Author Topic: Large Ferrite Core with IGBT-Brick Driver  (Read 2105 times)

Offline Phoenix

  • High Voltage Experimenter
  • **
  • Posts: 63
  • Karma: +1/-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 Technician
  • ***
  • Posts: 100
  • Karma: +2/-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 Experimenter
  • **
  • Posts: 63
  • Karma: +1/-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 Technician
  • ***
  • Posts: 100
  • Karma: +2/-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 Expert
  • *****
  • Posts: 1016
  • Karma: +16/-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 Experimenter
  • **
  • Posts: 63
  • Karma: +1/-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 Technician
  • ***
  • Posts: 100
  • Karma: +2/-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 Experimenter
  • **
  • Posts: 71
  • Karma: +4/-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).



Offline Phoenix

  • High Voltage Experimenter
  • **
  • Posts: 63
  • Karma: +1/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #8 on: February 24, 2019, 12:11:30 AM »
Hello  :)

I have relaunched this project and improved some things. I now found the cause for the fast temperature rise after short run times. I did not use any gate resistors before, because the gate signal looked very good and the rise/fall times were the fastest. But now I checked the gate signal while the bridge was running and I was shocked. There were large oscillations on the gate signal which brought the IGBT back into the linear region! I now added 2.3 Ohms of gate resistance which seems to be the best value after some testing.

The high transients spikes also disappeared because of my larger snubber capacitance and slower rise/fall times.

Here you can see the halfbridge inverter:



Here is a small test video with a power consumption of around 8kVA:

Right now, I am using 5 primary turns which equates to 14µH inductance with the secondary shorted. This results in 180A of peak current flow in the primary. The current signal has a triangle waveform. The peak current is so high that the primary wires repel each other when the arc ignites  :o I am planning to run the whole setup with more power. I could reduce the primary turn count to 4, which would lead to 280A of peak current flow.

I would also like to run the bridge on 3 phases to get 565V bus voltage. But first, I need to build myself a 3 phase variac out of 3 variacs.

Greetings,
Phoenix

Offline PI.VSM

  • High Voltage Enthusiast
  • *
  • Posts: 1
  • Karma: +0/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #9 on: March 11, 2019, 01:16:44 AM »
Dear mate.

Could you please kindly tell me how do you calculate the value of the snubber capacitor?

Regards

Offline Phoenix

  • High Voltage Experimenter
  • **
  • Posts: 63
  • Karma: +1/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #10 on: March 11, 2019, 07:11:30 AM »
Hello PI.VSM  :)

I did not calculate how much I need, I just checked the output of the inverter and the DC Bus with my oscilloscope while the transformer was connected. With only 1 microfarad, there were some small transient spikes left, but with 2 microfarads they are all gone.

I have just finished building my 15kW 3 Phase Variac and I am just waiting for the large DC Bus Capacitors to arrive. I will have around 740 joules of energy stored in the new DC Bus caps at 660V.

Here you can see a test video of my variac:

Greetings
Phoenix

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1016
  • Karma: +16/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #11 on: March 11, 2019, 08:47:51 AM »
Dear mate.

Could you please kindly tell me how do you calculate the value of the snubber capacitor?

Regards

Hi PI.VSM and welcome to HVF

I wrote a chapter on selection and estimation of snubber capacitor values in my DRSSTC guide: http://kaizerpowerelectronics.dk/tesla-coils/drsstc-design-guide/snubber-capacitor/
http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline Phoenix

  • High Voltage Experimenter
  • **
  • Posts: 63
  • Karma: +1/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #12 on: March 14, 2019, 07:39:04 AM »
 Here is the upgraded inverter for 3 phases:




I am using 2 350V 6800 microfarad capacitors for the halfbridge, which store up to 830 Joules of energy. The full bridge rectifier can handle up to 75A.

Offline Phoenix

  • High Voltage Experimenter
  • **
  • Posts: 63
  • Karma: +1/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #13 on: March 14, 2019, 08:40:27 PM »
The test run today was a complete success. The input power was up to 14kW high.

Here you can see a video:

Here are some images of the arcs:





I am planning to run the inverter on even more power, something like 20kW

Offline Hydron

  • Administrator
  • High Voltage Engineer
  • *****
  • Posts: 231
  • Karma: +6/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #14 on: March 14, 2019, 10:00:55 PM »
Wow, that's some serious arcs coming from a relatively modest transformer! Definitely beats what I managed with 4 microwave transformers in series.

Have you considered removing the electrolytic smoothing cap? Might make it sound scarier to have the hum from the mains ripple.

