I'm suffering from something similar to kotomile (see this thread: https://www.miataturbo.net/forum/t35902/) but I have no vaccum leaks in the MAP line, which is tee'd in to the FPR line.

The fluctuating AFR is only present in the 3.000 to 4.100 rpm range when cruising (MAP values around 70), but not entirely consistent. The following two screenshots are from the same log:





'92, N/A, MSPNP, no AFM, K&N intake, LC-1, GM IAT.

I have reset the LC-1, uploaded firmware to MS and tuned the VE table using MLV. To no avail. EGO settings are equal to MSPNP base map.

What is going on? Any suggestions?

 

Between running a new (smaller) line, changing the location of (read: adding more right angles to) the MAP line, and finding a small leak in one of my T connectors, mine has subsided and the car is running much better (of course). Good luck with yours, hopefully my experience will save you some grief.

 

Damn, I'm having the same problem. It's fine until I go over 70 mph, once over my afr's start bouncing like crazy along with everything else.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

 

I do not believe it's related to the MAP line. The small MAP variations does not correlate with the large AFR variations. If you take a look on the log it's clear that the ECU stays within two VE cells next to each other. Some thing is messing up the PW. EGO correction?

 

OK, I think I have nailed it: looking through the log and checking if any parameter was correlating with the problem I got this:



Seems to be a noisy power supply issue. I will plug in a noise reduction filter ASAP and run a new log.

 

PROBLEMS CONTINUE!!!!

Installed a L/C noise filter with no change! The problem prevails and the car runs like ****.

The AFR is peaky and unstable. Looking back on some logs made in April confirms my anticipation that I am facing a problem which has evolved.

The first screenshot is from a log in April:



This one is from today:


Something is not right and I dont have a clue where to look. Help needed!

 

Try putting a fuel filter inline with the MAP line.

 

Thanks Matt. I have investigated that to day by reducing the section of the vacuum tube from the tee and it has a significant impact on reducing the flickering MAP signal. Combined with a fuel filter as a buffer volumen might do the trick.

What puzzels me is the fact that this problem has developed over a few weeks without any changes to the car. It ran almost perfect and my logs where smooth and fine. Is it possible that I have a fuel pressure problem and that he FPR distorts the MAP signal? A clogged fuel filter maybe?

 

I made some changes to my setup and put my MS vacuum line directly on the back of the IM, on the bung where the auto tranny is sourced from. I got the dreading sin waves on my map signal. Tee'd it back into the FPR line, and it was normal again.

just fwiw.

 

This mirrors my findings. When I had the TEC3r I got MAP oscillations above a certain rpm. Tried various sources for the map signal, with the amplitude of the oscillations largest when MAP was sourced from the port at the rear of the manifold and smallest when MAP was sourced by 'T'ing the FPR line (FPR acting as a damper). In the end I could only conclude that the oscillations were real. Many ECUs average the MAP reading, the TEC3R and TECgt have an option within the tuning software for this. In my case averaging over 2 data points completely eliminated the oscillation.

 

I'd post my two logs, but that would require effort.

 

Nope I was wrong, move along!

 

Wrong? Have I missed something???

 

I thought it might be a turbo thing as I have a MAP line hooked up to that nipple at the back of the intake manifold, yet my logs have nice flat lines.

Then I remembered the MegaSquirt's MAP signal comes from the front of the intake mani by the TB and the rear nipple is used for the supercharger bypass actuator.

So yeah I was being an idiot

 

okay well here.

this it tee'd into the FPR line:


this is straight from the rear hidden port used for the auto tranny:

 

The oscillation is real. I think the way the MS samples the MAP signal exacerbates the problem. It can be filtered out pneumatically for the most part. There is some info from my experiments here. The snubber mentioned in that link plus a switch to 1/8" diameter air line mostly solved the issues I was having, though the signal is still not smooth enough to run MAPdot based accel enrichment.