Offline flyback with centre leg gap...heating of windings

[treez]

Hi,
Some months back Easy Peasy on this forum gave a rule of thumb for how close to a centre leg gap in a offline flyback transformer a winding can be.
Ive search for the post but cant find it.
I am speaking of TEX-ELZ Litz windings.
I think he said the winding should best be at least three gap lengths away from the gap?
What are your thoughts?
Is it worth packing out the bobbin with tape to achieve the distance?

 

[treez]

Ok, its 2-3 times the gap length, as in post #21

https://www.edaboard.com/threads/500w-full-bridge-is-giving-too-high-delta-b-even-with-biggest-core-available.396124/page-2

 

[treez]

Hi,
I have just investigated a 128W Offline BCM Flyback with VAC input = 240VAC.
No PFC...Cin is 300uF. Vout = 26.7V
This is using a E42/21/15 Ferroxcube core with a 1.14mm centre leg core gap. (3C90)
The peak B is 325mT.
The switching frequency, at max power, is 27kHz.

This has survived soak testing at 23degC, in its enclosure, along with a 60W load in the (metal) enclosure, to get the enclosure hotter.
The bobbin is as follows...

https://en.aet.com.pl/Portals/en/Katalogi_pdf/Kat_rdz_i_korp/Kat_korp/korp_e/korpus%20EE42_15%20POZ%201SEK%2012PIN%20HY4214.pdf?ver=2015-06-11-125237-890

The first half of the primary is made of 2 strands of 0.56mm Enamelled copper wire. This is wound over the bobbin and is just 1.1mm away from the gap in the centre leg. By popular theory, this coil should be overheating, however, the thermal results for this SMPS are not bad, and it has survived hours of soak testing like this.
Can we suggest that having coils close to air gaps isnt as bad as first thought?
Incidentally sec is 11 turns of 3 strands of TEX F TIW.

 

[Easy peasy]

it is the 27kHz that is working in your favour re wire heating near the gap ...

put the gap in all 3 limbs and note a further reduction in wire heating ...

having said that, it is a large Tx for 128 watts, which gives it the requisite lower Trise ...

 

[treez]

Would have thought a simple cheap solution for "windings too close to a gap" would have been to just put down a layer of nylon fishing line as the first layer on the bobbin.....tape it down, then do your copper turns on top of that.
Has anyone done this. Its gotta be a really cheap/easy way?

This 60lb breaking strain fishing line is 0.84mm diameter and would be great to pack out any transformer bobbin...surely?......means you can use cheaper enamelled Litz wire instead of expensive Litz with 28 conductors etc etc

Daiwa Sensor Monofil

The Daiwa Sensor Monofil is the UK’s number 1 monofilament for good reason, with incredible knot strength, a thin diameter and a range of breaking strains to cover every style of fishing from float fishing for small silver fish up to deep sea fishing for

www.anglingactive.co.uk

 

[Easy peasy]

Usually the right sort of card at the appropriate width over the gap or over the winding width - is a pretty low cost and effective way to give a flat start to the winding - away from the gap, several turns of card usually, nylon fishing wire may not stand up to 110 deg C continuous operation ...

 

[treez]

nylon fishing wire may not stand up to 110 deg C continuous operation ...

...Thanks, i thought the same, but "Nylon served Litz wire" seems to be "de rigeur" for high power SMPS transformers these days. It always amazes me how none of it ever has a datasheet for things like max temperature withstand, or even isolation voltage withstand.

mwswire.com and New england wire both sell "Nylon served Litz wire" , and dont offer datasheets with full electrical parameters for it......you're just supposed to ask for "xx strands of AWG38 with Nylon serving", and presumably, be happy with it.

This is why instead i always use TEX-ELZ and just use enough paralelled strands till i get the needed amount of strands.....TEX-ELZ has a proper datasheet, with parameters for use with a transformer useage.

Mind you, regarding nylon fishing wire...nylon does have a melting point of 268degc...but it does burn when its heated to melting.

 

[Easy peasy]

the softening temp would be of more interest - of the fishing line

the nylon serve has no safety isolation properties and very little working voltage properties - it is just a placeholder/separator until the vacuum varnish gets in there and does it's job - this is why winding bifilar with it is problematic - there is essentialyno wire to wire isolation until varnished - esp above a few 100 volts.

You can get single and double teflon coated litz, together these give 3 layers & 2 layers if one suffers a pin hole - thus meeting safety isolation standards.

Teflon triple ins litz is hard to wind as it ends up being quite stiff and resistant to following the coil former.

 

[treez]

Thanks, i must admit if multi strand Litz is needed i like the idea of multiple strands of TEX-ELZ, because its malleable, and widely available....its just a bit unfortunate when the triple insulation isnt needed, but at least its not that thick a coating anyway?

 

[treez]

...on the same subject....(heating of windings when too near to the fringing field of a ferrite core gap)....
The following App note (on page 20) regarding a 100W offline flyback shows a flyback transformer using a EER3540 core with a 1.08mm gap. The primary coil is wound first on the bobbin and is of 0.6mm enamelled copper wire (single strand)

App note of Flyback SMPS...

https://www.onsemi.com/pub/collateral/an-4146.pdf

This primary coil is within 2mm of the core gap, and surely this means that the eddy current heating in it is going to be excessive?
Why did they not use Litz wire for this primary?