Information Overload - Troubleshooting random CLT noise

[Clonebug]

You should be able to unplug the CLT and it will default to a certain temp......maybe give that a try and even if it runs on cold start you can at least get a log of it for a few minutes.

[luftvagon]

Whats the default temperature? I'd probably have to warm it up, disconnect, and disable WUE.

[Clonebug]

I'm not sure what the default is...I'd just fire it up with a log running and check to see if it clears up with the CLT unplugged.

It won't matter if it runs a little rich for a while.

[Jadewombat]

What he said. You can just hookup a 50 or 100ohm resistor in place of the CLT to see if the problem repeats. I pulled the ballpark 50 or 100 from this graph:

[Piledriver]

Turn wue off temporarily (flatline it) and unplug the sensor.
If you are using a GM sensor for iat, flatline IT and plug wue into that.

I use the same $15 open element sensor for both, so simply swapping is trivial.
I lost the IAT sensor once, died just like you are seeing with the clt... led me to alter my IAT and WUE tables so full cold (open) gave me only 100% so it would run disconnected.

I have lived in a place where -40F was a sane reading, but I assure you, -40F is NEVER a sane reading.

[raygreenwood]

I'm 99.99% confident wiring is good at this point. I split the harness the other day. All my sensors wires terminate to the common ground that goes to pin #20 on microsquirt. I clipped the TPS, and now soldering back on. There was no change in behavior. I'm still having random dips on the CLT to the point where it will kill the engine....

Back to my original thought... if the temperature is "dropping" that means the resistance is increasing somewhere. Which could probably be the thermo probe going out, a bad cable, or something inside of the microsquirt.

The type of noise you are seeing......is also a COMMON voltage potential issue....caused by soldering of connections in a flexible wiring harness.....which is a no no. Soldered connections are far, far, far, far inferior to properly crimped, hermetically sealed connectors.

Solder has micro voids...always. These voids expand and contract with heat and cause potential problems as heat rises and falls.
This is why in millivolt sensitive fuel injection systems....and in aircraft wiring harness systems....there is not ONE production system and has not been in over 30 years....that uses solder in an unsupported wiring harness.....which means soft wire to component.

Solder however is used between wire and rigid PCB board. This is because one end....the board....is fixed and the wire can be strain relieved and held down by hot or cold elastomer. It is used rarely in automotjjve compared to land based machines....because of the very high vibration and heat differential in cars.

Stop soldering ANY fuel injection wiring terminals or connections. It can create mysterious issues just like you are having now. Crimping is a very highly developed industrial science. Use it.
Ray

[luftvagon]

Ray - good point. Coincidentally this is the only wire I soldered, and problem has started before I did this. Everything else was crimped using non-cheap uninsulated crimps, and heat shrunk.

So I have not had any issues lately after removing grounding of the shielding wire on the VR sensor cable . Took it to work today, 60 mile round trip, and a lunch errand, and it did excellent. Not one hiccup, reset or sense that the engine did something unplanned. May have even logged it - will check for noise shortly. Left it parked in the driveway for ~3hours, and took a ride around neighborhood with the kids, and it dipped again, and board reset a minute later while coasting in neutral... A 2014 Focus ST I had used to do that.. bad solder joint on MAP sensor.. there was a tech bulletin for it.. whooda thought.. You'd push the clutch in, and the car would stall.. Unfortunately no data log.

The microsquirt power comes from the alternator, switched by ignition. This has not bothered it before. This system run fine for years. It's a long run from the battery to the squirt, hence why I never run a cable back. I measured the resistance drop from negative terminal to microsquirt ground, and it was 0 ohms on my meter.

There is some voltage drop at microsquirt, but only with IGN-4 being powered on. The main power relay feeds the VR sensor, 4 high-z injectors, IGN-4, diyautotune ignition module, and microsquirt.

I guess there is still a chance that this issue is being induced by rapid heating of the engine, but not on the TPS wire. Oh; just remembered.. I soldered the male end of the CLT sensor, at the plug. The male end was part of the actual sensor before the sensor got desoldered.

[luftvagon]

To @Piledriver 's point, I'll never drive this thing at -14F and below. Hell, 32F is probably the coldest it will ever seen. It's a good fail safe.. so -14F is set to 100 WUE at this point. That should help with flood hiccups.. I've inspected the harness yet again. Everything looks good.

Everything microsquirt related has been grounded to sensor ground; with one exception:
QuadSpark Ignition Drivers from DIYAUTOTUNE https://www.diyautotune.com/support/tec ... egasquirt/
According to their documentation they should be grounded to the engine block -- which I think should be ok. Two drivers are grounded to the engine block.

The hunt continues

[luftvagon]

Well, this image speaks for it self...

[luftvagon]

Now the big question is.. is it the alternator or the ignition.

[luftvagon]

...and by the latest log it looks like it's ignition. Going to make another capture with the belt off/alternator unplugged.

[luftvagon]

[luftvagon]

I'm using spark plugs with built in Resistors.

[luftvagon]

I think I found the source of the noise. After rerouting IGN1 and IGN2 signal wires and re-wiring them with 2 pair shielded wire, I no longer have CLT noise.
MAT/IAT noise still persist.. which may not be noise at all, but just temperature variation. It varries by few degrees. Never by more than 5 degrees. I used 22 gauge shielded wire for IGN signal wires; all I had on hand.. which I dont think should be a problem since they go into transistors, and then engine ground via QuadSpark. Too low gauge wire? So far no issues (>60 miles). It's very easy to change, because its outside of main harness.

Now onto more fun stuff... Going to go semi-sequential for fuel. My injectors are already wired for 1,3 bank1 and 2,4 bank2.