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 8 9 10 11 next last

Move Thread

LAN_403

Shrad

Thu Oct 16 2014, 06:04PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

I think you'll see far less junk with a car battery indeed, and also with a dedicated scope as some computers can radiate too... laptops are not well earthed...

Electra

Thu Oct 16 2014, 07:38PM

Registered Member #816 Joined: Sun Jun 03 2007, 07:29PM
Location:
Posts: 156

The output of the mosfet (Drain), and Ground should look like a +13.8v dc level with Negative going pulses (just like the gate drive but inverted). Since you show positive going pulses in the last photo superimposed on some vaguely looking 60hz wave, One possibility is your circuit may working ok, but the issue may lay with the scope grounding. If you attach the scope ground clip anywhere else but ground, or leave it off you will get strange results.

The 100 ohm resistor should'nt be getting so hot at such low duty cycle either. So something to investigate too.

Patrick

Thu Oct 16 2014, 10:18PM

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

With the car battery i dont see any heat or switching. I may have killed the mosfet.

Electra wrote ...

The output of the mosfet (Drain), and Ground should look like a +13.8v dc level with Negative going pulses (just like the gate drive but inverted).

I thought of this while laying in bed, but now i cant see any signal across the load resistor. ill try grounding my laptop through the USB shield.

however a ground loop also worries me, if i ground with the scope, power supply and USB/laptop, thats 3 points trying to ground.

EDIT: FIXED IT !!!

Grounding the USB shield to the PSU negative terminal, imeadiatley turned it into the above wave form. And the resistor doesn't even get warm, the previous undulations must have created a higher duty cycle, and or not fully triggered MOSFET conduction. ( and youll notice the voltage switched is higher than any other voltage in the circuit, so its definatley PSU power going through)

Next up:

Differential measurements to be sure the resistor is seeing the wavform I want.
then, printing bobbins, and running a single loop primary.

EDIT2 : Finnally got some results !

Placed primary.

I don't know if this counts as 1.25 or 1 turn...

With out a safety resistor I think I blow out the MOSFETs with 200 Amp surges.

Here we see 10uS pulses of 13.5 V at 3.7kHz.

Here I think the primary series resistor limited the primary current too much, so the 1:3 ratio less than doubled the input voltage. 13.5V in, 23V out. I dont know why the output wave look like this, i expected a passive reset of the ferrite, but cant explain this or its short width.

in this screen cap, the secondary ouput flickers between 57 and 64 Volts. this voltage appears when I reduce the primary safety resistor to 4.1 ohms. The squeeeeeeeeeeeeeeeal gets louder too. now 13.5 x 3 turns = 40.5 volts theoretical, so i think some of the excess voltage is from kickback. As there maybe a tiny airgap, due to temporary hotglue holding the ferrite halves together.

Exactly 3 full turns of the secondary wire.

Seen in between the probe leads, a 680 ohm 1/2 watt resistor, across the secondary.

to the left, in series with the MOSFET and primary, a 4.1 ohm resistor. (15 || 10 || 10 = 4.1-ish ohms )

Patrick

Fri Oct 17 2014, 07:53PM

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

Im going to try all three cores later today (ive got to get my drone finished and graded by my professors soon first) but i still have some problems, as identified in the above post which i edited recently with new info.

Electra

Fri Oct 17 2014, 11:40PM

Registered Member #816 Joined: Sun Jun 03 2007, 07:29PM
Location:
Posts: 156

To get your LR time constant longer you'll need to reduce R.
t = L/R

If your mosfet can take 50A try R= 0.22 ohm. With only a 30v device you'll have to do something serious to clamp the back emf of the primary, or it'll almost certainly kill the mosfet by over voltage . A simple Zenner will probably be destroyed, unless you have a stupidly big one. If you got the secondary load calculated correct most of the energy should be transferred to that, but you always get some leakage.

Patrick

Sun Oct 19 2014, 08:21AM

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

so if your advising me to reduce it to 0.22 ohms, what is the T or L value you used to arrive at 0.22 ohms? i presume one of my oscillograms led you to this conclusion.

Electra

Sun Oct 19 2014, 07:52PM

Registered Member #816 Joined: Sun Jun 03 2007, 07:29PM
Location:
Posts: 156

The narrow pulses you see, is I think a result of the value of L being small, whether it is too small or not I don't know. For a conventional transformer the magnetizing current from the inductance might be something typically in the order of 5%, of the full load current.
I guess if you had a value for L, and you assumed it to be operating a continuous mode, the 10us would be for each half cycle I assume, so f= 50 khz you could get an aprox value of your reactance, and therefore the reactive current. (there's gota be a mathematical correct way using differentiation to work it out too).
A Value of 0.22 ohm (of a few watts) would limit the total peak current by default (ohms law), to well actually just over 60A in the worst case, with the psu still at 13.8 v. If you can't get a nice rectangular pulse through it by then. You need more turns after all, or reduce the pulse width and increase the frequency.
You don't need R once you are confident it works it only wastes power. Of course you know that.

I'm not suggesting running it unloaded, as it's not unlike a flyback as you have built it. Higher voltage FETs would have been ideal I suppose.
Don't mean to be critical in a negative way, just trying to input some ideas, even if I'm not always on the right track others can fill in the gaps.

Sulaiman

Sun Oct 19 2014, 08:16PM

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

With short pulses you should try to minimise leakage/stray inductance as much as possible,
this basically means having all conductors ( cap+ to coil to transistor to cap- ) as short as possible
and minimising the loop area that your conductors form / the current flows
putting the caps nearer/on the pcb will help

you may find a different primary winding better, marked in red;

I agree with the point about testing with a load,
reducing voltage stress on the switching transistors.

Patrick

Sun Oct 19 2014, 09:20PM

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

Sulaiman wrote ...

you may find a different primary winding better, marked in red;

I agree with the point about testing with a load,
reducing voltage stress on the switching transistors.

but i thought this path was less desirable.

Sulaiman

Sun Oct 19 2014, 09:50PM

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

You have the setup,
try the two and see the difference.

Then let us (me) know :)

I'm fairly confident that it will reduce leakage/parasitic/wasted external inductance
and if you consider the whole circuit for current, it's still one full turn for the single core

first previous 8 9 10 11 next last

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