[Diy_efi] current sense MOSFETs

[Phil Hunter]

I did a few searches of the archives, didn't find that this has been
discussed before, but I keep thinking this could be a useful addition
to the DIY'ers bag-o-tricks so I'll do a short writeup.

The basic idea is you can get some MOSFETs that have a small, accurate
percentage of the total current "bled off" to a fourth lead, from which
you can determine how much current is flowing between the source and
drain leads. The design "win" is the reduction in power to measure
large currents, and probably less complexity also.

Imagine a transistor made up of 100 identical MOSFETs on a chip. All
100 drains are tied together to one lead. Ditto 100 gates. 99 sources
are tied together w/ the third lead, and 1 source goes to the fourth.
In this case 99% of the current flowing thru the drain lead will flow
thru the main source lead, and 1% will flow thru the "sense" lead. 
If you were driving a fuel injector, and measured 0.01A in the sense
lead, you'd know 1.00A was flowing thru the injector coil (connected
to the drain). 

Looking at the heater control section of the DIY-WB, measuring the
current thru the sensor's heater is done w/ a 1 Ohm resistor having
about 1 Amp flowing thru it, thus a 1 Volt drop across it. If this
resistor could be eliminated, the DIY-WB could work on 1 less volt
of power, less still if you eliminate the voltage regulator also.
In theory, a current sense MOSFET could be a way to do it.

A quick look thru a Digikey catalog, I found 5 by International
Rectifier (HEXFETs), picking the first one, here's a URL:
http://www.irf.com/product-info/datasheets/data/irc530.pdf

Note there are 5 leads rather than 4, the 5th is a "Kelvin" reference.
Also note the current ratio (r) is ~1400:1 rather than 100:1 in the
imaginary example above. Max current is 14A, which would be 0.01A
max. in the sense lead. A look at Fig. 18 reveals how the Kelvin
lead is used, note the negative power supply for the op-amp in this
design example.

Disclaimer: I have not actually designed/built/tested/trouble-shot
with any of these parts...yet. It looks about as straight forward
as any other FET designs, but there may be "dragons" I don't know
about. They've been around a good many years, yet do not seem to be
very popular, there may be reasons for that. The "wasted" 1 Volt
in the DIY-WB is not an issue for me, so I'll leave any improvements
as "an exercise for the reader" :) For ignition coils, my intuition
is IGBTs are probably still the better way to go.

Just some "food for thought"...

regards,
philh
(digest)

_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi