Tasks such as this should not be attempted by man. / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Tasks such as this should not be attempted by man.

first previous 2 3 4 5 next

Move Thread

LAN_403

radhoo

Sun Feb 20 2011, 11:29AM

Registered Member #1938 Joined: Sun Jan 25 2009, 12:44PM
Location: Romania
Posts: 701

Congrats on this initiative, it seems to me it will be a lot of work.
Have you seen Uzzors2k's Big Mofo? Link here:

HV vs. good mineral oil:

ZakWolf

Sun Feb 20 2011, 11:46PM

Registered Member #3114 Joined: Sat Aug 14 2010, 08:33AM
Location:
Posts: 608

radhoo wrote ...

Congrats on this initiative, it seems to me it will be a lot of work.
Have you seen Uzzors2k's Big Mofo? Link here:

HV vs. good mineral oil:

could he drive that huge core with a zvs flyback driver ? and why did he only put one layer on ?

Dr. ISOTOP

Sun Feb 20 2011, 11:59PM

Registered Member #2919 Joined: Fri Jun 11 2010, 06:30PM
Location: Cambridge, MA
Posts: 652

One layer is a lot easier to wind than multilayer, and works well.
A similar beast is Steve's CCPS:

Adam Munich

Mon Feb 21 2011, 09:05AM

Registered Member #2893 Joined: Tue Jun 01 2010, 09:25PM
Location: Cali-forn. i. a.
Posts: 2242

Got 9 layers done. There's roughly 50T per layer and it appears that I can fit 15 layers per pancake.

This thing better turn 14V into 75,000 when I'm done.

Also, is the unscented mineral oil that pharmacies sell any good?

Microwatt

Tue Feb 22 2011, 05:29AM

Registered Member #3282 Joined: Wed Oct 06 2010, 05:01PM
Location:
Posts: 224

boil the mineral oil at 300f to remove water. looking good there!

Patrick

Tue Feb 22 2011, 06:04AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Use veg, canola, olive oil, or mineral, what ever is cheap and in large quantities where you live!

Wolfram

Tue Feb 22 2011, 10:19AM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

I think the resonant frequency might be a problem.

You have 4 sections, with an estimated 15 layers of 50 turns, according to an earlier post. This makes the total around 3000 turns, and for 75kV output, that's 25V/turn. Ferrite cores are normally limited by core loss, but since the run times are going to be short and the core is oil-cooled, you can run it close to saturation. The saturation flux density for normal ferrite materials is around 0.4T. The diameter of a typical flyback core limb is around 16mm, giving a total core cross-sectional area of around 2cm^2.

Plugging these numbers into transformer design equations tells us that the minimum frequency you can run it at and still get enough output voltage is ~80kHz. If your transformer ends up with a lower resonant frequency than this, you will hit core saturation before you get enough output voltage.

This is a problem, because your 3000-turn coil will almost certainly have a much lower resonant frequency than this. In this

thread, jpsmith123 describes a transformer he made that had 420 turns, and the output voltage started dropping off above 16kHz. The segmented pie winding will have a lower self-capacitance for a given number of turns, but running your transformer with 7 times as many turns at 5 times that frequency is not really realistic.

Or, to put it in different terms, if we assume an Al value of 2uH/n^2, which seems to be typical for flyback core-sized UU core sets, the total inductance of the secondary is 18 henries. To keep the resonant frequency above 80kHz, the stray capacitance has to be less than 220 femtofarads.

This is why you never see high frequency transformers for very high AC voltages, it just doesn't work too well.

You could make it the way they make DC flybacks, with multiple pie-wound sections and diodes between them, or a full diode bridge for each "pie" if you want fullwave rectification. This way, each section doesn't see the inductance and capacitance of the other sections. You would probably need a bit more than four sections though, to get the resonance frequency high enough.

I'm not trying to rain on your parade, I'm just telling you that what you're trying to do is not going to work, to hopefully save you from wasting time winding a transformer that will only be usable as a paperweight.

Dr. Dark Current

Tue Feb 22 2011, 10:52AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Anders: the L in the resonant frequency is determined by the leakage (and series if any) inductance if driven from a low impedance source (eg. a bridge), so the actual transformer/system resonant frequency depends on the series-L parameters and is much higher than the secondary-core self resonant frequency.

Wolfram

Tue Feb 22 2011, 12:23PM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

Thanks for the correction, Dr. kV. If I understand you right, the Fres is determined by the series/leakage inductance and the reflected secondary capacitance. That makes sense.

My main point is still valid though, for a slightly different reason. The reflected capacitance is going to be very high, due to the high step-up ratio and the leakage inductance is also hard to keep down in a transformer made for such a high voltage.

The reflected capacitance is the secondary self-capacitance multiplied by the square of the turns ratio. The primary number of turns can't be too high either, because then the step-up ratio becomes impractically low. Let's say you go for 10 primary turns, that makes the step-up ratio 1:300 if you have 3000 secondary turns. This means that the stray capacitance of the secondary is multiplied by 300^2 = 90 000, so even 10pF of secondary capacitance will appear as 900nF on the primary. This means that the resonant frequency will be quite low.

It might be possible to use a series capacitor to cancel out some of the leakage inductance, not sure if that would work.

In any case, making this work is going to be a lot harder than just using a lower step-up ratio transformer and a multiplier (if it's even possible). Insulating 75kV at 80kHz is not easy, members here have tried making lower voltage lower frequency transformers, and have run into these exact problems.

I would love to be proven wrong though, and the whole thing might be a valuable learning exercise even if it doesn't work, I just want to warn you beforehand that this has a very high chance of not working, so that you know what to expect.

Adam Munich

Tue Feb 22 2011, 10:50PM

Registered Member #2893 Joined: Tue Jun 01 2010, 09:25PM
Location: Cali-forn. i. a.
Posts: 2242

à² _à²

I'll set this aside for right now then and keep trying with the multiplier.

first previous 2 3 4 5 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob