RealMcCoy
01-21-2023, 11:50 PM
This one ran me around in circles for a minute... (more like hours) Figured I'd share for the next guy that goes through it...
The guy that owns the car is an experienced tech at another shop, and is no dummy, but wasn't as familiar with Chrome as I am. (turns out I wasn't as familiar as I thought I was...) He had wired in and installed a 99 ECU, and had me flash Chrome into it. The car previously had run fine with the 1G ECU, but was setting a knock sensor code. He replaced the sensor at the same time he replaced the ECU.
Symptoms presented immediately. it would start and run fine until you took off and got the rpms up. It acted like it hit anti lag or some type of limiter, and the CEL would pop on and stay on until you turned of the key and reset it, as well as just run like crap... We were convinced we bolted in a new problem with either the ECU, the wiring, or the tune...
I'll spare you the details of wild goose chases and snipe hunts embarked on, but here's what I found: Datalogs told me that once the issue presented, and the CEL came on, fuel was commanded at 11:1 AFR or less, timing was commanded in the single digits, and knock was fixed at 7-9. This showed all the signs of a rather brutal failure management strategy for a knock sensor failure. But why didn't it have any codes? and why was the CEL on? The CEL was easy... Chrome CEL on knock was enabled over 5 counts, and we were fixed at 7-9. Why no codes was a little harder, and I'm pretty sure there's a Mitsu engineer still laughing his ass off at pulling that prank... If you look at any 96 up Mitsubishi, P0325 is a knock sensor code. If you specifically look at the 98-99 turbo cars, guess what code they removed from the displayed codes menu..? Of course they didn't remove the sensor test, or the failure management strategy if it failed, there's just no code to display even though its set and being compensated for.
If you've ever experienced the failure management for a knock sensor code on a 1G, you know it's pretty mild. Pulls a little timing, fattens it up a little and it just doesn't run as strong... Don't know why they made the 98-99 so severe, but it's pretty brutal. Nothing like the 1G. It's nearly undriveable once engaged.
So on this particular car, he has a short in his knock sensor harness somewhere, the sensor wasn't the issue, and the 99 ECU with Chrome is much less tolerant, making us think we had bolted in a new problem when the whole time it was just the new ECU dealing with an existing problem in a completely different manner.
The car isn't fixed yet, he took it home to check out the harness and find the issue. I did prove all my theories by shorting the knock sensor to ground on a perfectly good running Chrome car I had at the shop, and exactly duplicated the symptoms his car was having...
The guy that owns the car is an experienced tech at another shop, and is no dummy, but wasn't as familiar with Chrome as I am. (turns out I wasn't as familiar as I thought I was...) He had wired in and installed a 99 ECU, and had me flash Chrome into it. The car previously had run fine with the 1G ECU, but was setting a knock sensor code. He replaced the sensor at the same time he replaced the ECU.
Symptoms presented immediately. it would start and run fine until you took off and got the rpms up. It acted like it hit anti lag or some type of limiter, and the CEL would pop on and stay on until you turned of the key and reset it, as well as just run like crap... We were convinced we bolted in a new problem with either the ECU, the wiring, or the tune...
I'll spare you the details of wild goose chases and snipe hunts embarked on, but here's what I found: Datalogs told me that once the issue presented, and the CEL came on, fuel was commanded at 11:1 AFR or less, timing was commanded in the single digits, and knock was fixed at 7-9. This showed all the signs of a rather brutal failure management strategy for a knock sensor failure. But why didn't it have any codes? and why was the CEL on? The CEL was easy... Chrome CEL on knock was enabled over 5 counts, and we were fixed at 7-9. Why no codes was a little harder, and I'm pretty sure there's a Mitsu engineer still laughing his ass off at pulling that prank... If you look at any 96 up Mitsubishi, P0325 is a knock sensor code. If you specifically look at the 98-99 turbo cars, guess what code they removed from the displayed codes menu..? Of course they didn't remove the sensor test, or the failure management strategy if it failed, there's just no code to display even though its set and being compensated for.
If you've ever experienced the failure management for a knock sensor code on a 1G, you know it's pretty mild. Pulls a little timing, fattens it up a little and it just doesn't run as strong... Don't know why they made the 98-99 so severe, but it's pretty brutal. Nothing like the 1G. It's nearly undriveable once engaged.
So on this particular car, he has a short in his knock sensor harness somewhere, the sensor wasn't the issue, and the 99 ECU with Chrome is much less tolerant, making us think we had bolted in a new problem when the whole time it was just the new ECU dealing with an existing problem in a completely different manner.
The car isn't fixed yet, he took it home to check out the harness and find the issue. I did prove all my theories by shorting the knock sensor to ground on a perfectly good running Chrome car I had at the shop, and exactly duplicated the symptoms his car was having...