Thanks for the reply Chris.
The application is for a 48VDc bus and 10A continuous.
I hope I can salvage the design with the IR21362.
I have a Power Ground plane on the PCB.
All the low side returns come back to a 0.02 four-terminal current sense
resistor in a star configuration.(2oz copper and 150mil wide tracks)
My driver is as close to the FETS as possible. All my bootstrap components
are as close to the 21362 pins as possible. I see very little noise on my
PWM output waveforms except for this nasty 50ns long 7V spike just before
the top FEt turns off. I can get my motors spinning and my load moving fine.
I am just worried about the long term reliability with the FETs running hot
and EMI from shoot through currents. Any ideas? I appreciate your time and I
can send you the scope trace if you have the time to look at it.

As regards the DC/DC and HP driver configuration, are there any specific
chips you are comfortable with? I would appreciate any hints.
Thanks again
Madhu


-----Original Message-----
From: pic microcontroller discussion list
[mailto:PICLIST@MITVMA.MIT.EDU]On Behalf Of Chris Eddy
Sent: Monday, September 17, 2001 11:22 AM
To: PICLIST@MITVMA.MIT.EDU
Subject: Re: [EE]: IR21362 & IRF540N


Yea, I been down this road.  The first thing that you must do is throw
the IRF bridge driver away. I was using their single channel devices for
a three phase brushless design. The grounding and layout requirements,
in my opinion, are not achievable, and when the high side bootstrap
driver fails it fails latched. You know what hapens next. I burnt
(smoke, sparks, flame) two prototypes before abandonning ship. I have
heard others that reject the IR high side drive chips outright. Use a
DC/DC and HP driver designed for the job. What is the current/power?

On the funky gate spikes, you absolutely must lay ground out in a
specific manner. If this driver chip is located remotely (and it is
since all six gates come from your chip) then you are just asking for
trouble. Solid and short grounds are manditory, and star configurations
where possible. If all this is right, and you get spikes, there are
strategies, just zap your results back onto the list and we can take
another stab.

Sorry for the cold water.
Chris~

Madhu Annapragada wrote:
>
> Anybody have any experience with the above combination? The 21362 is a
nice
> bridge driver and the IRF540 is a fairly common N-Channel MOSFET. I am
using
> this in a three phase Brushless DC application (20KHz PWM) and the trouble
I
> am having is that I am getting shoot-through currents. The 21362 is
supposed
> to introduce a 100ns Dead time between the top and bottom MOSFETS but I am
> getting this weird spike on the bottom MOSFET gate just before the top FET
> Gate-Source voltage goes to zero.  This spike (amplitude of 7V) is enough
to
> turn on the bottom FET before the top one fully turns off and consequently
> the bridge is heating up even under no-load conditions. I have enough
> bootstrap capacitance (10uF tantalum and a 0.01uF X7R ceramic in
parallel),
> fast bootstrap diodes (IN4148W) and even a free-wheeling diode between
drain
> and source of all the FETS(50ns trr).
>
> Any ideas on why the bottom FET gate-source voltage is showing a spike of
> about 7V way before (30-40ns) the bottom FET is supposed to turn on? Any
> insights would be greatly appreciated. IRF applications engineering has
not
> yet come up with an explanation and I am hoping that someone on the PIC
list
> might have gone down this road before.
> Thanks in advance.
> Madhu
>
> --
> http://www.piclist.com#nomail Going offline? Don't AutoReply us!
> email listserv@mitvma.mit.edu with SET PICList DIGEST in the body

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body