Codes stored in a 2004 Subaru

[mr.radon]

This is for a friend in NY. He's got a 2004 Subaru. His car has thrown P0453, P0457 10 times. 9 under warranty and now the first time out of warranty. (the dealer even replaced his car's entire electrical harnesses tryiong to fix this issue)

 

Does the ECU record the date, time and ODO when a fault code sets? I work in the heavy duty truck world and it does record this information. Anyone know?

 

Thanks.

[Legacy777]

No, unfortunately it does not record time & date stamp to my knowledge.

[OB99W]

[...] He's got a 2004 Subaru. His car has thrown P0453, P0457 10 times. [...]

Here are the definitions for those codes (plus one more):

P0453 Evaporative Emission Control System Pressure Sensor High Input

P0456 Evaporative Emission Control System Leak Detected (Very Small Leak)

P0457 Evaporative Emission Control System Leak Detected (Fuel Cap Loose/Off)

Here's a link to P04xx code info:

http://www.endwrench.com/images/pdfs/CodesRevealed.pdf

 

I included P0456 because it contrasts with P0457. If P0457 is being set, it seems that the ECU ''thinks'' that there is a more substantial EVAP system leak than would be the case for setting P0456. A leak of that sort should be possible to find as in http://endwrench.com/images/pdfs/EvapSpring04.pdf (assuming that the gas cap being left loose isn't the easier answer).

 

Date/time stamping would require a clock, a means to reset it, and a backup power source. The ECU does record ''freeze frame'' data, which is a snapshot of the conditions under which a code is set. That data in conjunction with EVAP leak testing and circuit checks should lead to a resolution of the problem.

[GeneralDisorder]

Have him call Subaru Of America. That's BS and there's lemon laws for crap like that. At the very least the warantee work that was done most recently should be under it's own warantee and they need to make this problem right since it's had the same issue since day 1 and it's obviously never been properly addressed.

 

GD

[mr.radon]

Update.

My friend got back from NY. Needs to get emmissions testing here, P0453 is setting again. Measured the Voltage, 4.63V on Vsensor for the Fuel Tank Pressure Sensor. Out of spec high. Doesn't make sense...so I told him to bring it by.

Got my friends car on my lift. Looked at the Fuel Tank Pressure sensor. I could NOT BELIEVE what the dealer did! The dealer cut off the Vref (red) and Vsensor (orange) wires to the sensor and crimped on two wires. They left the Vcommon wire (black) alone.

The two wires are 16AWG(Vref) and 18AWG(Vsensor), the 16AWG wire showed signs of having a thermal event. They ran these two wires from the sensor connector with zip ties to the grommet in the right rear fenderwell into the inside of the truck area, from there they wrapped the two wires in electric tape until they reached the engine ECU. At the ECU they used a power taps to connect to the OEM Vref and Vsensor wires. WTF???:mad::mad:

The old wires under the car were not even capped off. I'll post pictures when I get them downloaded.

I measured the resistance across the Vref and V common in the Fuel Tank Pressure sensor. (5K ohms) Then I tested the sensor, there is nothing wrong with it, ran the pressure up and down. Since the sensor has 5Kohms across it, so I got two 2.5Kohm resistors and soldered them together then soldered them to the three wires going to the sensor. Now the ECU will always see 2.5V on the sensor from the Fuel Tank Pressure sensor line.

So far no fault codes. I can't believe it. He took this car to the dealer nine times under warranty to get this fixed, the dealer wrote that he replaced the main wire harness, BS...that wire assembly still has the factory markings on it.

According to his warranty report, the sensor has been replaced 4 times. That sensor cost $150 either at the dealer or Rockauto. He got a lemon. I used to have a high opinion of Subaru, this has left me shaking my head.

This has also validated a principal I have,

NEVER GO TO A DEAL FOR ELECTRICAL SERVICE.

Edited November 15, 2010 by mr.radon

[hankosolder2]

If subaru's evap emissions control logic works anything like Hondas, I think the car will eventually set a code due to your resistor voltage divider fakeout. The ECU cycles various evap control solenoids and looks for an appropriate change in the tank pressure.

 

Frankly, I'd rather have a well done repair to the existing harness than have a dealer tear the entire car apart to replace it. It sounds like they half-assed it for sure in this case.

 

Nathan

[mr.radon]

I'm assuming the ECU is too sensitive. It sees a short (millisecond) over voltage on the Fuel Tank Sensor and reports it as a fault. I checked there is not time limit for the 4.5V setting off the fault code.

Maybe I should have soldered a 7805 voltage regulator to feed this sensor independent of the ECU. But what the dealer did was way half assed. Best of all is reviewing the dealer print out and it says right there, main wire harness replaced, but yet the main wire harness in the vehicle has never been replaced.

 

I would love to know if anyone knows if the ECU is programed to look for variation in output. If so I can get a bread board out and make a circuit to fake a slowly changing voltage. But I doubt it.

Better yet would be reflashing the ECU so it doesn't trip the fault code on at 4.5V.