Noisy data with engin on.

[PumbaScout]

I am trying to set up an ECU for a 95 LT1 Caprice engine that will be used in a 58 Chevy conversion. Currently I have the body off and I am working with motor in frame and ECU is on a table next to the car. I have been able to download and flash data on the bench and in the car no problems. When I try to use the gauge function when the engine is running I am seeing all the outputs go crazy as soon as I start the car. If I use a pot on the temp guages I get stable readings with correct scaling on the bench. I am not sure if I should be looking at issues with the homeade cable, baud rate, com settings, grounding issues with cable? Does the ECU need to be grounded to the engine? Or maybe the adx file that I started with is not right. I am just looking for a direction to start with. Any help is appreciated.

[MO LS Noobie]

I would confirm Cable integrity first, but I suspect a grounding issue. Double check your grounds from the battery negative terminal to engine, body and chassis. Next I would run it for just a couple of minutes with the alternator belt off. I have seen AC voltage from the alternator cause some weird problems, and also static electricity from some belts will interfere with CKP and CMP signals. I would also check the routing of the coil primary/injector wires to make sure they are next to any secondary voltage. Good luck with it.

[PumbaScout]

Thanks, currently no wires going to the alternator. The main coil wire is connected through a switch direcly to B+ but the original one is still sitting in the wire harness open. Do you know if the ECU should have its case grounded to the engine. Currently it is sitting on a table separated from the car and electrially isolated.58 first start wire harness.jpg

[MO LS Noobie]

The ECM case itself should not have to be grounded, it is grounded internally through the wiring. That being said you can never have too many ground wires tying the electrical system together. Double check all PCM grounds and make sure all power wires have steady battery voltage. With everything looking okay without the engine running I would suspect EMF from the coil/secondary wires inducing a voltage into the harness. Try separating everything in the secondary ignition by about 2 inches from all primary wiring. It looks like a fun project.

[kur4o]

tunerpro is known to have troubles controlling bus master.

To rule that out use eehack, and do some debugging if not working correctly.

[PumbaScout]

So after working on the car yesterday we managed to fix some issues we were having with engine. Optispark was the culprit for rough running idle. Replaced it and the engine runs great now. The issue with the ECU gauges in TunerPro is still an issue. I was able to learn that if I start the car with the link running then everything goes crazy. If I stop communication and restart communications while engine is running then most of the time it stabilizes and runs ok. It feels like communication protocols is getting out of sync sometimes. I am using a homeade cable and I am starting to think it may be the culprit. I am putting my wiring diagram here to see if maybe I am not wiring it correctly or my choice of FTDI chip type is bad. ECU cable.png

[PumbaScout]

I use eehack to upload new bin files but I am unaware you can monitor outputs with it. To be clear uploading and downloading bins works just fine with my setup. Only gauges have a problem in TunerPro