Peak-n-hold failure!

[rr]

Mike, there are only two things that would cause a failure
of this type: excessive heat, or inductive kickback spike.

The PnH driver has to dissapate more heat than a saturated
driver, because of the hold current/voltage. Your heat sink
may not be large enough: 12.6W vs. ~1.2W for sat inject.

As for the zener, the CS453 datasheet shows a 42V breakdown
for the driver output transistor. You should use a 30 to 33
volt zener @ 10W (10 W => that from data sheet).

I like using the TVS type of zeners for this kind of
application. They are designed for inductive spike killing.
Also, you mentioned that the ECM uses one zener for a bunch
of drivers. This may place the zener too far from any one
device.

These zeners really need to be right at the leads of the
component that they are trying to protect.

HTH

BobR.

>Now for the let down.  I went to go test drive the modified 
>ECM and as soon as I revved it once in the driveway, billows 
>of raw fuel smoke started emanating from the exhaust.
>
>A post mortem shows that two drivers failed with their outputs 
>going nearly to ground (147 ohms on one, 13.4k ohms on the 
>other).  All other drivers are > 3 megohms.  These two drivers 
>were holding their respective injectors open 100%.
>
>Soooo, I wonder if it's because of the way the ECM shares 
>one zener between all drivers, isolated by diodes facing the 
>opposite way?  Is there more flyback current when using peak 
>and holds?  I wouldn't think so.  The input signal is a nice 5v 
>square wave just like it's supposed to be, so I can't imagine 
>that did the damage.
>
>I guess the next step is replace the two failed drivers and 
>modify the circuit with separate zeners on each driver.
>
>At least the only smoke this time was all out of the exhaust. 8-)
>
>-Mike