limaynard

Just looked at your website - so you did them remotely - customer sends you a ROM and logs, you modify and send back - great idea...

tscompusa

iTrader: (31)

Ya i really enjoy doing it. Greg will get you sorted out with the xml files for the Airtrek.

limaynard

He has already sent it to me - I would have wasted months getting to that point...

limaynard

I did some Evoscan logging this morning. I was surprised at how much knock the standard JDM tune had

I downloaded the log to Excel, did some filtering and created a pivot table showing RPM rounded to nearest 500 and a calculated engine load rounded to nearest 10. The knock sum I diivided by 3 to give the degrees to reduce the timing advance and added back 1 (didn't want to be too conservative)

I cut and paste the timing table from ECUFlash to Excel and overlaid the above figures before cutting and pasting back into the timing table in ECUFlash - easy.

Took it for a run with a Knock alarm in EvoScan and it was considereably reduced - only knock when under full throttle and high boost.

Looked at the stats and it needs another degree of retardation in some areas of the map...

Just reflashed the ECU and I'll take it for a spin later...

So in summary, the timing needed to be retarded 1-4% from 120-200% load and 3000-7000 RPM

limaynard

All knock now gone. Retarded ignition timing another 1 degree in a very small areas of the map to remove final knocks. 5 degrees total in some areas...

Now to attack other areas of the tune. I will probably remove the Jaycar Electronic Boost Controller that I fitted, revert to stock solenoid and tune the boost maps.

Then who knows, maybe some tweaking of the air fuel ratio.

If anyone wants the files then let me know.

limaynard

Well, looks like I have finally chased down most/all of the knock.

On Australian VPower 98, I needed to take out up to 10 degrees at high RPM / high loads...

I also modifed the AFR to make it slightly leaner in the cruising part of the map and also a bit more lean in the high RPM / Load area while being careful not to overdo it - it is still very rich.

I also increased the boost limits very slightly in 3,4,5 gear table and in the overall boost limit table - not sure how the three boost limit maps work together but I'm sure trial and error will help.

It certainly feels quicker (very subjective)!

I am also planning to go back to the stock BCS setup (or pretty close) and remove the current (sort of) closed loop system I have as it work by using injector duty cycle to estimate load and then varies the solenoid duty cycle to decide how much air to allow through to the wastegate actuator based on a 2 custom programmed boost curves.

This always worried me as it is normally closed so any failure would cause no air to get to the wastegate actuator and with higher boost limits I don't want to take the risk. I can then tune the boost curve through the ECU...

GregM

iTrader: (1)

The boost limit table doesn't work. The 2 separate ones do. You may need to up the fuel cut as well. The xml I sent wasn't my latest, (just so you know Tom) and doesn't have the knock tables.

I have a Grimmspeed MBC in my Airtrek, but run their 3 port in my Evo. I don't think you'll get much boost out of the stock boost controller. I'd get a decent MBC, its simple and easy.

What wideband are you using? I found my fuel pump wasn't keeping up and I was running lean.

limaynard

Thanks, Greg.

I have just ordered a tactrix 3 port BCS to replace the stock. Thanks for the tip on the boost limit tables. I will tune using the 2 separate ones.

Not using a wideband sensor (as yet) .

I am goingto (try to) stop messing with it until the BCS is fitted as I will end up re doing it all.

limaynard

Fitted the Tactrix 3 port BCS. I had to fabricate a bracket but it was fairly straightforward. One thing though, they work so much better when you plumb them in correctly

Now I have sorted out the pipework and populated the WGDC tables it boosts to about 12.5 psi. I have set 100% duty cycle up to around 2500rpm then tailing off to 60% WGDC and then slowly climbing back up to 100% as the revs climb.

I didn't want to blow it up on the first outing although I did have a shock as one of the intercooler pipes blew off (not tightened up enough). It made the engine run like a dog until I put it back on...

I need to get my EvoDroid so that I can wire my 3 Bar MAP sensor into it so that I can log the boost and do some more accurate tuning.

Greg, I also spotted an issue with the XML definition - when you decrease the Boost Cut Delay timer, it also reduces the Fuel Cut pressure! I suspect that with the two tables having adjoining memory addresses that there is some overlap. I will do some research..

GregM

iTrader: (1)

ok, I haven't played with the boost tables as I'm running an MBC on my Airtrek.

limaynard

Still waiting to tune the boost curves properly - current duty cycle at 4000 RPM is around 69% (as opposed to stock of 82.5% with stock solenoid) and boost is probably 3 PSI higher.

Using the Tactrix 3 Port BCS gives a lot more range to play with but it could be easy to overboost without changing stock values...

Also, changed the values in the MUT Table to enable the 2-Byte Load Logging.

For reference, the 00 value is 6a36 (decimal 27190) and the 01 value is 6a37 (27191).

ReaperX

iTrader: (5)

Sounds like an interesting and fun rid you got there. Looks like you doing everything you need to do to work your way up slowly on her.

limaynard

Yeah, it is fun. Not sure where to go next to tune, I am slowly raising duty cycle on solenoid to get boost back to 1 bar and ,aking small incrementL steps to add back timing advance now the knock has gone.

Maybe it is time for more serious stuff. Anyone have any experience with the TD04 ported to take a 19t turbine?

fred sampson

Hey guys,

I have been told about the forum and joined up

I have the same issue disconnect the green wire and got the read but its 256k :S , how else can i get the full file to map it , been trying for last couple of days and i am unsuccessful

can someone please help me

GregM

iTrader: (1)

pm sent