Tim D

I see a couple of things from the log that might be worth a look (pic attached) 1. It looks quite rich (0.68) at the start of boost, before stabilising later. 2. Boost Mode is set to 'closed loop' but the Target Boost Table only has a row for TP = 0%. It should include rows up to TP = 100%. Boost Status is shown as OFF all the time, due to the above I expect. Hope that helps?

Hi Adam, IAT in that region was around: 10 C winter 35 to 40 C summer

Hi... Please could someone advise on a Det query on Impreza STI? I spent a lot of time setting up knock control (frequency analysis, det cans, inducing controlled knock etc...) and got a really responsive setup and running completely knock free. This was in the summer months when it was 20 degrees C. Now it's colder (freezing), on occasion, the system detects a single knock event around maximum torque (5500 rpm, 240 kpa, see attached). Would you use the IAT Ign Trim table to resolve this? It's currently set to remove timing at high IAT only (see attached). Thanks

Nice work Brendon, thanks for the feedback and apolpogies for my suggestion... Tim

Hi Brendon, With regards to the hot-start issue, in my opinion, this will be trial and error and I guess you need more fuel to get it started (assuming it doesn't already smell rich on cranking?). You could try increasing the values in Pre-Crank Prime table (since it's not even starting). If the 'hot-start' problem is repeatable, you should soon find out if you're improving the situation.

Hope I'm not treading on your toes Adam, I was just replying too....... :-) The RPM lockout for Idle Ignition Control is set to 1200, perhaps set this higher so that the ignition idle table comes into play and bring the RPM down. The Idle Ignition Table at zero error should be 16 (not 12) to help with a smooth transition between tables (maybe add 4 to all values as a starting point). The Overrun Deactivation Table looks to have rather low values (at normal operating) temperature, which may cause the RPM undershoot (as it goes lean on the approach to target idle RPM). You could even turn this off until you're happy with the idle. The Overrun Ignition Trim (linked to the above) also causes the ignition timing to be retarded (by 10 degrees) on the approach to target idle, further causing RPM undershoot. It may be useful to switch Idle Speed Control to Open Loop to carry out the above tweaks. Then once the idle is a good as you can get it, switch it back to Closed Loop. I have been through this process a few times and had great success with this approach :-) Re your comment "Ideally I would like the rpms to drop and then hold above the RPM target and then slowly come down to it" I agree, but can't see an elegant way of doing this. Hope that helps...

I was waiting for someone else to respond, but in my experience, tuning my fuel table I couldn't have predicted any of the cell values, I believe you need to 'access' each and every cell to do it properly.

Take a log and capture these parameters (assuming you're running closed loop), it makes analysing and tweaking a piece of cake!

Did you mean this...

Ah yes, agreed, that looks good, hope you didn't mind me commenting... Tim

Something else might be worth a look... the boost control looks a bit odd? As the RPM increases, the boost control goes from stage 1 to stage 2.... but not stage 3, it reverts back to stage 1 causing the boost to drop off. Wouldn't you want to maintain boost as far as possible?

Thank you Simon.

Hi, please can someone advise on Launch Control Fuel table? The help file says: "This table is provided to inject extra fuel during launch control based on two engine parameters" Do the cells in Launch Control Fuel table mean % addition of fuel? Does a value of 0 in the table means no change? I guess you'd only want to add fuel to table at 100% throttle? Thanks in advance for any help!

Out of interest, I just looked at my Target Lambda (as it was tuned by a professional tuner, I'm not experienced in this area!!) and mine is 0.79 agreeing with Brad's recommendations.

Here's my theory! I think you've reached the MAP limit. I expect car revs beyond 7000 rpm when not under load (ie MAP limit not reached) Also noticed the knock threshold is exceeded (albeit at lower rpm, so not the cause of this problem).

Thanks Adam, I will amend to 0.5 s as you suggest. Cheers Tim

I have a turbo dynamics query... Please could someone advise what I would use Deactivation Timeout for? It is currently set to 2 seconds on my ECU, hence it adds a 2 second delay when returning to Stage 1 (from Stage 2). I don't understand why a delay is needed at all? The car is a Subaru Impreza (EJ257). Thanks in advance....

Hi Rob, Interesting... In that case, I'm out of my depth, I will leave you in the hands of those who know... Couple of final thoughts... You can log to ECU at 100 Hz, that might show peaks that are 'missed' by laptop logging? You can use Knk Ign Trym Cylx to show where timing has been pulled. I'll keep watching this post out of interest though............

Hi Rob, Firstly, my experience is limited to my own car only... but I was puzzled why you were only logging/showing knk level for 2 cylinders? I understand wasted spark, but surely you need to monitor cylinders 2 and 4 to set the gains in the same way as you have for 1 and 3? Otherwise these may be exceeding your threshold and causing timing to be pulled?

Do you definitely only have knock detection enabled on 1 and 3? Could you post a map and log file...

I also have an EJ25 with forged pistons etc... I found that my engine 'background noise' was too 'noisy' to have a threshold of only +20%. I also noticed a momentary increase in noise during the transition from 'on throttle' to 'off throttle' (clearly heard on det cans and visible in screenshot attached). Cold running also produced different noises to hot running (due to forged pistons I guess?), not that I was driving it hard when cold! When inducing controlled knock to test my setup, I could easily detect knock whilst not giving false knock detection. See attached for threshold vs cylinder noise that worked for me. Note, where the threshold is exceeded, throttle is 0%, so knock system is inactive and timing is not pulled. Not sure if that helps!

I agree with Brad, the ECU doesn't appear to have knowledge of the wheel speed (assuming your laptop logged every parameter?), so that suggests that the cluster is getting CAN data directly and not from the ECU I think. Can you get to the cluster to see what's connected (maybe?)

For what it's worth, I wrote up my experience of setting up knock control on my Impreza STI 2007. It's all based on the Link help files... I believe what I have done is correct (comments welcome!) Feel free to have a look at the attached word doc... BTW, I found the sound levels of each cylinder to be quite 'noisy', so had to have a fairly large headroom in threshold to prevent false knock detection. Knock Setup.docx

Just an idea... Perhaps you could try logging to ECU at 100 Hz, only logging the relevant parameters at this rate though. Might help to understand whether the PC just can't display/log data fast enough (as Adam was suggesting)?

Hi, I only skimmed through this post, apologies if you've already fixed this...