Page 1 of 1

Erratic MAP Readings

Posted: Thu Feb 20, 2020 11:51 am
by lcalex
I have an '02 RSX with a K20A3 engine. I am using a PRA calibration because that's the only type that does not throw a CEL on my car. I am randomly getting erratic MAP readings, which I think is causing the short term fuel trim to go negative sharply, all the way to -27%. It gets so bad that the car will sometimes stall when I get to a stop with it happening. At first I thought is was the MAP sensor and I replaced it with a new OEM one, but the problem didn't go away. The MAP readings will return to normal when I restart the engine, then the readings will go all over the place again after the engine's running for a while. Any kind of input/help will be appreciated.

Re: Erratic MAP Readings

Posted: Thu Feb 20, 2020 12:12 pm
by Spunkster
What is the part number on your ECU? What error code were you getting with non-PRA calibrations?

Re: Erratic MAP Readings

Posted: Thu Feb 20, 2020 5:44 pm
by lcalex
I will get P1607 if I try to upload any non-PRA calibrations, so I loaded the European CTR factory calibration and imported the basemaps from a K20A3 map and started to tune from there.

Re: Erratic MAP Readings

Posted: Thu Feb 20, 2020 10:56 pm
by lcalex
The code of my ECU is 37820-PND-013. The car is an '02 Acura RSX, RHD, it used to be an automatic but it's converted to manual. I am not sure whether that is the original ECU that came with the car, or it's replaced with a manual ECU by whoever that did the manual swap. I am from Hong Kong, so the car should be close to EU/UK spec.

Re: Erratic MAP Readings

Posted: Fri Feb 21, 2020 10:01 am
by Spunkster
The JDM PND-013 is not supported and will generate errors and will not have full functionality if used. There are no calibrations specific to this ECU.

Re: Erratic MAP Readings

Posted: Sun Feb 23, 2020 8:15 pm
by lcalex
I think the problem is on the o2 sensor, I don't think the o2 sensor on my car is a wideband, and maybe its different from the narrowband on the European CTR. I hooked up an AEM wideband and the problem seems to be gone.