Winding a SMPS transformer

[Harvs]

http://youtu.be/qYK2l5A9KrM

Makes my 20-30min seem a bit slow!

[AndyC_772]

Yikes... I wonder what the leakage inductance is on that transformer, and what it would be if the windings were applied with a little time and care to keep them flat and parallel?

A transformer with a poor coupling coefficient can be a nightmare; the excess energy in the uncoupled inductance rings, can generate high voltage transients that stress the driving transistors, radiates EMI and gets hot.

That's why your cheap consumer electronics waste power and don't last long, right there 

[Psi]

I wonder how they solder the connections, since the wire is enameled and she is twisting it around the terminals.

Maybe they dip the terminals in some sort of acid or enamel solvent before soldering? Or maybe its spot welded/melted together

[Harvs]

A worry for me, if you look she first applies barrier tape to the ends, then proceeds to run the windings completely over it. If I'm not mistaken the whole idea of the barrier tape layer is to keep the isolated windings enclosed and provide the minimum distance from the other side windings.

[BravoV]

Makes my 20-30min seem a bit slow!

Well, if she had your speed, probably she won't able to pay her daily meals. 

[TerraHertz]

Wow. She is fast.

Andy_C: but her windings ARE flat and parallel. Watch closely. She's laying two wires, bifilar wound. She uses the tendency of the wire to space itself, does one layer then runs the second layer back the other way.

The leakage inductance - I'm pretty sure is a function almost entirely of the ferrite core geometry and gap, not the windings.  Anyway, bifilar winding. Best possible coupling.

Psi: the twisted wires will be soldered with an iron, the ordinary way. Can't recall just now what it's called, but the enamel will be a kind that decomposes at soldering temperature, and also acts as a soldering flux. It works very well. Just wrap the enameled wire round the terminal, apply heat and solder. Perfect joint.

[Psi]

Psi: the twisted wires will be soldered with an iron, the ordinary way. Can't recall just now what it's called, but the enamel will be a kind that decomposes at soldering temperature, and also acts as a soldering flux. It works very well. Just wrap the enameled wire round the terminal, apply heat and solder. Perfect joint.

That's pretty cool, i don't think i've ever come across that stuff.

[notsob]

PSI - I haven't used that type of enamelled wire, but have read of it. Here is a quick description and a seller.
http://www.synflex.com/en/produkte_pdf/?id=5&areaid=wickeldraehte
http://wires.co.uk/acatalog/ax_ec_wire.html

[amyk]

You'd think the video is sped up, if you've never seen this in real life.

They really are that fast, because they've gotten a lot of practice.

[SeanB]

Self fluxing enamelled copper wire. Absolutely the worst to use in winding a motor though, as it can run internally so hot ( with class H windings common these days) that the wire will self flux and short out. Just think though that she probably is being paid 2 US cents per transformer.

[Flávio V]

Self fluxing enamelled copper wire. Absolutely the worst to use in winding a motor though, as it can run internally so hot ( with class H windings common these days) that the wire will self flux and short out. Just think though that she probably is being paid 2 US cents per transformer.

(2 US$ cents each means that she makes 2,117.65$ per hour which is almost like a portuguese worker)

[peter.mitchell]

Self fluxing enamelled copper wire. Absolutely the worst to use in winding a motor though, as it can run internally so hot ( with class H windings common these days) that the wire will self flux and short out. Just think though that she probably is being paid 2 US cents per transformer.

(2 US$ cents each means that she makes 2,117.65$ per hour which is almost like a portuguese worker)

2c/transformer works out closer to $3.70 an hour, 3600 seconds in an hour, 100 second video, 36 x 5 (number of transformers done in video ) 180 x 2c/transformer = 360c, but then you take away the awkwardness of the dude with the camera over your shoulder...

A worry for me, if you look she first applies barrier tape to the ends, then proceeds to run the windings completely over it. If I'm not mistaken the whole idea of the barrier tape layer is to keep the isolated windings enclosed and provide the minimum distance from the other side windings.

TBH, since she does only the outermost winding and tapeings, i'm more concerned about the person who did the inner one and it's tape, and how thick it is.

[KJDS]

It's standard practice to only use 50 minutes an hour in a production timing, the rest being for P & T

[Ghydda]

The leakage inductance - I'm pretty sure is a function almost entirely of the ferrite core geometry and gap, not the windings.  Anyway, bifilar winding. Best possible coupling.

No the leakage inductance is a function of the spacing between windings. More spacing means bigger leakage. It's the flux that is does not encompass all windings that leads to leakage.

[Harvs]

A worry for me, if you look she first applies barrier tape to the ends, then proceeds to run the windings completely over it. If I'm not mistaken the whole idea of the barrier tape layer is to keep the isolated windings enclosed and provide the minimum distance from the other side windings.

TBH, since she does only the outermost winding and tapeings, i'm more concerned about the person who did the inner one and it's tape, and how thick it is.

Please excuse the dodgy screen grabs.  But if you look at how the barrier was applied to the isolated side, then the winding run over it, it now leaves no second isolation barrier between the windings that enter and exit on the far side, and the windings that enter and exit on the near side.  The barrier tape as I know it is there to enclose the windings and form a complete "loop" with the main isolation tape to ensure there's always a second barrier beyond the wire enamel between the primary and secondary.  Happy to be corrected though.

[SeanB]

True, this will not likely pass a 2kV flash test at 100% pass rate. You would find possibly 1% fail, law of averages on the thin spots of the insulation. It would have taken an extra 5 seconds per unit to place the 4 sleeves on the leadout wires for the inner side, probably cutting production by around 30%. As this worker probably does these in 2 passes ( start in morning doing the first windings and placing in a pile then switch when there are 1000 in the pile) this would be a big cut in production.