AFR values way off after dynoday
 

Setup: M45 Supercharger, MSPNP 90-93, Inovative LC-1, 550cc injectors (measured to 700cc at 2.4 BAR), IC, IAT, etc.

I've been running this setup for many years. I've just tuned idle, WUE and the VE tables. It was ok to drive but I knew it was a "crude" tune.
So I took the car to a dyno and let the tuner have a go. He had problem with the AFR in Tunerstudio being very bouncy. He struggled to get a stable reading. So I took out the wideband sensor and did a heat calibration and then we cleaned it with compressed air and did a free air calibration. I also changed the spark plugs (NGK xx 7 something).

This helped a little. So he re-scaled my tables and started tuning. He tuned the spark table etc and I got 155 HP and 183 TQ.

But still he struggled with the AFR readings and while cruising the car would not run 100% (shaking / stumbling). He then changed the number of injectors in Tunerstudio from 4 to 2 and he was able to tune everything better. I drove the car home (1 hour trip) and everything seemed perfect! When I got home I idled the car and the AFR was now sitting around 11...
The next day I took the car to work and now the AFR was showing 11-12 while cruising in fifth, not 14-15 as it should.

Any suggestions?

The Wideband is earthed on the engine (same place as the ECU according to the installation guide I followed back in the day: https://mx5unleashed.com/tech/megasquirt-pnp.html )


The runs fine but I can smell all the fuel. ​​​​
I'll post the tune / logs later.

Sounds like the Sensor is failing. Tuning off a faulty sensor sounds like a recipe for disaster if you ask me. If it were me, I would be replacing the sensor so you know you have accurate readings and go from there. Unfortunately, you will likely need to re-tune depending on what you find with the new sensor.

I changed the sensor in 2016 and I don't use the car that much. But maybe high AFR damaged it? I'll try to get a replacement or install the old one to see if there is any difference.

The datalogs are corrupted. Got saved as .mlg. Tried to eksport them from megalogviewer but I can't get any data from them.
I've had a similar problem before. But then it was the other way. I would be out driving for an hour or two and then the AFR would get LEAN at cruise, but it was fine at WOT.

The tuner also disabled EGO correction (if that is the correct term).
I'll log again (changed settings in tunerstudio) and upload the tune and logs.

I changed the O2 settings in tunerstudio (new computer) from Innovate LC-1 0-4 V = 0.5 -1.5 Lambda to Innovate PLX 0-5v 10-20:1 AFR. So now the WB-gauge in the car and in tunerstudio are in sync.
It just seems that my whole fuel table is rich. Maybe there was some sort barometric pressure difference from where I tuned the car to where I live (on the coast) ?

Anyway, the tuner will take a look at my logs and change the VE-values.

I've attached some small logs and my tune for anybody willing to take a look at it.

I'll be ordering a VTPS kit and maybe a new LSU 4.2 (o2 sensor).

You may also have an issue elsewhere in the fuel system: partially clogged filter, failing fuel pump, etc.

I was watching a good video about engine dyno mistake on High Perfoamnce Academy YT channel lately, it was interesting watch (the whole channel is btw), .
They also say wide band sensors seem to have an On/Off failure mode which would make your suspicions about failing sensor most likely not your problem.

Thanks for the input. I'll check out that video.

Upgraded fuel pump and new fuel filter 4 years ago.
The problem here is a rich AFR after tuning it to be "perfect". I'll change it to be leaner, and report back if it changes again.