Offline Phoenix

  • High Voltage Experimenter
  • **
  • Posts: 63
  • Karma: +1/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #15 on: March 14, 2019, 11:18:04 PM »
Thanks for your reply, Hydron  :)

Removing the smoothing caps is not really possible, because the transformer needs a lot of current delievered in a short period of time. There can be up to 200A flowing in the primary coil. If i hook it up directly to the mains, it could cause huge problems, because i am pulling such high currents at high frequency from it.

Greetings,
Phoenix

Offline Hydron

  • Administrator
  • High Voltage Engineer
  • *****
  • Posts: 231
  • Karma: +6/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #16 on: March 15, 2019, 09:27:48 AM »
You are right - I should have specified "and add a bit more film capacitor decoupling for the high frequencies".

I was thinking along the lines of induction cookers etc - they get good power factor by deliberately not including a large smoothing capacitor - high frequency currents are supported by smaller film caps but the overall input voltage follows the mains supply, leading to reasonable power factor.

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1016
  • Karma: +16/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #17 on: March 15, 2019, 10:31:10 AM »
Thanks for your reply, Hydron  :)

Removing the smoothing caps is not really possible, because the transformer needs a lot of current delievered in a short period of time. There can be up to 200A flowing in the primary coil. If i hook it up directly to the mains, it could cause huge problems, because i am pulling such high currents at high frequency from it.

Greetings,
Phoenix

That can indeed cause some problems, I tried doing this with my DRSSTC4, all details here: http://kaizerpowerelectronics.dk/tesla-coils/kaizer-drsstc-iv/

You are right - I should have specified "and add a bit more film capacitor decoupling for the high frequencies".

I was thinking along the lines of induction cookers etc - they get good power factor by deliberately not including a large smoothing capacitor - high frequency currents are supported by smaller film caps but the overall input voltage follows the mains supply, leading to reasonable power factor.

That only works for quasi-resonant inverters, so it kinds of stops at 2.3kW for a single 230VAC 10A outlet or you could go bigger on phase-phase 400VAC 16/32A or higher or with a huge step-up transformer, usually not very practical solutions compared to regular DCbus+inverter. But could be interesting to see a huge quasi-resonant single switch inverter doing 10kW :D
http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline rikkitikkitavi

  • High Voltage Enthusiast
  • *
  • Posts: 17
  • Karma: +1/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #18 on: March 16, 2019, 08:35:51 PM »
You can overload your fuses temporarily, but at a risk of overheating the wiring . How much depends on wiring and fuse type, but a typical 16A slow blow fuse can take 20Amps for quite some time.
It might be more or less illegal to do so, depending on country if done intentionally even if any elecitrcal motor would do so temporarily upon power on

Looks like it is a 16A CEE plug though, but I guess Phoenix knwos exactly how much his fuesbox can take :)

Actually I think that the power line can stand the overload longer than the electrodes...

BTW - nice work. I like your current limit circuit for the variac. But wouldnt it benefit from a larger wheel, to enable faster adjustment?
« Last Edit: March 16, 2019, 08:38:54 PM by rikkitikkitavi »
A man can not have to many variacs

Offline LongLiveRocknRoll

  • High Voltage Enthusiast
  • *
  • Posts: 1
  • Karma: +0/-0
    • View Profile
Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #19 on: May 11, 2019, 04:21:44 PM »
Here is the upgraded inverter for 3 phases:




I am using 2 350V 6800 microfarad capacitors for the halfbridge, which store up to 830 Joules of energy. The full bridge rectifier can handle up to 75A.

It's totally AWESOME!!

Could you show schematics for upgraded 3 phase-inverter please?

High Voltage Forum

Re: Large Ferrite Core with IGBT-Brick Driver
« Reply #19 on: May 11, 2019, 04:21:44 PM »

 


