Re: SSgap update 4-2

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Marco Denicolai by way of Terry Fritz <twftesla-at-uswest-dot-net>" <marco.denicolai-at-tellabs-dot-com>

Hi Terry,

please find mind comments below

Tesla list wrote:
> Using the IR2118 driver, I can really push the IGBT with solid signals.
> Below is the transmit LED vs. IGBT gate voltage:
> 
> http://hot-streamer-dot-com/TeslaCoils/MyCoils/SolidStateCoil/SSgap2001-4-3/IGBT
> Drive0.jpg
>

It looks SLOW! Something is wrong.

> 
> you can clearly see the IGBT gate turn on as a delay glitch at 300-400nS.
> The IGBT turn on and turn off is within 1uS of the logic signal's command
> back at the controller.  If I add 100 ohms to the gate drive the signal
> softens significantly:
> 
> http://hot-streamer-dot-com/TeslaCoils/MyCoils/SolidStateCoil/SSgap2001-4-3/IGBT
> Drive100.jpg

No way: it's not meant to be like that!
Terry, I used the IR2112 that employs the same charge pump and floating
supply. Also output current was the same, 200mA. With my 90 ohms gate
drive I have never seen such a slow gate drive. The capacitor between Vb
and Vs of your IR2118 must be at least 1 uF (microfarad). If that is ok,
check (and shorten) your gate wirings.

I see also you use some 18V for gate drive: I suggest going down to 15V,
as I did. If the gate gets to 20V you have a good chance of loosing your
IGBT (and that DID happen to me). 

Suggested readings (I have got several of them as PDF, if you need):

[Bax96]	Baxter, L.: Solving Noise Problems in High Power, High Frequency
Control IC Driven Power Stages. Motorola Semiconductor Engineering
Bulletin EB206, 1996.

[Ir936]	The Do's and Don'ts of Using MOS-Gated Transistors.
International Rectifier application note AN-936.

[Ir937]	Gate Drive Characteristics and Requirements for HEXFET®s.
International Rectifier application note AN-937.

[Ir973]	Chey, C., Parry, J.: Managing Transients in Control IC Driven
Power Stages. International Rectifier design tip DT 97-3.

[Ir978]	HV Floating MOS-Gate Driver IC. International Rectifier
application note AN-978.

[Ir983]	IGBT Characteristics. International Rectifier application note
AN-983.

[Ir990]	Application Characterization of IGBTs. International Rectifier
application note INT990.

[Mit95]	Mitter, C. S.: Application Considerations Using Insulated Gate
Bipolar Transistors (IGBTs). Motorola Semiconductor Application Note
AN1540, 1995.

[Tak95]	Takesuye, J., Deuty, S.: Introduction to Insulated Gate Bipolar
Transistors. Motorola Semiconductor Application Note AN1541, 1995.


> 
> I guess that may be a good thing since the really fast turn ons and offs
> may cause spikes and such that would be harmful.  The times are still well
> within what is required as long as heating does not start to be an issue.

Well, you know that V*I stuff, and I think your scope (or Matlab) can do
the multiplication. A moderate slow commutation is ok, but I'm afraid
your is too slow.


> 
> I have now run at 900volts and up to 40 amps with no problems.  I did blow
> one IGBT while fiddling with the scope buttons and trying to figure out if
> the collector voltage was 200 or 2000 volts...  It was 2000 ;-))  I also
> popped another while turning up the variac too fast and probably with poor
> drive signal.  In both cases it looks like the anti-parallel diode was
> blown short.  

If collector and emitter are shorted, you exceeded the maximum Vge.
Sometimes you can kill the IGBT by just probing its gate with the probe
tip. Sometimes it will die because it is a nasty #¤&"?!!&&**.

>I will have to wait till I get back to work to decapsulate
> the die and see for sure.  There is a chance the diode is not taking the
> powerful signals well.  It would not be hard to use an IGBT only device
> (half the cost) and maybe use a more conventional but hefty reverse diode.

You would like to do that with a MOSFET, because the intrinsic diode is
slow and dissipates heat. So you'd use a diode in series between Vcc and
drain, and another (fast) in antiparallel between Vcc and source. But
the IGBT diodes are surely fast (and soft) enough for you. A faster
diode would only arrange nice transients for you.

By the way, yesterday Thor hit with a streamer its RSG assembly (or
electrodes?) which wasn't grounded. Of course, one IGBT pair decided to
pass away... :(

Let me know if you want those PDFs.

Cheers

-- 
_____________________________________________________________

 Marco Denicolai           Senior Design Engineer 
 Tellabs Oy                tel: +358 9 4131 2769
 DSL Products              mobile: +358 50 353 9468
 Sinikalliontie 7          fax: +358 9 4131 2410
 02630 Espoo  FINLAND      email: marco.denicolai-at-tellabs-dot-com
_____________________________________________________________