Forum » General Tuning Discussion » WRX / STI Ignition Timing on Pump Gas... MBT possible?

WRX / STI Ignition Timing on Pump Gas... MBT possible?

General Tuning Discussion

Discuss all things tuning in this section. News, products, problems and results. 

Page 1


I have a doubt some time ago. I had road tuned several WRX/STI cars in my city. There is no dyno and the closest one is at 1000km from where I live, so we are forced to do the tuning at road and to find the best timing I usually use the rule of getting to the knock point and then getting it back a couple of grades to be safe.

We use 98 octane RON gas pump (Peru). By local laws I know they have 7.5% ethanol on it.

I feel these JDM WRX/STI EJ20 (specially) take a lot of ignition timing. Some times I keep advancing the timing and they keep taking it, but somewhere I feel there is no gain advancing the timing more. So I was wondering. Is it possible to reach MBT with pump gas in these engines? Or is it possible to even exceed the MBT point and start loosing torque or is impossible to get to MBT with pump gas? Or is the MBT in these JDM WRX/STI at a point that the engines do not knock (with pump gas) and you could exceed the optimal timing point?

Some times I feel a lot of more torque with not that much timing advance. And some times I decided to go further with timing and the cars felt not that strong. By rule I leave the timing tables where I feel the car runs better, but I know when road tuning with pump gas the car will be knock limited (in theory), but as I wrote before, I feel like sometimes I am exceeding the MBT point... Am I crazy?

Sorry for my english, I hope you can understand my questions.

Best regards,


Are you recording any knock events? I'm assuming you are reflashing the stock ECU.

With the JDM wrx/sti on moderate to low boost, it should be possible to meet or exceed MBT. There are several ways that you can collect data log pulls in 3-4th gear and use spreadsheets or similar to plot a simulated dyno pull while street driving. You can take 3 runs and average them together, then compare with another 3 runs with higher or lower timing. If power goes up you are on the right track, if power seems about the same, you can pull it back 2 degrees at leave it there. As long as you are not in a knock limited area of the rpm range this should get you close. By overlaying the graphs, you might be able to see certain rpm ranges might need more or less timing depending on the shape of the graphs.

Good luck

Thank you for your responses.

Yes, I'm loggin knock events and no problem at that area. I'm reflashing the stock ECU with Tactrix cable and Ecuflash. Romraider is used to log.

I will use Virtual Dyno software to plot the results. I've read that maybe it is not exact but can give me some ideas how things are going.

So, I assume MBT is possible then. Have to trust the simulated Dyno results.

Thank you very much.


Try to use the same stretch of road from the same point every run.

the VD software might not be accurate on the figures but that's not what we are using it for in this instance.

You are using it to compare the changes made graphically and for that it does a good job as long as you don't log too many parameters at the same time and the conditions as similar.

I have found the WRX to be a bit more Knock limited VS the STI by up to 4 degrees in most cases. This is comparing ej255 vs ej257.

Hunt the knock threshold then pull 2 degrees as you find it, and also don't over speed the td04 as it just turns into a hairdryer..

AVCS tuning can also make a big difference and will affect the knock threshold.

Enjoy and watch out for the birds...

I will follow those steps for sure.

AVCS tuning its something I've not researched about yet. Do you have any recomendations for reading about?


The ROMRaider howto has a bit on it if memory serves me correctly but the different sources and opinions on the interwebs made my head hurt so I decided to test it for myself.

What I would normally do is pull 10 degrees of timing across the board, then do some logs with 0-30 across the avcs map in 5 degree increments logging power ,torque, boost and VE.

Then extrapolate the results, compare and build a table accordingly.

Log some on the new table and attempt to fine tune it a bit then start bringing timing back at 2 degree increments.

This normally give you good results but does take allot of time if your road tuning.