When I set mine up, they were totally independent. Separate Jesse Prather sourced crank sensor, MAP, IAT, Water temp, lambda, knock sensor, power. The only thing I shared was throttle position, which isn’t used for anything except transients. With everything separate, I don’t have to be very smart. This was very early in the concept of parallel; no sharing of sensor inputs. I used all GM sensors; available at every NAPA.

So OEM ECU works fine. My next step would be to activate the parallel ECU and let it do nothing. What happens? You can check every sensor and data log without the parallel ECU having any control. Does this work?

 

Either way now that I have the harness back out of the car, I will cut the 5v ref to the stock ECU and retain only the ms2 vref.

I'm confused as to what you're saying here: "but as you should share the signal ground and actual ground I'd only use one or the other, probably the stock ECU one". Did you mean *shouldn't* by chance? If I don't share the sgnd and main ground with the stock ECU, I would need to find another grounding point on the car correct? If the ms2 inherently does not separate signal and power ground internally, would my sharing of the signal ground with the stock ecu signal ground induce a significant amount of noise into the stock ECU's signal circuits?

Unfortunately I don't have a scope, $200 bucks for a half decent scope isn't really worth it for me considering I'd probably use it once a year at most. I did find a neat handheld scope on amazon that's based off a DIY kit: It only has one channel and i'm sure the resolution isn't great but could be pretty handy tool for a home garage tech like myself!

Is there a standard R56 and R52 resistance that nb owners tend to use? I've played around with both pots and haven't found really any change in the signal regardless of their settings.

I vaguely remember reading about "diode isolating" the ckp/cmp signals from the stock ECU. Again, not an electrical genius, but could the conditioning circuits in the ms2 be creating some kind of noise in the cmp/ckp signals that are being fed back to the stock ECU. I'm guessing that is the purpose of the diode-isolation...?

Unfortunately it's not so easy to add and remove connections for testing purposes. I created a boomslang harness using the stock ECU board mount connector and the 3 individual male connectors replicating the stock ECU harness connection scheme. So essentially everything that's shared with the stock ECU is a straight pass through and the remaining connections are ran out to the DB37 connector for the ms2. I've run out of the crimp connectors that are needed to install additional wires into the harness, though I could purchase more I guess.

 

If you had it running off the MS2 I assume your potentiometer voltages were halfway correct. I forget the exact values, but google should find them.

I have one of these, used it to find trigger problems on my MS3x. They used to be cheaper, but what isn't any more. There are custom firmwares for these too. I also have a faster 4 channel rigol, still use the pocket one more often.
https://www.ebay.com/c/1261520748

It's probable the MS2 is messing with the signals from the triggers. It might be strong enough to still trigger the megasquirt but too weak to do the same on the stock ECU. With a scope you could confirm pretty quickly by unplugging the MS2 and seeing if things change while cranking. If you get too much impedance or resistance on the signal from the crank sensor line, as you might trying to run two ECUs off one sensor, it won't be able to reach a full 5V before switching off again. Might be you could try lowering the pullup resistor value, look up formulas for resistors in parallel, so that the open collector on the sensor can charge up to 5V faster with more current available. A scope would be amazingly helpful. See what the stock ECU sees loaded on a running car then you know what you are after. If you reinstall the MS2 and the signal goes lower then you found the problem.

Diodes cause voltage drop, and I think the cam sensor is pretty weak as is. A digital buffer would probably be better, a very high impedance input that should not interfere with the source signal at all but will drive an output with enough current to trigger something else. Should work if I'm not mistaken.