pulsed 10A current source

Define "fast"

Consider using a supply that's a little higher than you need for 10A to your resistive load. You can then regulate the output (turning the regulator on and off under PIC control) with a high current linear regulator. The sag for your longest pulse has to accommodate the droput voltage for your regulator at the end of that pulse.

--
What\'s the resistance of the load, and what\'s the minimum and
maximum off time?

Of course you are going to have to increase the supply voltage, assuming you're going directly into a resistive load.

One really simple way is to use a voltage regulator following the capacitor and switch the output voltage of the voltage regulator. Provided the input voltage to the capacitor does not drop below the minimum required for regulation you'll get constant voltage output, and therefore constant current into your (fixed) resistive load.

You could also replace the MOSFET with a switched current source circuit (again with increased input voltage), but that is going to be more complex, and possibly more difficult depending on your specifications for overshoot and speed.

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

About 2 Ohms and the min/max off times are 3ms/1 sec.

Thanks!

Fast enough to slam the current up to 10A and back down to zero again in under 25us. Rise and fall times are not critical beyond that loose definition.

Thanks!

This is a trivial project if you have a 10A regulated supply, which is also a trivial project, or a $75 ebay purchase. Inexpensive power MOSFETs have no trouble switching 10A in less than 25us, even with wimpy gate drive. We're used to designs where we switch 25A in 10ns, which is a bit more of a challenge. :-) Perhaps you can tell us a bit more detail about your project and we can make some good suggestions.

BTW, if you want a linear mode 10A MOSFET current source, that's possible, but it's a bit more complicated issue.

--
 Thanks,
    - Win

I realize it's trivial if I want to use a 10A supply. My thinking was that I'm only averaging a few watts into the load and it's sort of over-kill to use a 300W power supply (not to mention the size and weight of a 300W regulated supply). I'm on a student budget so using a low current supply such as a 2A 24V switcher I already have is the preferred route.

My simulations indicate that there's no problem with the mosfet switching speed (though I have no experience with the accuracy of the model in this regard). I will breadboard the thing and make some measurements when the mosfets arrive, hopefully during spring break. I am going to use Si4420DY mosfet(s). Probably overkill for this app, but I will be using them for other projects down the road...

If there's no easy way to do it I can live with it. I'll add a bunch more capacitance and maybe use multiple, parallel outputs from the PIC, selected as a function of the on-time or off-time, with output each driving a mosfet with different current limiting resistors.

Thanks!

Yeah- it's called using junior high school arithemtic and 10th grade physics to size the capacitor. Then 5% of 20V=1V and this is 1V*C coulombs lost to 10A * 200us= 2mC at longest pulse width, making C=2mF=

2000uF. So a 2000 microfarad at 25V WVDC or more should do it, low ESR, and ripple current rating of approximately 2x power supply max DC current capacity.

Thank you.