Author Topic: Large Ferrite Core with IGBT-Brick Driver  (Read 307 times)

Offline Phoenix

  • High Voltage Enthusiast
  • *
  • Posts: 40
  • Karma: +0/-0
    • View Profile
Large Ferrite Core with IGBT-Brick Driver
« on: July 17, 2018, 07:19:13 PM »
Hello  :)

I wanted to share my latest project on this forum so here is my post about it:

The schematic for the GDT Driver is a modified version of the 1.3b DRSSTC Driver. I simply removed all the components for the Interrupter and OCD path. The IGBT-Brick is a Semikron SKM400GB126D Halfbridge. The transformer has 5 primary turns and 530 secondary turns using 0.28mm copper wire.

Here is a picture of the Halfbridge:


The large ferrite core (it consists of 2 U-Cores and 2 I-Cores):


The only issue i had so far is that one TVS-Diode blew up. This happened because i am using 440V TVS-Diodes and my input voltage is 380V DC high. So if there are any transients above 60V the Diodes will blow up. I am going to put 2 Diodes in series to fix this problem. The IGBT Brick ist still fine after the explosion.


Here i have a video. The power consumption is about 6500W high in this video. I estimate the arcs to be about 70cm long in total.

And here i have some arc pictures:










Greetings,
Phoenix

Offline Teravolt

  • High Voltage Enthusiast
  • *
  • Posts: 34
  • Karma: +0/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #1 on: July 24, 2018, 08:53:47 PM »
hi Phoenix nice work, how many turns on primary and secondary and where did you get that geat ferite core material? did you use a resonant cap on the primary like in ZVS circuits

Offline Phoenix

  • High Voltage Enthusiast
  • *
  • Posts: 40
  • Karma: +0/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #2 on: July 24, 2018, 11:04:54 PM »
Hello Teravolt

Thank you for your reply  :)

The Primary has 5 turns and the secondary has around 530 turns of 0.28mm wire. I got the ferrite cores from Mouser. I am using 2 I-Cores and 2 U-Cores for the large, finished core.
Here are the part numbers:
B67345B0001X087 (U-Core)
B67345B0002X087 (I-Core)

But remember that i had to carefully grind of some material from the I-Cores because they are not made for such an application. There was a small airgap around 100 µm wide but i managed to decrease the gap to about 15 µm using sandpaper and a glass pane. This increased my output by a little bit and the whole core is more stable now.

There are no resonant capacitors like in a ZVS, just the output of the halfbridge fed directly into the primary. Keep in mind that your IGBTs/Mosfets need a bodydiode (built in or external), to absorb the Back-EMF of the primary.

I have also improved the layout of my DC-Bus to decrease transients:


Greetings
Phoenix

Offline Teravolt

  • High Voltage Enthusiast
  • *
  • Posts: 34
  • Karma: +0/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #3 on: July 25, 2018, 04:24:52 AM »
I have seen this experiment done before on 4hv and tried it unsuccessfully my self. did you have to adjust your oscillator and pulse width to maximize the arc?
some times there is a thin spacer added between the mating faces of the cores maybe to prevent eddy currents. A resonant cap in parallel with the primary your primary might reduce the average power you need. Do you plan on caring your experiment further?

Offline Mads Barnkob

  • Administrator
  • High Voltage Senior
  • *****
  • Posts: 692
  • Karma: +4/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #4 on: July 25, 2018, 12:47:10 PM »
You got some nice pictures there, good and simple project.

I got a similar project, stranded in a box for some years now, to drive some old x-ray transformers and these all uses a capacitor in series with the primary coil, if I remember correct, is not a resonant capacitor, but a DC-blocking capacitor to prevent the bridge from blowing up at core saturation?

I am rather interested in your MMC (this time Multi Mini Core) design, it does not seem like you have any clamps or anything else holding it in place, except gravity. I have a lot of brick ferrite bars and could virtually build any shape of transformer by stacking them, but I have not done any of that so far as I did not think it would work that well and know how to hold the construction together.

http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline Phoenix

  • High Voltage Enthusiast
  • *
  • Posts: 40
  • Karma: +0/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #5 on: July 25, 2018, 01:58:11 PM »
@Teravolt

I think you need to adjust your oscillator to exactly 50% Duty-Cycle to make the whole thing work properly. Are you using a TL494 oscillator? I am using a small, portable function generator to provide the signal. Currently, the setup is running on 20kHz, but i can go down to 8kHz before my GDT saturates. You need to be careful with your GDT and you have to make sure, that the GDT does not saturate with lower drive frequencies. Here you can calculate how many turns you need on your GDT using the information in the datasheet. It turned out to be fairly accurate for me: http://thedatastream.4hv.org/gdt_saturation.htm

If i wanted to use a resonance circuit, i would have to make sure that I drive the LC circuit with its own resonance frequency. The ZVS is self-oscillating so the resonance frequency is always hit automatically. I am also worried about high impulse currents because of the high input voltage (380V DC in my case, a ZVS runs on lower voltage) generated in the LC circuit. Such a event could get quite destructive.

