Attempting to write new DBW ECU ROM w/o success

[twg]

I have 2004 USDM FXT ecu... 2 in fact.. both have been ecutek'd... unfortuntely, I don't have the original fXT ecu, looking for an oem right now...

I'm changing some simple things like disabling some CEL codes... and also raising the rev limit...

I'm pretty sure i'm doing the changes properly... even raising the rev limit, I used the 1D address from the file posted by another user.. and the values make sense... I changed them to the STI limit, so nothing crazy..

however, whenever I make any change, be it CEL disable, or rev limit increase.... when I flash the new ecu, my CRUISE and CEL light are on, and when I start the car, the idle is really lumpy and surges...

I haven't driven the car yet...

Am I doing something wrong, or is there a ECUTek' checksum routine that prevents any changes to the ECU data ?

btw: the 2 ECUs were tuned by different tuners...


***EDIT***

Actually I tried to flash the 2nd ROM into the 1st ROM, and the same thing happened... so it seems like the flash itself isn't working properly... or I'm doing something wrong...

The SW seems to indicate the flash is ok however...

[Freon]

What happens when you flash back to your original backup of the ECU? I hope it works fine again. If not, there may be a problem with your backup, the cable, the software, etc.

Maybe a simple checksum could force a CEL, that doesn't sound too crazy. Did you try to pull the code?

[cboles]

One issue I will be addressing shortly in the next release is a feature where the kernel resets the ECU after a reflash. I have heard reports of rough starting the first or second time after a reflash without the ECU reset. Those who reported that said that the problem went away after that.

Colby

[twg]

I have 2 ROMS:

Rom 1 and Rom 2

I have images of both ROMs..

Currently, I have Rom 1 in my car... when I reflash anything other than the original ROM 1, I get a flashing CRUISE and CEL and the car idles very rough... I've checked the code, there is no code (at least not thru the Tari logger).

I have also "reset" the ecu using the tari logger hoping that will fix it, but it didn't.

So far, I have tried simple modified versions of ROM 1 - didn't work

I have tried an exact ROM 2 image - didn't work

cboles: Do you have an image of a stock 2004 FXT ecu that I can compare it to ? I'm wondering if there is additional code in the ECUtek tune that is preventing me from flashing the ecu ?

[cboles]

FXT ROM:

http://forums.openecu.org/viewtopic.php?p=4214#4214

[Freon]

I have the same problem with my 2004 STI 500J AF241.

I flashed a ROM where only the CEL for P0420 was changed and it wouldn't idle correctly. Idled at 1200rpm, 25:1 AFR (twenty five to one), and 16in Hg instead of 600-800, 14.7:1, and 19in.Hg.

Something is amiss. Factory ECU CRC check on bootup or something?

[cboles]

Here is your solution:

http://wiki.openecu.org/index.php?title=Subaru:DBW_Checksum_Algorithm

I'm going to add checksum correction to EcuFlash right away...

Colby

[Spiider]

I recognise some of those item Freon mentioned as normal after reflash behaviour for my car as well.

When I reflash, I normally get 25:1 and 1200rpm etc, until I warm it up for the first time and go for a little drive. I think this mught be some sort of state that occurs after a reflash. I know that I don't have any 25:1 spots in my fueling map. Lol.

[AutoXer]

I thought EcuTek'd ECUs were "locked"? What am I missing here?

[Freon]

I definitely did not attempt to "drive through" the problem. After I flashed back to stock I had no idle issues. The car was already warm so idle settled to ~800rpm @ 14.7:1 within about 10 seconds on the stock reflash. I only let it idle for about 20 seconds on the modified ROM.

1200rpm idle may be normal for a cold engine, but 25:1 AFR definitely isn't. In fact, I typically find immediate startup AFR is about 10.5:1 and it slides to 14.7:1 in about 5-6 seconds after startup. The vacuum was clearly off as well. Even on a cold engine start @ 1400rpm vacuum is barely affected.

I'm positive what I saw was not "right" at all.

I will try to run another test tonight with checksum disabled or corrected and see what happens.

Autoxer: My stock ECU does the same thing as twg's Ecutek. Perhaps Ecutek's checksum is different, or needs to result in a different value, but it seems like it is essentially the same problem. TWG, I'd try the same fix on your 2nd rom and see what happens.

[cboles]

I don't think there is an EcuTek issue here. It is just the normal checksum system.

[swifty]

Have you compared your ecutek'd roms with each other or with the standard rom that Colby posted ? Can you 'spot the difference' ?

The reason I ask is that I was reading a thread on scoobynet the other day where a number of ecutek tuners were discussing remapping customers cars that had already been mapped by another ecutek tuner. They were saying that if tuner A maps a car then tuner B looks at it, tuner B cannot see the changes that tuner A has made. They can only see the stock map so second time around the amount of mapping involved is the same as the first, ie they have to start from scratch. The license remains on the ecu so a new one is not required.

If this is the case, it would imply that all the ecutek changes are held separately to the original maps and are somehow only accessible by the tuner who made them, some sort of authentication involved, so when a second tuner comes along their changes must wipe out / overwrite all the changes that were made by the first.

I'm just curious to know if the changes made via ecutek will be viewable when a rom is downloaded via ecuflash and opened up using ecuEdit.

thanks

[fastchip]

A graphic view of the DBW crc area

[Freon]

swifty: if that were the case, I would say the reason would be that the Ecutek software doesn't read the ROM at all, it only writes. All source files would be present on the PC, not in the ROM. That seems like a simpler explanation.

Either that or the PC side software checks two keys. One for Ecutek license period, and a second for that tuner's specific ID.

I'm not sure it matters. I'd bet you can openly read, modify, and reflash a previously Ecutek'd ECU without any more issues than doing the same on a stock reflash when using the openecu tools. TWG should be able to confirm if he can hack up his second rom and try again.

[Freon]

Seem to have found it. I'll modify my tweaked ROM again tonight with this fix (exactly as pictured) and try flashing again.