Using Logic ICs to make short pulses

[Jdo300]

Hello All,

I'm designing a driver board to control some MOSFETs. I have some mic4427 MOSFET drivers to drive the gates into avalanche mode but the function generators I am using to trigger the chips only put out 50% duty cycle square waves. I'm working in a wide frequency range from 1 kHz to about 500 kHz and I simply want to use the shortest possible pulses to turn on and off the MOSFET driver chips.

I am wondering if there is a way I could use some logic chips to produce extremely short pulses ~1uS to trigger the MOSFET drivers? I would simply feed my function generator output into the chip(s) and they would, in turn, make the shortened pulses to drive mic4427s. If not, does anyone know of a simple way I could accomplish this? I'm trying to think of the most bare bones way to do this with the least components.

Thanks,
Jason O

 

[berkeman]

Can you just capacitively couple the square wave, in order to get pulses on the edges? You can use the positive pulse to turn on the FET. You may need to diode clamp the negative-going pulse, if it exceeds any reverse breakdown voltage specs of what you are driving.

Put an explicit discharge R after the C, and tune the RC time constant for your pulse width.

 

[chroot]

It's a simple solution, berkeman, but the "pulses" are not going to be anywhere near a microsecond long -- ideally, they'd be deltas if the function generator has a nice fast slew rate -- and they're going to be rather poorly shaped.

If it were up to me, I'd use a monostable multivibrator (a "one shot"). A tiny IC won't take up much more board space than a large-valued capacitor, and it'll be a lot more configurable and precise.

- Warren

 

[Jdo300]

Hi Warren & Berkeman,

I really like both of your approaches and would like to test out each one to see what works best.

@berkeman,

I would not be driving the MOSFET gate directly from the function generator + short pulse setup, the MOSFET driver chip will drive it directly. I just need a signal to the FET driver that is fast enough and with a short enough width to do it. According to the datasheet for my FET drivers, the input needs 1.5V to V+ to turn on and is monitored by a comparator which will smack the gate of the MOSFET on in under 43ns once the threshold is reached. In my case, V+ is a 9V battery. Could you tell me how to wire the RC setup to test out your idea?

@Warren,

I really like the multivibrator idea because it is closest to what I was imagining before. But what kind of IC specifically should I use?

Thanks,
Jason O

 

[chroot]

Jason,

The LM555 or equivalent is the simplest and probably best choice. You can look up its monostable configuration in its datasheet.

- Warren

 

[xez]

I agree with chroot; use a monostable multivibrator chip.
1us is hardly 'extremely short' relative to the speeds of
basically *any* common digital logic chip you can
easily buy today will generate pulses of 100ns to 5ns and
correspondingly operate at frequencies from 10MHz to
200 MHz, and it's hardly uncommon to find 500MHz to
2GHz rated parts if one needs them.

Even the decades old 555 timer/multivibrator will generate
about 1uS pulses in astable mode, though it isn't the
best choice due to retriggering issues and that being
up there in the range of its speed capability.

Anyway, try this multivibrator (or a similar one with
schmitt trigger inputs and edge triggering)..

Use the TTL 74LS prefix equivalent part version
(check the speed first) if you'd rather not work with the
relatively more electrostatic discharge damage prone
CMOS parts.

http://focus.ti.com/docs/prod/folders/print/cd74hc123.html
http://www.ti.com/lit/gpn/cd74hc123
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=470996&Row=841592&Site=US

330pF and 5.6k will be just about a 1uS pulse time
for that part.

Beware -- check the output drive current specifications
of the MOSFET drivers and the gate capacitance of the
MOSFETS; it's entirely possible that large higher power
MOSFETS will be difficult to turn on and off in 1uS
simply because the gate capacitance can become
rather large and that (as well as other inductance and
slew rate limitation issues) limits the maximum frequency
one can toggle them on and off at given a reasonable
amount of gate drive current.

For low frequency repetitions of short pulse
periods, however, at least thermal dissipation issues
due to gate drive power shouldn't be a limitation since
you're letting them stay off for 500 times the desired
pulse period.

Also beware that the logic levels and drive requirements
of the MOSFET gate drivers are compatible with the
logic levels of your multivibrator chip.

 

[xez]

PS -- oops I thought you had said 500Hz to 1kHz rep.
rate.

Actually at up to 500kHz pulse rep. rate you better
make sure that your driver and MOSFET gate current
and power dissipation / thermal specifications aren't
being exceeded by driving the MOSFETS on and off at
such a frequency. It's not unthinkable that the larger
capacitance MOSFETS may have issues pulsing at quite
that speed for extended periods of time.

 

[Jdo300]

Hi Xez,

Thanks for all the great info. The MOSFETs I am using are IRF840s and you're right, they have a lot of gate capacitance; that is why I need to use the FET drivers so I can get them to turn on fast at high frequencies. I can directly drive them up to 1MHz off of my function generator but I have a Coax line between the MOSFET drivers and the board containing the fets and there starts to be a lot of reflections and other garbage on the line once I get up around 600 kHz and up. But for the range I'm using they are switching on very cleanly. I'll check out the links you posted, thanks for your help :-).