Chokes: How to do the Gap right?

Sadly TwinX is not working at the moment. So you miss pictures of my first choke winding.

I try to do my own chokes for Power Supply filtering.
I allready found out what wire size to use.
Now the wire is on the bobbin and I have some EI sheets lying on my bench.
But nowere I could find any info about how the Gap is practicaly done.
Every choke info page says: "Ah and yes ofcourse you need a gap!"
or: "One can calculate the gap by:xysde/45hf"

But what if I now know: 1mm is what I need. (Just for example. I do not know it for now)

Will I just skrew the I beside the E and leave 1mm air gap?

Maybe someone here knows the official way.

Thanks,
Stephan

Huh?

lg = (uo * L * I^2 * 10^4) / (B^2 Ac)

Ac = cm^2 (core cross-sectional area)
uo = 4 * pi * 10^-7 H/m
lg = m (gap length)
L = H
I = A
B = T (max flux density)

This ignores some parasitics but should get you close. Or you can pick a number and see if you saturate the core, change the gap till you don't. Maybe 0.5mm ?

this is priceless

> chokes for Power Supply filtering. ... how the Gap is practicaly done.

You don't have CJ's experience, you also don't have his silly engineer to satisfy.

I assume CAP-Input, with choke after first cap. Choke-Input is tough.

Use copy-paper for shim. Build your gear with chokes E up so you can mess with the gap. Rest I-lams on top, no shim. Check ripple. Try with one sheet, two sheets of copy-paper. Try again with a clamp. Use the thickness which gives least ripple at normal DC load current with reasonable clamp pressure.

If too thin, the iron is soaked in flux, inductance is low. If too thick, too much "air", inductance is low. There may be a maximum in between.

1mm seems way too big. "zero" might be right: a lam-stack "butt" joint (not the butt joint in that video) is a pretty big air gap. I think one reason CJ loves his hammer is that a sloppy-stack of lams won't give a small gap.

Think whether copy-paper is safe enough and permanent enough, or you should ask CJ for some of that Nomex which can't burn and won't go brittle.

Exactly.
The enginner used to wheel that but stack back and forth so many times, we would get mad at him and take it out on the core, which helped demag it for the next test session!

Stevie Ray Vaughn used to hammer on his output transformers, how can a coked out Texan figure this stuff out>

Bored in the hotel room I guess.
They used to pray every morning in order to get off the powder, it did work, but then Jimmie gave up his seat and the rst is a Waylon Jennings type story.


Yes, if you butt stack a choke, the gap will depend on who cut the core steel.
There is a de reeleing machine used to cut lams. If the spring tension on he de reeler flutuates, you get lams that do not sit flat.
Results: Natural air gap, no gapping material needed.

How does this guy know all this stuff, thats why I hang around,.

So cj ctartys cutting lams, and I tweak the de reeler, now the lams are so perfect, you could but stack a 4 1/2 inch core and it looks like the Hubble, after the corection that is.
Smooth as a babys but.
Only problenm is now we have to add 001 nomex.

Nomex is a fire proof material used in Nascar suuts.
You can hammer the hell out of .003 Nomex, against a shrp lam edge, and it weill not pierce the nomex, thats how tough it is.

New England and Farve so far.

Exactly.
The enginner used to wheel that butt stack back and forth so many times, we would get mad at him and take it out on the core, which helped demag it for the next test session!

Stevie Ray Vaughn used to hammer on his output transformers, how can a coked out Texan figure this stuff out>

Bored in the hotel room I guess.
They used to pray every morning in order to get off the powder, it did work, but then Jimmie gave up his seat and the rst is a Waylon Jennings type story.


Yes, if you butt stack a choke, the gap will depend on who cut the core steel.
There is a de reeleing machine used to cut lams. If the spring tension on he de reeler flutuates, you get lams that do not sit flat.
Results: Natural air gap, no gapping material needed.

How does this guy know all this stuff, thats why I hang around,.

So cj ctartys cutting lams, and I tweak the de reeler, now the lams are so perfect, you could but stack a 4 1/2 inch core and it looks like the Hubble, after the corection that is.
Smooth as a babys but.
Only problenm is now we have to add 001 nomex.

Nomex is a fire proof material used in Nascar suuts.
You can hammer the hell out of .003 Nomex, against a shrp lam edge, and it weill not pierce the nomex, thats how tough it is.

New England and Farve so far.
Nice snow shovel pass, eh?

Only about a dozen, but all run at a minimum frequency of 30 kHz and use ferrite. SMPS design was my day job for a couple of years. No crowbars on ferrite.

Ok cool. :thumb:
It sounds doable for me. With or without Nomex.
Thanks for you kind help.

I am doing a troubleshooting box for some old EAB tube channel strips.
And need to know how many I can recover before I start the big box. :twisted:
So this is the perfect time to try to do my first chokes. Winding was allready a pleassure.

best greetings from Germany,
Stephan

sorry,go to the butcher shop, i know a guy who winds macintosh transformers with butcher paper, but i ain't nanming names.!


you go by heat.

heat = saturation.

dont tell anybody i told you this.

if the gap needs to be big, you got a lousy design.