course.

Cheers,

GJ

[Refugee]

You can see the layout of the windings in these photos showing a push pull output transformer.
They are in two sections under and over the secondary.
I fixed this output transformer that had an intermittent short across half of one anode primary.
You can see the smudge in the impregnation where a bodging worker had stuffed the loose wire from the end of the second half of the winding down the side of the link wire where start goes across the speaker winding to the end of the first half.
It was an easy fix once I took it apart and found the fault.

[joebog1]

I have had orders from "engineers" that require I wind part of the low resistance secondary in resistance wire. I have measured the performance and cant see any difference on the DC side. That could be a cap input or a choke input.

I have played with sectioning the secondaries to balance the DC resistance. No difference.

In output transformers I have also wound partially with resistance wire. Measured performance ? No difference.

If one is seriously interested, have a look at Williamsons origional design. He wound BOTH halves of the transformer together, SIDE BY SIDE. That meant that both halves used exactly the same length of wire. Secondaries were interleaved 4 times for the home built versions. The interleaving insulation involved linen cloth ( MUST be un dyed
dipped in parrafin. between the layers.

Today one can wind a transformer BETTER than Williamsons origionals simply by winding in a full width layer with multiple interleaving of the primary and secondary layers.
Winding side by side as Williamson did was perfect for balance and not so good for leakage inductance. In general, providing the transformer is "adequately" designed, the slight imbalance wont matter a fig. I have wound output transformers with vertical sectioning rather than horizontal sectioning, with no change in performance, BUT the two halves had almost perfect DC resistance balance, equal leakage inductance, and somewhat higher interwinding capacitance. Capacitance was not so large as to upset audio frequencies.

I "decent" transformer has a few things that I think ( MY opinion ) more important than DC balance. GENEROUS lump of iron to start helps. Low flux density, within reason because lots of wire means lots of inductance.
Interleaving insulation, be it onion paper or mylar tape. Anchoring of the beginning and end of the winding wires. Termination cables/wires with standard colour coding is nice to have, and should be difficult to pull out of the assembly ( see Refugee's post). Some form of cooling!! I am rather against transformers like Quad filled with bitumen.

Instrumentation transformers are a whole different kettle of fish for those finding fault with my brief explanation above.

Joe

[GrimJosef]

Quad's use of bitumen does make the cooling problem tougher. And if a fault develops elsewhere then the resulting transformer overload causes the bitumen to melt and run out all over the insides of the amp. And the outsides. And, if you're very unlucky, your furniture and carpet !

But ... they are the quietest transformers, electromechanically, that I've ever come across. They don't hum or buzz when so many others do. Radford wound electronically great transformers, but loads of their mains ones buzz. I own two MA50s and an STA15 and all three buzz. I have a customer with a pair of Leak TL/50s and one of his mains transformers buzzes. I'm currently repairing an Audio Innovations 500 and the hum from its mains transformer is substantially louder than the hum from its speakers. I own a Dynaco ST70 whose US mains transformer needs a step-down transformer to go from 230V to 110V. Both transformers buzz. I've had amps sent to me where the mains transformer buzz is the only thing the customer wants fixing and, sure enough, when I checked there were no other faults. Of course 240V and 50Hz in this country don't help.

Cheers,

GJ

[kalee20]

I've also wound an output transformer, with secondary full-width; primary, 4mm Nomex tape in centre of bobbin to divide in two, with the two halves therefore side-by-side, exactly equal length. To keep capacitances the same, rather than one half being inner connection anode, other connection HT and the other half vice-versa, I wound one half clockwise and the other half anticlockwise.

And sectionalised: 3 secondary sections, two primary sections. So the winding was full-width secondary; 2 x half-width primary; full-width secondary; 2 x half-width primary; full-width secondary.

It took ages. Was it any better? I doubt it. And as joebog1 points out, leakage inductance is compromised somewhat doing this.

Although this was an output transformer, the leakage inductance can start to become significant even in a mains transformer, for the abrupt current pulsing in a capacitor-input filter.

GJ's observations on the Quad transformers are interesting! Is there any clue how hard the cores are fluxed? Because that makes a big difference how much they hum.