permanent cel disable

[cboles]

They are all in one contiguous area. The actual location depends on the ROM. You want to be pretty sure you are in the right place before you start poking at the bytes...

[akcel]

i guess i should wait for the cable to come then to pull my own ecu rom or maybe i should go pull my carpet and get the id and junk off the case.

its an 02 wrx manual USDM

[NeverLies]

Post some patterns that you have already found (at least 10 bytes before and after the "identified" CEL), we will let u know what is the good one.

[Freon]

Anyone have an idea on the DBW ecus? I'm guessing it isn't going to be the same fix.

I'm pretty darn certain I've found the code block with all the CELs on my 500J ECU (512kb)

[akcel]

if that were the right one for cat efficiency p0420 it would be something like 04 04 20 or 03 04 20 yours is 04 20 02

I could be wrong. im by far no expert

[cboles]

This isn't a HC16 based WRX. The CEL codes could be stored differently. I haven't had a chance to look at the problem in IDA yet, but it should be pretty easy to figure out.

Colby

[x99percent]

im trying to locate cyl misfire 1 2 3 4 caused by my flywheel in the ecu via the hex editor the codes are p0301, p0302, p0303, p0304.

I would think it would be a better idea to lower the misfire sensitivity rather than disabling the misfire codes entirely...??

[akcel]

that may be an option....but if it doesnt work...i have no problem with taking out the codes. the car runs fine and the misfire codes come in at the most random of times...idle one day then cruising on the free way another and so on.

[JRSCCivic98]

It's been my experience that extending the sensativity threshold of what the ECU sees from a sensor before it sets a code is a much better way to get rid of CELs rather then disabling the CEL altogether. That way you don't run into the possibility of having issues that you don't know about because you never set a CEL for it.

For example, on WRXs, if you push the EGT limit up a CEL for no uppipe CAT will not be set because the temp range will not reach that limit. That way you still have a working EGT probe (which is used for other various map calculations), but you will only set a CEL when the probe fails, if ever. O2 sensor readouts have these same threshold limits... you just simply push them up to a level that represents a realistic number. So if you put a catless exhaust on the car the CEL won't go off for exceeding the factory limit. In my map I have "CAT Threshold" maps that have higher values then the factory ones do. So, my O2 sensors still work and have output, but I don't set a CEL because the "new" limit is offset to match a catless exhaust for instance.

[MalibuJack]

Has any of this information or data been seen/experimented with on the Mitsubishi ECU's yet? I'm fairly certain there are quite a few people interested in that.

[akcel]

It's been my experience that extending the sensativity threshold of what the ECU sees from a sensor before it sets a code is a much better way to get rid of CELs rather then disabling the CEL altogether. That way you don't run into the possibility of having issues that you don't know about because you never set a CEL for it.

For example, on WRXs, if you push the EGT limit up a CEL for no uppipe CAT will not be set because the temp range will not reach that limit. That way you still have a working EGT probe (which is used for other various map calculations), but you will only set a CEL when the probe fails, if ever. O2 sensor readouts have these same threshold limits... you just simply push them up to a level that represents a realistic number. So if you put a catless exhaust on the car the CEL won't go off for exceeding the factory limit. In my map I have "CAT Threshold" maps that have higher values then the factory ones do. So, my O2 sensors still work and have output, but I don't set a CEL because the "new" limit is offset to match a catless exhaust for instance.

thats true. i cant wait for my cable to start playing wit this thing. i cant wait to put my cel bulb back in and have it work on actual cels lol. and then i could go get carb'd without having to swap my flywheel....I LOVE THIS SITE

[xswrex]

Anyone have an idea on the DBW ecus? I'm guessing it isn't going to be the same fix.

I'm pretty darn certain I've found the code block with all the CELs on my 500J ECU (512kb)

look for an area in the rom just after these codes. It will be filled mostly with 1's, 1 is enabled 0 is disabled

[akcel]

so its supposed to be 04 in the 00 and 20 in the 01 columns?

maybe i've been looking in the wrong areas lol

[twg]

I think I've found the CEL table in my 2004 USDM Forester XT

CEL address start: 0x47658
CEL address stop: 0x482CE

When using EcuEdit, use row length of 10, Power `1 byte, View in Hex. As someone has said above, the first 2 codes is the CEL code, followed by a numerical value.

I compared the data that I have and found that the way that other tuner softwares disable the CEL is as follows:

There is another table called the CEL enable table located just prior to the CEL entires, so in my case, they are located from 0x475B4 to 0x47657

This table consists mainly of 1s, those denote CELs that are enabled, and 0s indicate those that are disabled.

The first entry of this table (at address 0x475B4) corresponds to the first CEL code entry, ie. if this address contains a 1, this means that the CEL code indicated at 0x47658 is enabled, and if it's a 0, means it's disabled.

Some other examples, the P0420 code is listed at address 0x47B44. The CEL enable table entry is at 0x475F3, so to disable the P0420 code, you would change the value at 0x475F3 to a 0.

I've created a formula for finding the corresponding CEL entry given the address of the CEL code (ie. if you did a search for 04 20, and found address 0x47B44, inputting this into the formula would give you 0x475F3).

Note that this formula only currently works for my ecu, which is the USDM 2004 Forester XT, A2ZJ500i_AJ470-7312:

This is all done in hex math

(CELCodeAddress-0x47657)/0x14+0x475B4=CEL enable table entry

If I have time, I'll check to see if the STI works that way as well...

[twg]

The STI seems to have the same format:

(OBD2code_address-0x450CB)/0x14+0x45028

This would be the correct formula for the A2ZJ500J ecu...

not tested, would appreciate if someone could test to see if this works...