max power for a flyback drivers? / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

max power for a flyback drivers?

1 2 3 next

Move Thread

LAN_403

cedric

Fri May 20 2011, 11:15AM

Registered Member #2941 Joined: Fri Jun 25 2010, 08:08AM
Location:
Posts: 143

hello everybody.
I am currently buying some big ferrite from china to build a big fly-back (the ferrite wait about 500 g)
I want to use it to produce a single high voltage pulse from low voltage input,
My plan is to use a big mosfet in serie with the primaries of the fly-back transformer to power it .
the output voltage will be Dependant of the duty-cycle of the mosfet.
the system have to work as a single event operation but be capable to operate at up to 20 khz.
the average output power should be in the kw range.
the question is :
is there a max practicable power for this topology?
in this set up the freewheeling diode is important to protect the mosfet ,but does it contribute to improve the performance of the fly-back by helping the field to collapse faster,or does it wast energies?

Shrad

Fri May 20 2011, 11:30AM

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

I'd use a MOT with a 2 to 5 turn primary and original secondary, and discharge a capacitor (bank) via a big SCR

you can achieve more than 10kV for one spark with a decent power this way, it hsa been successfuly used to trigger big laser tubes in the past

Ash Small

Fri May 20 2011, 12:44PM

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

There is a maximum power for this topology and it depends on the core material and size of core (cross sectional area).

The most efficient frequency and maximum frequency also depend on core material.

Some manufacturers publish this information, some do not.

Do you have manufacturers specifications for these cores?

If the cores are anything like the ones in the specs below, they should cope with power in the kW range at 20-25 kHz.

cedric

Fri May 20 2011, 01:17PM

Registered Member #2941 Joined: Fri Jun 25 2010, 08:08AM
Location:
Posts: 143

for those who are interested, here is the link for the core:

I am not worry about the core itself but more about the electronic,I don't see on the web any setup using one switch (a mosfet ) in serie with a fly-back,for more than 200 W of power ,I like to know why (I know a half bridge or full fridge configuration would be better,)do I simply have to scale up the mosfet max current rating?
in my set up ,the fly-back should never be allowed to resonate,all the energy from one pulse have to be completely dissipated before the next one ,so the pulse shape of the secondary stay the same for any frequency.
I have a working prototype witch operate at 60 W giving me a max frequency of around 1khz.
the load of the fly-back is a 1nf high voltage capacitor,in serie with a hv diode,after the pulse the voltage in the capacitor is 5kv.

Steve Conner

Fri May 20 2011, 02:58PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

High power flyback converters are less efficient than forward converters. If you don't care about high efficiency, then the flyback will be fine.

The freewheeling diode doesn't affect the collapsing field: it's the other way around. The collapse generates a positive drain voltage at the FET that adds to the supply voltage.

If you can do it with a forward converter, that would be better. Ie, the output pulse is generated while the MOSFET is ON, not when it switches off.

The klystron modulators for particle accelerators like SLAC and Fermilab are built that way, with Metglass cores, 3.3kV IGBT brick switches, and peak power outputs of many megawatts. (I don't know exactly how many :) )

Ash Small

Fri May 20 2011, 03:03PM

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

I'd suggest an IGBT may be better than a MOSFET.

IGBT's are generally better for high current and low frequencies.

But you do want a fast switch off, so a MOSFET may be better here.

I think the priority is the switch off time waveform, this should be as short as possible.

As far as I know, the freewheeling diode is just there to protect the MOSFET.

EDIT: a snubber accross the primary should eliminate ringing, I think.

cedric

Fri May 20 2011, 03:14PM

Registered Member #2941 Joined: Fri Jun 25 2010, 08:08AM
Location:
Posts: 143

thanks for the answers...
I may try with a forward configuration,but for me the fly-back look more simple (more voltage gain per turn on the secondary),
igbt have terrible turn of time compare to mosfet so I will stick to them...

Mattski

Fri May 20 2011, 04:05PM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

The selection of IGBT vs. MOSFET is dependent on the voltages and currents involved. IGBT's are certainly better at extremely high levels of each, but at lower power levels MOSFETs are better with smaller conduction and switching losses.

Forward converters are still quite simple to build, you need to do a little more winding to make a center-tapped primary and add the reset diode, but you still get the simplicity of having a single low side switching transistor.

Also with a forward converter you don't need to worry nearly as much about saturating the core since most of the primary winding current does not contribute to core flux, whereas in a flyback converter all of the primary current goes into magnetic flux. Means you can use a smaller core for the same amount of power.

Pinky's Brain

Fri May 20 2011, 04:29PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

You don't have quite the same design parameters converters do, output inductance is presumably more important to you (to get a fast pulse) than efficiency. So the theory behind converters doesn't seem entirely appropriate.

Without an air gap you're not going to approach the power of the forward transformer ...

Antonio

Sat May 21 2011, 01:50AM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

There is no free-wheeeling diode in a flyback converter. Use the "quasi resonant" configuration. The same used in most CRT TVs and monitors.

1 2 3 next

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