Mazzilli ZVS Driver driven from more than 40V (how many times has this come up?) / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Mazzilli ZVS Driver driven from more than 40V (how many times has this come up?)

previous 1 2 3 next

Move Thread

LAN_403

...

Sat Apr 07 2007, 08:50PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

no, you need a powdered iron inductor. I have seen a few people make it work with ferrite, but the old 'yellow' cores are correct. There was an explanation buried within the old.4hv zvs flyback driver thread if you are curious...

Marko

Sat Apr 07 2007, 09:04PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Yellow inductor is actually very lossy, and here it is saved only because it concucts ripple with heavy DC bias.

I used ferrite too without problems but in that case rod is preferable due to DC bias and much more turns are required.

Sulaiman

Sat Apr 07 2007, 09:14PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

For this Royer/Marzilli/CFPR etc. invertor I have successfully used for the dc inductor
MPP cores (the best), Powdered Metal-, Ferrite- and Air-cores.

For a beginner I recommend ... Air-core

A reel of wire rated for the dc current is excellent
and will not saturate causing strange symptoms

An air-cored inductor is of course not ideal with respect to;

-efficiency,
for a given inductance/current a ferromagnetic core will be less lossy due to wire resistance

-cost,
an entire reel of wire costs more than a suitable core/wire inductor

-size,
a reel of suitably rated wire will be larger (probably) than the invertor transformer.

On the plus side is the lack of saturation effects,
and most of us have a (temporarily) spare reel of suitable wire.

For the less technically advanced it can be very difficult to design a suitable dc inductor.

uzzors2k

Sun Apr 08 2007, 10:06AM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

I tried a ferrite rod, but it shorted the PSU. So far the yellow choke hasn't given me any problems. The Mazzilli ZVS driver is pretty good, the only heating I'm getting is on the gate resistors, even the caps keep their cool. I upped the frequency to 71 kHz, as I'm too lazy to rewind the primary. That far from the flyback's resonance frequency the arcs are thin and weak, but at least the core isn't saturating. Getting good pics of arcs is tough, so I took a video clip too, it'll open with quicktime.
] 1176026578_95_FT23210_zvs_driver__25v__uzzors.zip[ /file]

1176026578 95 FT23210 Zvs Arc Uzzors 25v

1176026578 95 FT23210 Vert Arc Uzzors25v

Now I'll drop the frequency down 25kHz again, and make a 25+25 turn primary.

Tonic

Sun Apr 08 2007, 03:16PM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

On scheme there's a note saying that MOSFET's must have Rds smaller than 150 mOhm. What is this resistance and why it's that? I've heard about someone who built this driver with FETs with Rds larger than 150 mOhm and they've burned, but I still don't know why it occured. I'm asking because I'm thinking about giving separate supply (IR2153 driver pissed me out - I couldn't achieve any results, so I'm returning to simple Mazilli driver for a while) and replacing IRFP250 with another for higher voltage (there I have plenty of IRF450).

Sulaiman

Sun Apr 08 2007, 04:26PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

Rds means 'Resistance form Drain to Source'

What you are looking for is low Rds(on),
which is Rds when the transistor is fully ON

IRF450 has Rds(on) of 0.5 Ohm at 25 Celcius
but Rds(on) rises to 1 Ohm at a junction temperature of 150 Celcius

So for example, running a 'ZVS' invertor with say 5A dc
each transistor will pass 5A for 50% of the time
If Rds(on) is 1 Ohm then each transistor will dissipate 5^2 x 1 = 25Watts for 50% of the time
= 12.5 Watts average per transistor, this is called conduction loss.

In a ZVS there is not much switching loss (due to Zero-voltage-switching)
so you would budget 13 Watts heatsinking per transistor for 5 A dc.

IF you ran at 10A dc then each transistor would have a conduction loss of 50 Watts average etc. etc.

Myke

Sun Apr 08 2007, 04:30PM

Registered Member #540 Joined: Mon Feb 19 2007, 07:49PM
Location: MIT
Posts: 969

when the MOSFET is on it is the resistance between the source and the drain. I think it really has to do with heating of the MOSFETs at high powers.

uzzors2k

Mon Apr 09 2007, 12:44PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Meh, too many turns and the step up ratio is lost, higher resistance wire, all sorts of problems. The best preformance I have had is with 8.5+8.5t and a 5nf resonant cap. I'll have to get a better power supply before I can compete with blackplasma and "...".

I'm using IRFP450s with no problem. No heatsinks, and 30V of power during an arc. I can't measure the power since my multimeter is dead, but given that the mosfets are still lukewarm after long runs I assume the current passing through them is pretty low. The core isn't very big, just 130mm sq., so that limits current too.

How can I determine the power handling of a ferrite core if I know the cross-sectional area?

Dr. Dark Current

Mon Apr 09 2007, 03:23PM

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

Uzzors wrote ...

Meh, too many turns and the step up ratio is lost, higher resistance wire, all sorts of problems. The best preformance I have had is with 8.5+8.5t and a 5nf resonant cap.

5nf? and I thought 5uf is about optimal value

uzzors2k

Mon Apr 09 2007, 07:03PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Actually the capacitor forms the resonance frequency with the primary inductance, so the optimun value depends entirely on voltage, primary inductance and frequency. At low voltages with few turns a large resonant capacitor would be best.

EDIT: Whoops, the resonant cap is 50nf. 5nf would send the mazzilli driver into Tesla Coil frequencies...

previous 1 2 3 next

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