A zener or MOV is definitely near the top of the list for = the=20 fix.

Peak & hold drivers were not used in the EC2 because = their=20 advantage is to very accurately control the IPW at idle where = emission=20 control is most difficult, not a big consideration on the EC2. For = their=20 own reasons, Mazda didn't use them in the RX-7 ECU either. = P&H=20 drivers also add complexity and failure modes but not a big = deal. =20

The final fix will utilize the diode clamps but not as originally=20 configured. By isolating their connection to +12 (easy to do, just = one=20 trace cut) the fix can be the addition of a single part to the junction = of the=20 clamps. Might as well use the diodes since they are there and = simplify the fix. The wattage of the part will need to be higher = (than=20 individual snubbers) but my studies so far indicate that the dissipated = power is=20 surprisingly small.

Tracy

----- Original Message -----

From: Al Gietzen

To: Rotary motors in = aircraft

Sent: Monday, March 12, 2007 = 11:33=20 AM

Subject: [FlyRotary] Re: A = solution?=20 Injector flow rate mystery solved
Subject:=20 [FlyRotary] Re: A solution? was : The truth??? / Injector flow rate = mystery=20 solved

What about MOV's? = Typical=20 transient response time measured in nanoseconds...

= <Marv>

My = expert, who has=20 designed ECUs for other applications, had the following comments that = may be=20 of interest:

As far as the ECU = goes, the=20 diodes should be replaced with

tranzorbs or zeners of = about 30=20 volts. I don't think a pull down resistor

of 1K as suggested = will do=20 anything at all. It should be roughly 8 times

the injector = resistance=20 (assuming the injector hold current is less than 8

times the pull-in = current --=20 typically it's between 4 and 6). And if you

want a true peak and = hold=20 injector driver, then the National LM1949 is the

way to = go.

The resistor only = makes=20

sense at very high = RPMs,=20 probably higher than what the engine can do, so I

consider it = optional. The=20 relay contacts are not the issue. The issue is =

protecting the drive = transistor=20 which will fail at voltages lower than what

will cause an arc in = the relay=20 (not to mention RF noise). The other issue

is minimizing the = current flow=20 as the magnetic field collapses.

As previously = mentioned, the=20 =91correct=92 way to drive a high flow rate injector is using a =

peak and hold=20 circuit.

FWIW;

Al=20 G
 