new supply: Bridge with BIG IGBTS

This forum is for specialized infomation important to the construction and safe operation of the high voltage electrical supplies and related circuitry needed for fusor operation.
User avatar
Mike Beauford
Posts: 419
Joined: Wed Apr 25, 2007 2:24 pm
Real name:
Location: Morton Grove, IL

Re: why does my driver get hot

Post by Mike Beauford »

Hi Roman,

I ran into something similar with this circuit (See Attached Diagram). Richard mentioned this circuit a couple of months back, and I tried it out from a Nut's & Volt's magazine article. It turns out the circuit had a number of issues, one of which was it was drawing abnormally high current, the schematic was incorrect, which caused the main chip to overheated within a 10-15 seconds. I ended up going with the manufactures reference design, which worked like a charm. I have not looked up your chip's reference design, but I would highly suggest you take a look at it. Most of the time you can spot immediately what your doing different that might be causing the issue.

My 2 cents.

Mike Beauford
Attachments
QA-Figure5.pdf
(184.64 KiB) Downloaded 269 times
Mike Beauford
myID
Posts: 139
Joined: Wed Jun 25, 2008 2:16 pm
Real name:

Re: why does my driver get hot

Post by myID »

Hi-
it is the non-inverting One getting hot.
Thanks anyways!
Roman
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: why does my driver get hot

Post by Jon Rosenstiel »

FWIW… In the process of building my SLR driver I noticed that after I increased the number of turns on the GDT (used a slightly larger core and smaller wire) the gate drive IC's ran considerably cooler.

Jon R
Cliff S
Posts: 48
Joined: Fri Nov 21, 2008 1:25 pm
Real name:

Re: why does my driver get hot

Post by Cliff S »

Roman,

You need to look at the primary current and voltage waveforms. Indications are the transformer is saturating. As a sanity check, replace the transformer with a 100 ohm resistor (>3W rating). The drivers should be able to drive into 100 ohms without major heating. (Should be under 0.3W in the I.C. if the datasheets are accurate). If the I.C.’s are getting very hot even into 100 ohms, they you have a layout issue. Possibly some ground bouncing causing substrate latching/conduction in the I.C. But, I’d bet it’s just your transformer.

Listen, if you are doing all of this for fun, and a learning exercise, great. I’m sure you are learning a lot. But if you have more important things to spend your time on, send me your mailing address. I’ll ship you a gate drive PWB to drive your H-bridge. It’s painful for me to watch you go through this….


Best Regards,
Cliff Scapellati
Executive Vice President of Engineering
Spellman High Voltage Electronics
631-630-3110
myID
Posts: 139
Joined: Wed Jun 25, 2008 2:16 pm
Real name:

Re: why does my driver get hot

Post by myID »

Hi-

listening to all your tips and learning a lot I finally am pumping out some power.
I used a different IGBT board (from an old xray- supply) because I understood that brute force is not the way to go for now.
I built a driver with a TL494 giving me the waveform I need for driving a SLR output stage.
My GDT driver also works perfect now- evertything stays dead cold from 10kHz up to several 100kHz- the output from the IGBT board is a nice square wave without load.
I connected a big ferrite XFMR with a winding ratio of approx 1:9.
I choose a 60W 220V bulb as a load- and- it works. The output of the xfmr is approximately 8 times the input Voltage.
I easily can get up to 270V- then I stopped because I did not want to blow the bulb...
The only problem I still have is that my primary current still does not look like SLR.
This should not cause any problem so far because I have a pretty long dead time. I guess the reason for this is that I have no clue what Fres is. I do not have the ability to measure L and C of my circuit. On my IGBT board are rather big capacitors and serial inductivities so I assume Fres must be pretty low...
Looking forward when I will run with high Voltage/ Power I will find the correct settings first.
Whatever- thanks to everybody (especially Cliff...) for helping me.
Now comes the fun part of building a suited HF HV xfmr- we will see- and- sure: finding Fres...