Here i have the schematic of the modified 1.3b Driver in case you are interested:


@Mads

The only problem i had so far is that my bridge is getting quite warm even after short runtimes. My switching speeds are not bad, the rise/falltimes at the bridge-output are 80ns fast. Cross conduction can't be the reason either, because the inverter pulls no current when there is no load connected while the gatedrive is running. In your opinion, how hot can a heatsink get in an hard-switched inverter? I think it is normal for the IGBT-Bricks to get hotter because of the hardswitching. I am used to the stone-cold heatsink of my DRSSTC (resonant switching), so i am a little bit worried. I have already ordered a fan to keep the heatsink temperatures down, so far i did not let the heatsink get over 40°C.

I also think that this capacitor is for DC-blocking. What is the value of the capacitor? I do not think that the capacitor is a resonant capacitor like in a DRSSTC, because such a resonance circuit in an hard-switched inverter in CW mode would be very damaging for the IGBT´s.

The ferrite core is only held together by gravity, but the PVC-pipe of the secondary guides it in place. I believe that you could glue together the cores using some epoxy glue with mixed in ferrite-dust. If you want to separate the cores again, you could carefully heat the ferrite parts with a heatgun until the epoxy softens again.
« Last Edit: July 25, 2018, 02:17:40 PM by Phoenix »

Offline Teravolt

  • High Voltage Enthusiast
  • *
  • Posts: 34
  • Karma: +0/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #6 on: July 25, 2018, 09:00:08 PM »
I ordered a cuple of cores to mess with. does your transformer have any sort of resonance or do you think it is acting like just turns ratio? when I look at the inital arc it seems to be about 5-10kv with a lot of current behind it?

Offline Netzpfuscher

  • High Voltage Enthusiast
  • *
  • Posts: 26
  • Karma: +0/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #7 on: July 26, 2018, 07:58:11 AM »
A while ago I've build a ferrite transformer too. But I used the SLR-Topology described here: https://www.stevehv.4hv.org/ccps1.htm
This topology is acts as a current source, so its short circuit proof, soft switching, due to the waveform there is no catastrophic current rise like in the drsstc for example and the insulation of the transformer winding is much easier because of the sinusoidal current waveform. I have pushed easily 2kW through simple 20A IGBTs with no big heat. The tuning is easy and there is no failure if you set the resonant frequency wrong, it only transfers not so much power and the softswitching is gone if you set the frequency to high (not for to low frequency).



High Voltage Forum

Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #7 on: July 26, 2018, 07:58:11 AM »

 


