High Voltage Planar Ferrite transformers. ( Intial Experiments ) / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

High Voltage Planar Ferrite transformers. ( Intial Experiments )

first previous 11 12 13 14 next

Move Thread

LAN_403

Ash Small

Tue Oct 28 2014, 12:08PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I've not read the links JP posted yet, as I've been away for the weekend, but the 'voltage adder' thingy is what I'm using in the project I linked to above somewhere. I'm using parallel MOSFET's too, but I'm currently waiting for more HV high speed MOSFET's to become affordable. (at least 1kV breakdown voltage and sub 10nS switching times)

There are some links to particle accelerator pulse generators in my thread.

I'll read through the links later, and maybe post again.

Patrick

Wed Oct 29 2014, 02:30AM

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

Ash Small wrote ...

I'm using parallel MOSFET's too, but I'm currently waiting for more HV high speed MOSFET's to become affordable. (at least 1kV breakdown voltage and sub 10nS switching times)

Yeah this is what I was waiting for but 1Kv start getting ineffcient, and slow, this was the case in the 90's, and early 00's. Now MOSFETs are so much better in Rds and cost is unbelievably low, but its mainly in the sub 500 volt , or better still sub 100 volt region where MOSFETs really shine.

Ash Small

Wed Oct 29 2014, 11:12AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Patrick wrote ...

Ash Small wrote ...

I'm using parallel MOSFET's too, but I'm currently waiting for more HV high speed MOSFET's to become affordable. (at least 1kV breakdown voltage and sub 10nS switching times)

One reason for paralleling them is to reduce Rds. Also, I mentioned above somewhere that the 'voltage adder' curcuits I've seen (see links in my thread) all have capacitors at the MOSFET which only partially discharge during the MOSFET 'on' time. Each MOSFET switches it's own capacitor, which presumably also aids 'load sharing'.

Patrick

Fri Oct 31 2014, 05:35AM

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

Ill be etching new boards soon, these will have 5 transformers, for a total of 60 volts per turn. with 200 secondary turns this will mean 12000 volts.

Sulaiman

Fri Oct 31 2014, 06:04AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Patrick,
before you start the new pcb layout
read the post from our resident smpsu guru GeordieBoy (Richie)

some of the best, most concise guidance ever.
some of it applies to your application

Dr. Slack

Fri Oct 31 2014, 09:13AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

Hi Patrick,

Inspired by your multi-core layout, I began revisiting my old design for a direct transformer driven magnifier TC. This is basically an SSTC, however the SSTC, having an air-cored transformer, has a very high magnetising current. The popular solution to this these days is to resonate the primary to make a DRSSTC. However, I hanker after reducing the magnetising current by coring the transformer, to leave something that's simpler to drive.

Transformer spec is 200kHz. 50kV output. With the largest easily commercially available ferrite core being UR64, 290mm2, that could give around 50v per turn when pushed, so 1000 secondary turns does it. I considered a design splitting it as 100 turns on each of 10 cores.

Now, the question is - arrange the 10 cores mechanically like yours and wind 100 secondary turns round the lot, or wind 10 individual 100 turn transformers and put their secondaries in series?

While they have identical first order parameters - volts per turn, total volts, nominal power, insulation needed from hot secondary wire to grounded core, I didn't need to go very far down the thought experiment to realise one significant second order parameter was very different, residual capacitance. With a single secondary winding, the hot output turns are in the same vicinity as the groundy turns, and the secondary winding diameter is considerable, meaning a lot of parallel wire capacitance, and a considerable length of HV wire passing through a lot of grounded cores. With the individual secondaries, each section only has a small voltage across it, meaning the capacitance is much lower. The early radio people thought the same way, you would see an inductor wound from several flat windings connected in series with airspaces between to reduce the capacitance.

Now, I don't have time to do any experimental work, and frankly not even enough to model what's happening in sufficient detail to get a numerical result for my handwaving argument. But I think the difference between the two may be enough for you to want to consider carefully. I expect the difference would increase as the number of cores increase.

Before you embark on any serious amount of building, I would recommend you build two transformer systems, one with a secondary round two cores, the other with two individually wound cores with discrete secondaries in series, and see if their resonant frequencies are significantly different. It occurs to me that you don't even need to implement the primaries or test at high power, measuring the resonant frequency of the secondary inductance would be enough.

Patrick

Fri Oct 31 2014, 06:27PM

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

Suilaiman,
I visit Geordie's site often, and have for years, it was a huge guiding factor in me oing 4HV, and my other succesful projects. Ill see if he's posted anything new.

Dr. Spark,
Fair points, and they worry me. but with 200 secondary turns spaced out from inside grounded groove to HV top groove, i didnt think would have much capacitence.

if it does, that really sabotages things. ill still go through with a 5 core proto, becuase that will confirm or deny problems for sure. I do have a working HV oscilloscope probe upto 40Kv.

But let me ask, why are 10,000 turns in a TV flyback not a problem? they work at 15Hz and more.

Sulaiman

Fri Oct 31 2014, 07:28PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

To get high volts/turn from ferrites you need high frequencies
the capacitance of the windings form a resonant circuit with their inductance
and operation above the resonsnt frequency would be difficult
and the capacitance(s) are non-ideal / lossy so heat would be generated by dielectric losses.

your windings are close to the primary and ferrite so relatively high capacitance

As an exercise, assume 'stray' capacitance from each winding to core is Cs
and Ns turns secondary winding, each core set having inductance factor Al.

For 5 cores with one winding;
L = Ns^2 x 5 x Al
C = 5 x Cs
so Fo = 1/2.pi.sqrt(5.Ns^2.Al x 5.Cs) = 1/2.pi.sqrt(N^2 x Cs x 25)

For each of 5 separate cores with Ns turns each;
L = Ns^2 x Al
C = Cs
so Fo = 1/2.pi.sqrt(Ns^2.Al.Cs)
There will be five series connected transformer secondaries,
each with a resonant frequency five times higher than one large multi-section transformer.
This is why separate transformers are used for a 'voltage additive' system .....

Also, physical problems such as bending/flexing are less with five separate transformers
and more compact (yet still well cooled) physical arrangements are easier with five separate cores/windings,
and a lot cheaper to replace one of five failed units rather than one of one units.

Dr. Slack

Fri Oct 31 2014, 08:07PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

A lower Fres needn't sabotage things, but if you're close to the limit, then every little helps.

Patrick

Sat Nov 01 2014, 02:22AM

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

but if i series the coils, (and the alcholL makez this ThingkIng hard) i dont know if itworks the same?

lets see:
12 x 200 = 2,400 x 5 = 12,000

12 x 5 x 200 = 12,000

ok i see now. but the last coil still sees the worst Vpk and Cs. but i guess its better than the whole coil around being bad.

EDIT: but how do i get the "jumper" from teh outer coil of one to the interior or the next coil, with out blowing out the second coil?

first previous 11 12 13 14 next

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