You're up against proximity effect which will be more prevalent than skin effect. "Soft Ferrites" by E.C. Snelling covers proximity effect. If you have an university nearby, they usually carry this book. There is an optimal wire diameter to use, but there is around a 2 AWG variation from the optimal size which will work well in practice.
Yes, but interleaving does not mean what you or he thinks it means (apparently).
You should start out placing the entire primary on top of the rest, and then dial that in, THEN place it at the base of the bobbin, and dial in the secondaries (again), THEN you can start working on optimization elements, though in your app it would entirely unnecessary.
A would also look into a increase in the primary gauge choice.
Ok. I still think the current is high for that since a single #26 is only good for one amp. and it would take more than 3 #30 to get that. I would then jump to 4 x #30. If the turns count is low, you could simply place it in its entirety into a Teflon tube and forget about tape altogether.
Reducing leakage induction.
pot
Within my reach, I have 19mm wide tape (from Farnell or RScomponents). As he speaks of 50 W output, core size is not very small. As he don't need lots of the tape he can cut it by hand.
rad
me
rHe mentions it is line-powered, so you have to account for over voltages and whether single or double/reinforced insulation is requried, even when the output voltage is 5V. If Hammy doesn't have EN
60950, or 60065 at hand, he may search for ECMA-287. It can be downloaded free and gives good guidance on creepage, clearance, insulation requirements, etc.Best regards,
Wim PA3DJS
On a sunny day (Sun, 20 Jun 2010 13:22:14 -0400) it happened Jamie wrote in :
Well, if they keep replacing parts, when they get to the brain, he wont be him anymore.
isolation.
Only up to about 1000V isolation (that's generous). Without impregnation, that paper looks like it is not even there to the excited electron. Even through several layers of it. D'oh!
Much easier to find the current common item, DuPont NOMEX insulating paper.
Jeez. I have been away from ebay so long I never even gave it a thought, but OF COURSE!
On a sunny day (Sun, 20 Jun 2010 11:59:21 -0700) it happened Perenis wrote in :
isolation.
Sure, but how much voltage you have between layers if you wind in a normal way is something you can calculate.
1000 V is a lot, much more then was dealing with. Say if you have 1 turn per volt, for arguments sake, and 100 turns per layer, 3 layers, then there is no problem. IIRC I used about 1 turn per volt 440 turns for the H scan coil drive winding Long time ago, BW TV, my school days, tubes, worked nicely.
True, all but for heat considerations that cause me to lean toward the high side of any such window. Sort of like the buffer we use in selecting a cap voltage for a design.
Sure. One should... where it is NEEDED. No need... no do.
In this app, I'd be willing to bet that the operational efficiency gain would be no more than a single percentage point or two.
The UNIT is, not the transformer.
I would feel reasonably sure that the hammster does not need to worry about his AC flashing through to the operator. Andif it doe, he'll trash it and build another.
He has stated SEVERAL times that this is NOT a product!
He knows about that already. SO DO I. The only difference is that *I* DID read the thread, and *I* DO know what his considerations for this are. It is for his personal lab.
First, he wants quick and dirty to prove the design. Then, he MIGHT fine tune it, characterize it, or 'dial-it-in'. If it works right on the first jab, he might just be done with it and get back to his other bench work.
Exactly.
Thanks for reading the requirments.;-)
isolation.
How many layers did that end up as?
All it takes to breach the mag wire at the vpt is about 300 turns, so it also would have to have been flat wound. Can't go that high with a scatter wind, and regardless of your claim, an overvoltage event would punch through that like a needle through skin if the lead-out wire (or turns) are anywhere near the lead-in.
On a sunny day (Sun, 20 Jun 2010 15:25:31 -0400) it happened Hammy wrote in :
Any mains isolation could be solved by putting windings on separate legs and using 2 U cores.... Earth the core. Safer not possible.
On a sunny day (Sun, 20 Jun 2010 12:29:18 -0700) it happened Perenis wrote in :
isolation.
Well, for that winding about 5 (with 100 per layer, flat wound). Critical 'overvoltage' events would mean a spot deflection of 3 x teh screen width? Are you dreaming? The HV coil was on a different leg and done in a different way. He is not doing a HV coil (18kV).
It is NOT a line transformer!
No s*it, Dip Tracy.
the outside wraps, even the ones around the ferrite halfs can sometimes be peeled off an re-used. the keyword is sometimes.
I'm quite aware of how many tranformers are needed as I was the first person to even ask such a question, before every other expert went off about how to vacuum impregnate a coil.
So what you're trying to say is you don't know a damn thing about winding transformers without bobbins, which is how how most tranformers used to be made and quite a few still are.
Have something to add? Share your thoughts — no account required.
Ask the community — no account required