S300 on B16A Lean and idle issues when warm.

[racin_jayson]

Setup:
I am running a P28 Ecu with S300, Car is a 1991 Civic, my engine is a B16A with an OBD0 - OBD1 Conversion, it is turbocharged, I am using all factory sensors except for a honda OBD1, 4 wire O2 sensor and a GM 3bar map sensor. The engine is fully built, now at 9:1 compression, and was dyno tuned in 2007 to 17psi boost and it has run problem free to this very day with no issues whatsoever. It has a healthy ignition, fuel and air intake system. (I have obviously checked the basics)

Issue:
I recently bought a PLX wideband to monitor my AFR's and went through hondata's steps for installation with no issues. When I connected my laptop to my ecu it required a firmware update, (it has been years since I connected my laptop) I went through the update process, reloaded my calibration and adjusted the settings to read my new wideband. The car started fine, AFR's were normal, but as soon as the car gets close to operating temperature it suddenly drops the idle and leans way out. Thinking that I may have done something wrong with the wideband install I reverted back to the original settings, and re-installed the original 4 wire narrowband and am having the exact same symptoms.
I took a datalog of the issue, you can see at several points I am holding the throttle to keep the engine from dying but towards the end of the datalog (about 8:45) you can see how the engine runs on it's own just before dying.
If I attempt to drive the car in this condition I can apply part throttle but the engine has no power, if I give any more than part throttle it bogs and will not rev up, if I let off it dies and stalls. It is basically not driveable at all.
One thing I noticed in the datalog, the ECT correction is at ~-40% when the symptoms arise, not sure what this means though. The datalog was taken with the narrowband sensor in place, but IIRC with the wideband the AFR's when the engine was dying were around 20:1

I don't understand what happened, this car has been problem free for 5 years, and ran like a top right up until the moment I parked it to install the wideband. I have checked and re-check every single wire from the ecu, gone over every single component on the engine and searched through my calibration for a parameter that I may have accidentally screwed up, I have searched every forum online and have found nobody with the same issue so I have turned here. Hopefully someone can help me out.
Thanks in advance.

[Spunkster]

You appear to have a problem with battery voltage. It is lower than it should be and then continues to drop. It should be about 13.5-14.2 volts. Also there may be a problem with your thermostat as the car is not warming up as it should be.

[racin_jayson]

I must have a weak alternator as I just installed a new battery, I will change my alternator and see.
A new thermostat wouldn't hurt as it has been 5 years since I built the engine. I will try these and see where it gets me.

any thoughts about the temperature compensation pulling too much fuel out? It one point in the datalog I can see the short term adding 40% and the Coolant temp compensation pulling 40% out. then suddenly the short term goes back to zero... could low voltage be causing these issues?

Thanks for the reply, at least I have some new things to look into.

[Spunkster]

Fix the other 2 issues first then see if you still have a problem.

[racin_jayson]

Replaced the thermostat with a new OEM unit and replaced the alternator, voltage is better now.
As before, when car is warming up as soon as my ECT climbs to around the 150 degree range (plus or minus a few degrees) While datalogging I can watch the ECT.c go from zero to -39% and the car idles extremely low until it dies, when cracking the throttle the ECT.c goes to zero briefly and the engine revs up then goes back to -39% and wants to die. I am 100% certain that the ECT.c is causing my issues, but I do not know why it is doing this? Is there a parameter I can adjust to stop this?

I have similar symptoms to several people in this thread:
http://www.hondata.com/forum/viewtopic.php?f=19&t=9796

Why after a firmware update would this problem arise?

I really need some help here...

[Spunkster]

Post a new datalog now that the voltage is correct

[racin_jayson]

Problem solved! The temperature range on my water temperature compensation parameters only went up to 65 degrees C.
I'm not sure why this suddenly happened but I restored air and water temp compensation defaults and the problem is gone.