Having fun at 350 amps

[Winfield Hill]

boB wrote...

On 1 Feb 2020, Winfield Hill wrote:

** Our HUF75652 MOSFET can handle 33kW for 20us.

How about a lower RdsOn FET ? A 7 lead part, maybe
the Infineon IPT015N10N5 100V 1.5 m Ohm I think
it is also a bit larger die.

RdsOn isn't so important for linear applications.
That's an impressive part, but our TO-247 with
lead frame has more thermal mass. It can handle
about 3x more heat in 20us, which is important.
The TO-247 is also easier to get heat into a 100W
heatsink, needed for long pulses or high rep rates.
Some other TO-247 or SOT-227 parts might be better.
My linear circuit would also be happy with multiple
MOSFETs and mR source resistors, drains in parallel.

For high current shunts, 500A @ 50 mV is typically
used or 1000A @ 100mA shunts.

Yep, I have a nice stock of big parts like that.
Our 5 to 20mR values are chosen with high voltage
drops in mind, needed by the circuit. (E.g., 4V
compares to Vgs=4V, 1V for opamp control.) Small
size, low inductance is necessary for fast di/dt.
The Ohmite FCSL110 wide 5W sense resistors also
come down to 1mR, and in an FCSL150 10W version.
They're better suited for small circuit boards.


--
Thanks,
- Win

 

[boB]

On 2 Feb 2020 09:33:32 -0800, Winfield Hill <winfieldhill@yahoo.com>
wrote:

boB wrote...

On 1 Feb 2020, Winfield Hill wrote:

** Our HUF75652 MOSFET can handle 33kW for 20us.

How about a lower RdsOn FET ? A 7 lead part, maybe
the Infineon IPT015N10N5 100V 1.5 m Ohm I think
it is also a bit larger die.

RdsOn isn't so important for linear applications.
That's an impressive part, but our TO-247 with
lead frame has more thermal mass. It can handle
about 3x more heat in 20us, which is important.
The TO-247 is also easier to get heat into a 100W
heatsink, needed for long pulses or high rep rates.
Some other TO-247 or SOT-227 parts might be better.
My linear circuit would also be happy with multiple
MOSFETs and mR source resistors, drains in parallel.

For high current shunts, 500A @ 50 mV is typically
used or 1000A @ 100mA shunts.

Yep, I have a nice stock of big parts like that.
Our 5 to 20mR values are chosen with high voltage
drops in mind, needed by the circuit. (E.g., 4V
compares to Vgs=4V, 1V for opamp control.) Small
size, low inductance is necessary for fast di/dt.
The Ohmite FCSL110 wide 5W sense resistors also
come down to 1mR, and in an FCSL150 10W version.
They're better suited for small circuit boards.

Didn't think you were using the FET as a current source. You just
need the largest die and best heat transfer possible. TO-247s do not
necessarily guarranty that the die is large but your part is pretty
good.

For short-ish pulses, you can also use inductance to limit current. We
used to do that to test power FET current handling capability.

Sounds like you're on top of things.