* Recent Topics and Posts

post Re: Choke-input DC-filter
[Electronic circuits]
johnf
Today at 08:58:07 PM
post Re: CRT hacking ideas....
[Electronic circuits]
badpeter
Today at 07:28:00 PM
post Re: Ideas needed!
[General chatting]
MRMILSTAR
Today at 05:13:57 PM
post Re: Why did 4HV die??
[General chatting]
MRMILSTAR
Today at 05:07:41 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
Bert911
Today at 02:14:55 PM
post Re: Choke-input DC-filter
[Electronic circuits]
kamelryttarn
Today at 12:53:58 PM
post Re: Choke-input DC-filter
[Electronic circuits]
johnf
Today at 12:09:27 PM
post Choke-input DC-filter
[Electronic circuits]
kamelryttarn
Today at 11:43:58 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
shrad
Today at 08:57:42 AM
post Re: adventures in homemade flyback bobbins.
[Transformer (ferrite core)]
Patrick
Today at 04:06:00 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
johnf
Today at 03:50:56 AM
post Re: CRT hacking ideas....
[Electronic circuits]
petespaco
June 19, 2019, 11:29:28 PM
post Re: A dynamical arc model v2
[Dual Resonant Solid State Tesla coils]
Uspring
June 19, 2019, 01:40:16 PM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
Mads Barnkob
June 19, 2019, 10:49:45 AM
post Re: A dynamical arc model v2
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
June 19, 2019, 10:43:32 AM
post Re: Light, laser and optics - Link collection
[Light, lasers and optics]
shrad
June 19, 2019, 10:03:54 AM
post Re: How go get single mode optic fiber to work
[Light, lasers and optics]
shrad
June 19, 2019, 10:02:21 AM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General chatting]
shrad
June 19, 2019, 09:35:52 AM
post Re: CRT hacking ideas....
[Electronic circuits]
badpeter
June 18, 2019, 11:35:16 PM
post Re: CRT hacking ideas....
[Electronic circuits]
Mads Barnkob
June 18, 2019, 08:22:23 PM
post CRT hacking ideas....
[Electronic circuits]
badpeter
June 18, 2019, 07:20:16 PM
post Re: Is one of these good?
[Dual Resonant Solid State Tesla coils]
VNTC
June 18, 2019, 12:16:43 PM
post Re: A dynamical arc model v2
[Dual Resonant Solid State Tesla coils]
Uspring
June 18, 2019, 11:45:53 AM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils]
Netzpfuscher
June 17, 2019, 09:01:29 PM
post Re: A dynamical arc model v2
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
June 17, 2019, 08:54:41 PM
post Re: Bought huge capacitors
[Capacitor banks]
MRMILSTAR
June 17, 2019, 04:34:02 PM
post Re: How energy flow through a DRSSTC?
[Dual Resonant Solid State Tesla coils]
Uspring
June 17, 2019, 03:13:03 PM
post Re: A dynamical arc model v2
[Dual Resonant Solid State Tesla coils]
Uspring
June 17, 2019, 02:40:06 PM
post Re: What happens when L and C are not matched?
[Electronic circuits]
klugesmith
June 17, 2019, 01:17:36 PM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
June 17, 2019, 11:13:33 AM
post Re: Where to buy Tesla coil capacitors?
[Spark gap Tesla coils]
Mads Barnkob
June 17, 2019, 10:46:56 AM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
Bert911
June 17, 2019, 08:01:10 AM
post Re: A dynamical arc model v2
[Dual Resonant Solid State Tesla coils]
VNTC
June 17, 2019, 04:19:07 AM
post How energy flow through a DRSSTC?
[Dual Resonant Solid State Tesla coils]
VNTC
June 17, 2019, 04:05:33 AM
post Re: Tungsten
[Spark gap Tesla coils]
thedoc298
June 17, 2019, 12:02:39 AM
post Re: My first DRSSTC on bricks
[Dual Resonant Solid State Tesla coils]
Laci
June 16, 2019, 09:00:31 PM
post Re: Bought huge capacitors
[Capacitor banks]
dexter
June 16, 2019, 08:25:16 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
hightemp1
June 16, 2019, 07:06:48 PM
post Re: Bought huge capacitors
[Capacitor banks]
MRMILSTAR
June 16, 2019, 06:07:06 PM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils]
Hydron
June 16, 2019, 04:51:26 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
petespaco
June 16, 2019, 03:39:30 PM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils]
Netzpfuscher
June 16, 2019, 03:39:15 PM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils]
Hydron
June 16, 2019, 03:18:25 PM
post Re: A dynamical arc model v2
[Dual Resonant Solid State Tesla coils]
Hydron
June 16, 2019, 03:10:15 PM
post Re: Bought huge capacitors
[Capacitor banks]
Max
June 16, 2019, 12:25:59 PM
post Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
[Electronic circuits]
Bert911
June 16, 2019, 09:38:15 AM
post Re: Top of secondary
[Spark gap Tesla coils]
MRMILSTAR
June 16, 2019, 05:45:47 AM
post Re: Tungsten
[Spark gap Tesla coils]
MRMILSTAR
June 16, 2019, 05:40:52 AM
post Tungsten
[Spark gap Tesla coils]
thedoc298
June 16, 2019, 03:51:12 AM
post Re: Last question
[Spark gap Tesla coils]
thedoc298
June 16, 2019, 03:44:43 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