ECM case grounding

[daleulan]

On that ECM the internal signal ground is connected to the case. So you should not ground the case otherwise you will get ground loops and sometimes that can cause the ECM to burn up - if the engine-to-chassis ground strap is not perfect that can happen when you hit crank. Some applications had an inline 0.1uF capacitor connecting the case to the chassis of the vehicle. I thought this was funny in a production vehicle but that's how it was done, probably to avoid ESD/EMI sensitivity. That ECM is on a 2-layer PC board which is not very good from an EMI point of view, it may crash if you climb into the vehicle with wool clothes or that sort of thing. You can try to add the capacitor if you get that, it may help. All of the newer 4+ layer controllers (1227730 for example, although you can't use that one for your application) are far superior in this regard. Visteon (Ford) grounded their case differently and in that case the ECM should be grounded. It depends on the original installation.

[Mastiff]

If I remember correctly, the 7747 ECM aluminum case is at ground potential via pin D7, which is system ground. Follow this ground and it's tied to chassis ground, or battery negative.

Yeah, I guess the idea is to not have non-ECM stuff using the ECM as a ground source. If the case is grounded, some fraction of starting current (for example) would go back through the ECM wiring, through the case to ground. If the case is ungrounded, then this is not possible. Right now I have some relays attached to my case and they are using the case as ground. Seems fairly innocent, but I should probably move those grounds someplace else so the ECM doesn't need to share.