Proper IGBT choices and understanding types of IGBTs / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Proper IGBT choices and understanding types of IGBTs

Move Thread

LAN_403

Ken M.

Sun Oct 17 2010, 08:59PM

Registered Member #618 Joined: Sat Mar 31 2007, 04:15AM
Location: Us-Great Lakes
Posts: 628

With the thought of building a new SSTC, be it DRSSTC or SSTC I felt that I needed to do a little more background research on what semiconductor would be more efficient and handle the stresses of TC coil work.

I remembered I had some IGBTs from a few years a go and thought that I could use those, wanting to make sure they are suited for the work they would be subjected to I looked at the data sheet and to my dismay these IGBTs (fair-child G20N60B3D) have a horrendous turn on and turn off times -

Current Turn-On Delay Time td(ON)I TC = 150oC,
ICE = IC110
VCE = 0.8 BVCES,
VGE = 15V
RG = 10Î©,
L = 100Î¼H
- 25 - ns
Current Rise Time trI - 20 - ns
Current Turn-Off Delay Time td(OFF)I - 220 275 ns
Current Fall Time tfI - 140 175 ns

which is slower then the fabled 40n60's that are used on sub 100KHz coils and was thinking of using them on a coil that operates in the 200-300KHz, or 300-500KHz range.

With that I felt I needed to find a new IGBT to use knowing HVwiki would have some IGBT suggestions I looked it over and the usual 40N60's are there looked over the data-sheets to get a good idea of switching speeds and power limitations to look for in an IGBT. While looking through Digikey their filter system has the choice of PT, NPT, Field Stop, and Trench, as well as combination of them. I managed to find out some info and some over view of each but I can't seem to find much more info especially as far as comparison on Field stop vs Trench, so I was wondering if anyone had some info to explain the differences and benefits for our application regarding these types of igbts.

So far though I believe that NPT is better for TC use then PT due to them being a bit more rugged and because they are rated for Short circuit conditions which would be a benefit for TC use especially in the unfortunate situation of shot through, and Field stop is Better then NPT because it has a slightly lower switching and conductive losses but still doesn't out perform PT, other then FS IGBT's are short circuit rated.

Source for the comparisons on PT, NPT, FS vs Mosfet and each other:

Mostly ATM I'm just wondering if a Field stop trench IGBT would be a good choice although I still need to do a little more digging on trench IGBT's.

Steve Conner

Mon Oct 18 2010, 08:46AM

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

PT, NPT, field-stop, trench and so on, are all just different manufacturing processes. You can optimize all of them for either fast switching or low conduction loss. The IGBT "recipe" is like mixing together BJT and MOSFET: the more MOSFET you put in, the faster it goes, but the worse the conduction losses.

The "40N60"'s full part number is, well I can't remember, but it ends in "A4D". The "A4" is a different grade of IGBT die optimized for high speed. The "D" just means it has an antiparallel diode.

Your IGBTs are the "B3" series, which are optimized for low conduction loss.

I've used the "40N60" at 220-250kHz in DRSSTCs. Steve Ward has used them at 360kHz. The low duty cycle helps, you can run them at much higher frequencies than you could CW.

My record for CW power is about 1.5kW at 250kHz from a halfbridge of the Fairchild HGTG30N60A4D. That was running in a resonant inverter with zero current switching. They blew up after a few minutes at that power level, but the heatsink was pretty useless, which wouldn't have helped.

Short-circuit rating is irrelevant to TC use. It's only useful in systems that have desaturation protection. Essentially it means the device can withstand a short for a few microseconds, long enough for the desat protection to kick in and shut the system down.

Mads Barnkob

Mon Oct 18 2010, 02:28PM

Registered Member #1403 Joined: Tue Mar 18 2008, 06:05PM
Location: Denmark, Odense C
Posts: 1968

Current Turn-On Delay Time td(ON) - 25 - ns
Current Turn-Off Delay Time td(OFF)I - 220 275 ns

Either you mistyped or... these are not horrendous swtiching speeds, they are fairly fast...

Ken M.

Mon Oct 18 2010, 05:01PM

Registered Member #618 Joined: Sat Mar 31 2007, 04:15AM
Location: Us-Great Lakes
Posts: 628

Thanks Steve, I understood that the PT, NPT and so on were manufacturing methods, but I figured one method was more useful for what we'd use them for. Hence why I was trying to find out which type to look for.

Mads, yes they are fairly fast seeing as they would would switch faster then 2MHz but I believe thats for soft switching, and not including propagation delay which still isn't that much But I thinking of finding something closer to the 40N60 or better speed while keeping switching losses to minimum.

Oh and that's not a typo that was copy and pasted from the data-sheet.

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