Unexplained AFR readings..

Friend of mine has BBC motor on a dyno and after flow matching all his injectors, finds that monitoring #8 cylinder and tuning the motor based on his AFR readings, gets a good fuel map and the expected AFR readings through out the RPM range up to 5500rpm. Then he moves the O2 sensor to #7 cylinder and finds that cylinder runs very rich (10:1) at low engine speeds. As RPM increase the AFR will match #8 at about 4400rpm after that it goes very lean up to 5500 rpm, around 15:1.

Being batch fired he has moved the injectors and changed the connectors between batches and still no change. He doesn't have a scope to monitor the pulse to the injector, but he can't seem to find an explanation for this result.

Does anyone have an answer to why this may be happening?

Thanks!

Dick

air flow maybe ?

Engines are very rarely ever truly even in their air flow through the cylinders, that is why some ECUs have individual cylinder trims available when running sequential fire. Since your friend is running batch fire I would recommend monitoring AFR's after the collectors or turbo if there is one, that was you will be tuning the average AFR of all of the cylinders. You can also get this for bank-bank to balance V engines, you just need to monitor each bank after the runners have merged.

Andre done a webinar on individual cylinder trims whilst monitoring AFR on each cylinder here:

https://www.hpacademy.com/previous-webinars/064-individual-cylinder-fuel-tuning-4-channel-lambda/

But like I said, this is only valid if you are running sequential fire and have the ability to monitor each individual cylinder

Air flow could be an answer, but both cylinders are at the back of the motor, so the air flow should be similar.

When monitoring #8, from idle to wot the AFR's read what they should be. I don't have numbers off the top of my head, but he felt he had a well tuned motor until he moved the sensor to #7 and found it so rich on the bottom end and lean enough at the top end where it would hurt the piston. When someone runs a boat they can have a tendency to hold the throttle wide open for 10-30 minutes or longer and an AFR of 15:1 would certainly hurt something.

He is in the process of setting up monitoring, with new equipment, for all eight cylinders to see if this scenario still holds true. We shall see what he finds after he gets the system setup.

I know there can be variations from one cylinder to another but they shouldn't be so far out of whack that it can hurt the motor.. or can they??

Like you say, averaging may be needed but I think the end goal would be to have all the cylinders close to a proper tune.

Thanks!

Dick

I watched the webinar on individual cylinder trims and it was mentioned that errors of 5% are not uncommon, but as high as 10% can show up and is the point to look closer. Very good info, as always.

A lambda of .8000 = 11.76 AFR on the #8 cylinder at Max torque and a lambda of 1.050 = 15.435 AFR on #7 at Max torque is what I believe or is close to what my friend was seeing. If I remember the math, this would equal a 25% change. So there is something a miss between the cylinders. He bought an injector flow bench and found the best matching set to run in on the motor. It will be interesting to see what his new setup shows between all the cylinders and if this issue still exists. At least I have an idea that 10% is max and that with a boat engine under load all the time, it will be critical to understand what is happening..

Thank you for the input!!

Dick

Have you tried a compression and leak down test on the engine?

It was a rebuilt engine, just being run on the dyno to learn some things before it goes into a boat. So I doubt they are looking for a mechanical issue to explain the difference. By now he should have the 8 O2 sensor system setup and if it shows the same issues, as is seen in the video, then I will suggest he look into the health of the motor!!

I will keep you posted.

Dick

I've run into a bank fuel trim beeing 40% different from each other at idle on a 300ZX TT engine last week. I found this very uncommon, but couldn't elaborate more, as the car was having oil pump issues and we had to get if off the dyno.

The max. flow difference between the injectors at 3bar was 5%, so something must have been going on.

I would suggest that the airflow is drastically different, something like on your application.

I spoke with my friend yesterday and he has the 8 WB O2 sensor system setup and made a couple of pulls on the dyno, as well as some steady state runs to see what was happening. As I mentioned before, he is running batch fired injectors. He found out that the four cylinders on one batch have a good fuel curve and they match, relatively close. The other batch, all four cylinders are very rich at lower rpms and become quite lean above 4400rpm and at max torq. He as to remove the manifold to swap the wiring harness around, but he believes it is the driver for one batch or something in the wiring harness.

I will keep you all posted on what he finds..

Thanks for all the great input!!

Here is an update on this issue:

My friend setup his dyno headers with 8 WB O2 sensors and found an imbalance between batches with AFR's drastically changing with on batch of four cylinders. He is running MEFI3 and MEFI4 ECUs and he tried several different wiring harnesses and different ECUs and saw basically the same problem, one batch would give good AFR's through out the table and the other batch was very rich at low rpms up to about 4400rpm and then it would go lean.

I told him to look in the software for settings on firing the injectors to see if he could get the two batches closer to proper operation. He took a tune from another boat motor that had been tuned elsewhere and found a setting on firing the injectors was different. So he set his tune to that setting, which I believe was firing all the injectors each revolution of the engine. That resulted in a lean tune throughout the rpm range, around 17:1. He adjusted fuel pressure to compensate so that he could make a pull, and found all the cylinders were much closer to each other than before. He then went backed and made changes to the fuel map to bring the AFR in line with what he should have had in the first place. We surmised that because the injectors were being fired more often that the dead time had increased and the actual injector flow time had decreased causing it to run lean. But it seems all the injectors are being fired at once causing it to the resulting AFR's to be more closely balanced. After trying several different ECUS both MEFI 3 & 4, that there must be an issue with the drivers in that type of ECU. MEFI's are built by GM and supplied for crate motors and boat motors from Mercury Marine. The software to tune these ECUs seems very simplistic and hard to make changes. Fuel map can't be changed without shutting down the engine, so settings for firing options are not explained well at all and he had to make assumptions on what was actually happening when changing the firing options. As he said, I don't why it works this way, and all the ECUs did the same thing when set to batch firing, but it is working now and he is going to check other engines run by this same type ECU to see if is the norm for them.

Thanks everyone for the help and info.. As usual, learn something new every day!!

Dick