RomRaider

Ok, before I get started with my comment, let me mention Im in above my head with understanding how everything works together in reading the ECU accurately.

With that said, I've been trying to log with my current version of romraider on my 2008 STi, but can only pull some extended parameners. I've been trying to log knock (or hopefully the complete lack there of) but there is no check box to select Fine Learning Knock Correction or Dynamic Advance (seems to be a correlation for detecting knock in what Ive read).

Is this just because romraider doesnt yet fully support the 08 STi or am I doing something wrong? Ive checked the FAQ's but couldn't really come up with a good answer.

Thanks,
Chris

You won't be able to log any extended parameters (indicated by * at end of name) with the 08 STi because the logger defs have to be customized for each ECU. This means that someone would have to use Ecuflash and a tactrix cable to read a copy of the ROM from their ECU and then post it up in order for me to add support. Since Ecuflash doesn't currently support the canbus ECUs (such as the 08 STi/WRX, 07+ LGT/FXT), there's no way to add extended parameters or tuning support for these cars. Apparently Colby of Ecuflash has made good progress in hacking canbus, so sometime in the future this may be a possibility (a new tactrix cable will be needed as well).

How about the knock sensor #1 (and #2) selections on the lower left selection menu (on/off). I've check marked both of them and had what appears to be an event (1) a couple times under partial load. Is this an accurate detection of knock or just a fallacy in my mind?

Thanks again,
Chris

knock signal #2 is not applicable to your car as it has only one knock sensor (unless they changed that on the 08s -> all other Subarus except for H6 models have only one knock sensor).

You can log knock signal #1, however there is not enough resolution when logging for it to be typically useful (meaning you'll miss quite a few events unless they are prolonged or sampled at just the right time).

I see. Correct me if I'm wrong, but wouldn't a significant knock event usually be within a certain range, and not just a certain point? In essence the probability of having a knocking problem would be at least recorded within the recording interval, though not the exact extent of what it is?

I appreciate the responses so far.

merchgod, have you received a copy of the 08 STi rom from Colby? I believe gabedude said that he has already extracted it from an ecu (using a prototype cable I assume) and that he is getting ready to start making the production CAN cables.

The ECU reacts quickly and you will not always "catch" the knock event by logging the knock signal. This is because when logging, you are taking snapshots of each parameter at particular point in time. The more parameters you log, the greater the sampling interval. Even if you only logged "knock signal #1", it still may not be useful as a lot is occuring in-between each line of your log. The reason logging FBKC, FLKC (change), IAM (change) are more useful than knock signal is for several reasons. FBKC has a delay after the knock period has ended where it maintains the negative correction (therefore you can catch it in the log). FLKC is learned correction, so that correction remains when changed (same with IAM - if your IAM was 16 and then is 12, you know knock had occured). The knock signal is simply a switch - either on or off so there's no history involved and there's no delay (when there's no knock as perceived by the ECU it is immediately "off").

No, I haven't. I sent him an email on Friday but haven't heard anything from him. I'd like to get started on it so that by the time the new cable is released I can have some preliminary definitions. But Colby is often hard to get a hold of.

Great, thanks merchgod