Kind regards and thank you again
Roman

P.s.: I also bought a Bertan 105 30kV 1kW (I know low end Voltage wise) supply which I was able to fix with help from Cliff but still I want to learn more and build my own supply!
Attachments
IMG_3179.jpg
IMG_3184.jpg
IMG_3183.jpg
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: why does my driver get hot

Post by Jon Rosenstiel »

Hi Roman-

SLR topology doesn't like resistive loads, connect the hf transformer's output to a multiplier or, if you've got the sauce (and the nerve) a Jacobs ladder. Either scenario should result in a more normal current waveform.

Jon R
myID
Posts: 139
Joined: Wed Jun 25, 2008 2:16 pm
Real name:

SLR Power supply Xfmr

Post by myID »

Hi-

I am wondering if there is a rule of thumb for the SLR capacitor and inductivity of the primary. I know this is easy to calculate but what is a good combination of these values for frequencies around 50 kHz? A primary with around 4 - 10 Wdg seems to be in the ballpark? Or just when you use additional inductivities?
When I build a PE secondary holder with 2 big chambers (for center tap) each having 4 small chambers (approx 1kV per Chamber) will a GOOD xfmr wire stand that voltage?
All suggestions are really welcome because already up to here it was a big learning process and now starts what is considered "the black art" I guess.
As pointed out before- I have two big output stages properly designed (by somebody else) in oil already (hf hv xray tanks) but my goal is to build something on my own to learn- including a HV HF xfmr to drive my 1kW 40kV Glassman multiplier...


Sorry for changing the topic but I have the feeling it will be easer to keep the oversigt this way for all.

Kind regards
Roman
Darwin Wandler
Posts: 1
Joined: Sun May 12, 2024 12:09 am
Real name: Darwin Wandler

Re: new supply: Bridge with BIG IGBTS

Post by Darwin Wandler »

I am using 2 1200V 400A Powerex 1/2 bridge power blocks and the power required to drive the gates needs a full bridge IGBT driving a 1:4 gate driver transformer 1.5" ferrite with 10 primary windings and 4 10 winding secondary gate driver secondaries which is driven by the UCC37322 gate drivers. (could also use a Mosfet full bridge to drive IGBT gates) my output transformer is 20Turns 10 AWG on 4 core 4" diameter by 1"thick stacked core magnetic ferrite from Magnetics with a single winding secondary of 3/8" copper tube. The power blocks are on an aluminum water cooled heat sink.
Without the double bridge the UCC drivers just fried.
Also running at 30Khz 40Khz tops due to the long shutoff time of the IGBT's
240Vac input 40A at resonance
output tank has 3.75uF 1000A capacitor with a 7.8uH tank coil

I was thinking about Mosfets but the power levels on the output tank bridge are too high to spend the cash on what it would cost in fets

I got the Powerex 1/2 bridge IGBT blocks for 20$ each and bought 26 for experiments. Same with 1000A caps bought for 5$@ and got 50 and gate snubber caps. I didn't put snubber caps on first IGBT driver bridge and had random issues in startup. Put caps on driver gates and now system is very reliable.

Even if I screw up the power blocks will just conduct the current and blow my breaker. No harm no foul. Made for dummy mistakes.
the only way to get more speed out of the IGBT's is using the new silicon carbide IGBT's
I run the entire unit with a TL494 and soft start circuit driving 2 UCC37322 with the TL494 emitters driving the UCC37322 chips to keep them in a power off state until the system starts up. I think this eliminates the 322 getting hot issue due to always being on and the 321 off when oscillator not running.
Last issue was enough power to drive the intermediary bridge driver. I needed to get 15V 30amp all I could find without buying multiples until I got it right) power supply. The power to drive the monster Powerex blocks was an eye opener.
This was a fantastic learning experience without multiple blow ups.
Andrew Seltzman
Posts: 811
Joined: Sun Feb 01, 2004 8:02 pm
Real name: Andrew Seltzman
Contact:

Re: new supply: Bridge with BIG IGBTS

Post by Andrew Seltzman »

For those interested in building IGBT based supplies, this might be of interest. I build an -80kV, 40A power supply a few years ago (tested up to -100kV, 50A) using a 3 phase full-bridge IGBT layout driving loosely coupled resonant transformers.
http://www.rtftechnologies.org/emtech/r ... t-smps.htm
Andrew Seltzman
www.rtftechnologies.org
Post Reply

Return to “High Voltage - Fusor Input Power (& FAQs)”