Erratic lambda reading

[robespierre]

Hi,

I am trying to solve an irritating issue with the fueling of my Evo which is equipped with a M800. I spent a fair bit of time leaning it out to achieve better fuel economy during part throttle and cruising conditions...basically anytime when the car is off boost. The problem is that the lambda values seems to drift every now and then without any changes in the map. For a couple of days, the lambda sensor might be showing 1.03 for the 2500RPM/50KPA cell and then suddenly, it might go to .85 in the same cell the next day and stay that way for a few days before reverting to the original value. The drifting happens more often during partial throttle conditions but it happens during warm idling as well.

I'm at sea level and the ambient temperature right now is quite constant (+/- 2 Celsius...averaging 34 Celsius). The datalogs show the fuel pressure to be stable so the fuel pumps (Bosch 044) and injectors (Bosch EV14s) should be ok. Battery voltage is stable as well. There's neither temp/pressure compensations nor accel enrichment activated on the fuel/ign maps so the fueling should really be consistent. The lambda reading comes from the single lambda option in the M800. What do you think could be the issue?

Rgds,
JR

[Mazi]

Did you check the fuel temperature? Bosch 044 pump is an relevant fuel heater.

[yotamhks]

at first i would check what with throttle enhachments.
afterwards, check what happens to battery voltage,
if it changes - your battery compensation isn't correct.
at the very least - check for exhaust leaks,
altough if you don't report the problem at idle, it might not be it.

if you have a log it is just a matter of comparing injection time. if in the 1.03 and in the 0.85 you get the same
injection time(dont forget to compare injection timing)
then your problem is mechanical.

[IDP]

Intercooler temperature?

[robespierre]

The air temp readings are stable as long as the car is on the move. Air temp sensor is installed on the upper intercooler piping...about 8 inches away from the throttle body. Looking over some old logs, the injection timing (in degrees) appeared to be remain the same even when the AFR fluctuated but I will need to do some new logs to verify that.

A new issue came up today...my fuel pressure sensor (Ti 500PSI sensor) started going cranky. I set my base fuel pressure at 50PSI (vacuum hose disconnected) and while my HKS meter showed a steady 3.5bar (thereabout 50PSI), the Ti sensor started drifting between 40PSI to 65PSI! Anyone else using these sensors? Are they known to go defective? I have been using it for only 6 months.

Rgds,
JR

[robespierre]

Did you check the fuel temperature? Bosch 044 pump is an relevant fuel heater.

Have not taken a close look at the fuel temp yet but now that you brought it up, is there an estimate on how much each 5 Celsius increase in the fuel temp changes the AFR?

[robespierre]

There's definitely something wrong with my Ti 500PSI fuel pressure sensor but if I didn't do any compensations based on fuel pressure at all, would funky readings from the fuel pressure sensor cause erratic lambda readings?

[Mazi]

Try to use another sensor with 100PSI range. I think that from this type of sensor you can't expect less then 1% accuracy. In your case 5PSI. Also the noise figure is very bad if you are using sensor with to high range.

Once again... Check the fuel temperature. I have almost the same problems on P 997TT engine. Whenever fuel temperature climbs over ~70C engine starts to behave erratic.

[Holmz]

There's definitely something wrong with my Ti 500PSI fuel pressure sensor but if I didn't do any compensations based on fuel pressure at all, would funky readings from the fuel pressure sensor cause erratic lambda readings?

Could you tell by disable compensation based upon fuel pressure?