* Recent Topics and Posts

post Re: Slow rise & fall time
[Solid state Tesla coils]
Laci
August 13, 2018, 10:20:43 PM
post Re: EMI & placement of driver, interrupter and MIDI controller
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
August 13, 2018, 11:29:30 AM
post Do you really know what 3, 6 and 9 stand for?
[General chatting]
Downunder35m
August 13, 2018, 05:19:03 AM
post Re: Slow rise & fall time
[Solid state Tesla coils]
the_anomaly
August 13, 2018, 02:18:42 AM
post Re: Slow rise & fall time
[Solid state Tesla coils]
T3sl4co1l
August 12, 2018, 11:35:33 PM
post Re: Slow rise & fall time
[Solid state Tesla coils]
the_anomaly
August 12, 2018, 10:25:17 PM
post Re: Slow rise & fall time
[Solid state Tesla coils]
T3sl4co1l
August 11, 2018, 10:19:18 PM
post Re: Slow rise & fall time
[Solid state Tesla coils]
Downunder35m
August 11, 2018, 04:31:42 AM
post Re: Slow rise & fall time
[Solid state Tesla coils]
the_anomaly
August 10, 2018, 10:36:15 PM
post EMI & placement of driver, interrupter and MIDI controller
[Dual Resonant Solid State Tesla coils]
DaiKen
August 10, 2018, 09:34:15 AM
post Re: Slow rise & fall time
[Solid state Tesla coils]
Downunder35m
August 10, 2018, 05:18:22 AM
post Re: Slow rise & fall time
[Solid state Tesla coils]
TDAF
August 10, 2018, 02:30:23 AM
post Re: Slow rise & fall time
[Solid state Tesla coils]
Mads Barnkob
August 10, 2018, 12:00:38 AM
post Re: Slow rise & fall time
[Solid state Tesla coils]
Laci
August 09, 2018, 10:07:01 PM
post Re: Slow rise & fall time
[Solid state Tesla coils]
T3sl4co1l
August 09, 2018, 09:53:57 PM
post Re: Slow rise & fall time
[Solid state Tesla coils]
Downunder35m
August 09, 2018, 07:11:40 PM
post Re: Slow rise & fall time
[Solid state Tesla coils]
Hydron
August 09, 2018, 06:21:27 PM
post Re: Slow rise & fall time
[Solid state Tesla coils]
Mads Barnkob
August 09, 2018, 03:07:50 PM
post Re: Slow rise & fall time
[Solid state Tesla coils]
Laci
August 09, 2018, 02:52:15 PM
post Teardown of a Eaton PowerWare 60kVA UPS system
[Electronic circuits]
Mads Barnkob
August 09, 2018, 12:25:32 PM
post Re: Slow rise & fall time
[Solid state Tesla coils]
Mads Barnkob
August 09, 2018, 09:10:09 AM
post Re: Slow rise & fall time
[Solid state Tesla coils]
oneKone
August 09, 2018, 04:23:26 AM
post Re: Slow rise & fall time
[Solid state Tesla coils]
TDAF
August 09, 2018, 02:30:24 AM
post Re: Slow rise & fall time
[Solid state Tesla coils]
Downunder35m
August 08, 2018, 11:34:19 PM
post Slow rise & fall time
[Solid state Tesla coils]
Laci
August 08, 2018, 10:08:48 PM
post Re: Drsstc 3
[Dual Resonant Solid State Tesla coils]
oneKone
August 08, 2018, 06:54:28 AM
post Trying to understand the inner workings of the resonance between coils
[General chatting]
Downunder35m
August 07, 2018, 11:52:20 PM
post Re: Primary coil designs and configurations
[Solid state Tesla coils]
Downunder35m
August 07, 2018, 04:27:26 AM
post Re: Is it possible...
[Light, lasers and optics]
profdc9
August 07, 2018, 03:21:31 AM
post Re: Drsstc 3
[Dual Resonant Solid State Tesla coils]
oneKone
August 07, 2018, 12:58:50 AM
post Re: Is it possible...
[Light, lasers and optics]
Hydron
August 06, 2018, 11:08:25 PM
post Re: Is it possible...
[Light, lasers and optics]
Mads Barnkob
August 06, 2018, 08:37:39 AM
post Re: Primary coil designs and configurations
[Solid state Tesla coils]
Mads Barnkob
August 06, 2018, 07:52:15 AM
post Re: Is it possible...
[Light, lasers and optics]
profdc9
August 06, 2018, 03:20:35 AM
post Re: A little help with first Tesla Coil
[Solid state Tesla coils]
Teravolt
August 06, 2018, 02:44:53 AM
post Re: Is it possible...
[Light, lasers and optics]
FilipŠebík
August 05, 2018, 08:34:12 PM
post Is it possible...
[Light, lasers and optics]
Longranger
August 05, 2018, 07:40:01 PM
post Re: How to machine end-caps for secondary coil
[General chatting]
Downunder35m
August 05, 2018, 07:37:22 PM
post Re: How to machine end-caps for secondary coil
[General chatting]
TDAF
August 05, 2018, 06:13:11 PM
post Re: How to machine end-caps for secondary coil
[General chatting]
Downunder35m
August 04, 2018, 07:25:14 PM
post Re: Isn't there a better way to make a custom toroid?
[General chatting]
Nijin
August 04, 2018, 07:22:38 PM
post Re: Isn't there a better way to make a custom toroid?
[General chatting]
Downunder35m
August 04, 2018, 07:16:57 PM
post Re: Primary coil designs and configurations
[Solid state Tesla coils]
Downunder35m
August 04, 2018, 07:01:16 PM
post Re: How to machine end-caps for secondary coil
[General chatting]
Hydron
August 04, 2018, 05:35:03 PM
post Re: Primary coil designs and configurations
[Solid state Tesla coils]
Teravolt
August 04, 2018, 04:55:04 PM
post Re: Choosing Primary impedance and Fres
[Dual Resonant Solid State Tesla coils]
Nijin
August 04, 2018, 04:48:00 PM
post Re: Copenhagen Suborbitals: Launch of Nexø II rocket (new date! 4th August)
[Science, research and news in other fields than electronics]
Mads Barnkob
August 04, 2018, 04:32:49 PM
post Re: Choosing Primary impedance and Fres
[Dual Resonant Solid State Tesla coils]
Hydron
August 04, 2018, 12:47:06 PM
post Re: Choosing Primary impedance and Fres
[Dual Resonant Solid State Tesla coils]
Nijin
August 04, 2018, 12:15:39 PM
post Re: Choosing Primary impedance and Fres
[Dual Resonant Solid State Tesla coils]
Uspring
August 04, 2018, 11:44:50 AM

Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30