PDA

View Full Version : soldave's Evo I tuning thread


soldave
09-02-2009, 12:40 AM
Andy: Wasn't sure whether to put this in here or in the general forum. Hedged my bets on here but feel free to move it.

Now this thread is a continuation of a thread I've had going over and DSM Talk (http://www.dsmtalk.com/forums/showthread.php?t=203327). As quite a bit of my talk is around the ECU, I thought I'd transfer the thread here so I can get advice from you guys and share experiences.

Did a little testing at the weekend and was shocked when I finished my first power run. My Pocketlogger was reporting 0 knock counts for the entire run. Time to put some much needed timing back in! I ended up putting in 2 extra degrees of timing from 4000rpm upwards, and even then a 4th gear run only gave me a split second of 5 knock counts before dropping off again (as you can see in the graph and linked datafile below). I would have added another degree of timing in along the range (except for at the 6500rpm mark where I got the knock) as I reckon I could get even more power out of the car, but I ran out of time. Am using timing to increase power now as I'm peaking at about 1.8-1.84 bar of boost and don't want to go any higher than that.

I haven't done a 1/4 mile run for what seems like an age, but I did a couple of 1/8 mile runs last night. Managed to ease the clutch out on the 2nd run and launch with just a couple psi of boost (as opposed to my usual floor the gas, dump the clutch). Got a bit of tyre spin in 1st but controlled it, 2nd gear hit pretty hard and I'm pretty sure for a split second the clutch was slipping (could do without that, but then again it is a Centerforce single plate only rated to 300hp and I'm getting waaay over that now). Ran pretty well and was very pleased to see me get a 7.52s for the 1/8 mile; shaving 0.11s off my previous best. DSMs that have been running those sort of times have then gone on to run 11.3-11.5s 1/4 miles so we'll have to see how I go on with that one. I know that the way the Evo gear ratios are I will always get lower exit speeds and that will hurt me a little but it's feeling pretty good at the moment.

http://img172.imageshack.us/img172/3360/090112211901ik0.jpg

Here is the latest log from a power run with the "winter tune" in place.
http://img172.imageshack.us/img172/4876/logfromjan10thtv1.gif

And click here to download the full log in CSV format (http://soldave.ismysite.co.uk/jan10th.csv)

And then finally for reference, here is the timing map I am currently running in my car:
http://img120.imageshack.us/img120/245/timing10janhg9.gif

Will post up my full bin file if anyone is interested.

Andy, we talked about leaning the fuel a little between 2000 and 3500rpm to aid in spooling, but what load levels do you think I need to be leaning out at, or should I just drop fuel by half a point on the map along the entire top half of those rpm values?

soldave
07-07-2009, 02:37 AM
A rebuilt engine and a car badly in need of a tune... seems like an ideal time to revive this thread and talk about my (mis)fortunes of ECU tuning. Since the engine blew I have take a degree of timing out of the top 3 rows of the timing map shown above as a precautionary measure, and I have also turned off my boost controller until I get a tuning session done. So while the car moves pretty well, it feels so underpowered at the moment and my power figures are about 330-340ps (about 100ps lower than I was before).

Once I've done some tuning I'll report back with the results and what I changed to bring up the power but retain some safety in there. Andy's new knock-cut feature should help a ton with that, too.

AndyF
07-07-2009, 08:20 AM
Will be good to read about it all again Dave.

Should hopefully have your new ROM ready for you this evening with a bit of luck.

Will add in the knock cut, cel flash and also map tracing for you ;)

Do you want the multi-stage launch control as well?

Andy

soldave
07-07-2009, 09:01 AM
Multi-stage launch control would be great to play around with as a crude type of traction control.

AndyF
07-07-2009, 09:26 AM
OK, I'll add that in as well.

If you want to test out the O2 sensor feedback disable to allow the closed loop target changes to work, I'll put the code in the email and you can add it if you want to.

Cheers

Andy

P.S. You need to modify your 'limp home' ignition timing map as well as there is a huge chunk of it that is not correct.

P.P.S. There are also certain parts of your low octane ignition map that have more advance than you standard map so probably best to change those as well ;) ;)

soldave
07-07-2009, 10:05 AM
Thanks for the P.S.'s . Don't think I've ever looked at my other maps apart from the main one but will copy them across from my stock bin.

And I'd be happy to beta test and give feedback on disabling O2 sensor feedback. Just to confirm, it should give me the target AFR (or thereabouts) when the car's idling, right? Not just when driving?

AndyF
07-07-2009, 10:40 AM
It should do Dave but obviously this will depend upon on where you have your wideband placed etc. When I run with mine in the tailpipe I get wierd readings at idle due to the turbulence a the exhaust tip pulling air back into the end and making the readings appear lean. With it in the downpipe then it's not so much of a problem.

Andy

soldave
07-07-2009, 11:48 AM
It should do Dave but obviously this will depend upon on where you have your wideband placed etc. When I run with mine in the tailpipe I get wierd readings at idle due to the turbulence a the exhaust tip pulling air back into the end and making the readings appear lean. With it in the downpipe then it's not so much of a problem.

Andy

Well that's one thing I've learned today. My O2 sensor is in the downpipe but will see what kind of readings I get from it.

soldave
13-07-2009, 01:15 PM
Thought it about time for a bit of an update. I only managed to do a couple of logs this weekend, but they were successful, and despite being a little down on power due to not doing any actual tuning during the session, I was getting almost zero knock (just 2 knock counts for 0.1s at around 6500rpm). Might just throw in a bit more fuel there to sort that out. I was hitting about 1.65 bar of boost and I think holding about 1.6bar right up to redline.

Note that this was with a degree of timing taken out of my current map (which you can also see below, so you know what I'm starting with) at the top end, to be a bit more conservative pre-tuning. Note that the fuel map below does look a little strange, but that's mainly as I'm running meth which starts to come in at 1bar. Once I've got the fuel load mapping sorted out I can make sure that's all working as it should. Have to say that even though my overall power figures were a little down as noted on my Rev/Speed Meter (391ps) it felt quick and there is still some boost and timing left to be put back in there. You can tell my power is a little down too, looking at my injector duty cycle.

Am also including a CSV of my full run, for completion's sake.

http://i673.photobucket.com/albums/vv99/soldave/website%20pics/testing.gif

http://i673.photobucket.com/albums/vv99/soldave/website%20pics/ignition.gif

http://i673.photobucket.com/albums/vv99/soldave/website%20pics/fuel.gif

soldave
15-07-2009, 12:42 PM
Well, I'll find out how good my rebuild is this coming Sunday, as I've just managed to sign myself up for the upcoming drag tournament this Sunday. I'm in the same class as usual, which is 6.5-7.0s for the 150m run (I know - it's pitiful that in Okinawa we don't even have a 1/4 mile drag strip!). Have got a 2nd place finish before, but as long as the engine and everything holds up I'll be pretty pleased. Will either be doing some tuning tomorrow night or on the drive up on Sunday morning!

AndyF
15-07-2009, 11:10 PM
Dave,
Find attached a modded bin and xdf for you :D

Multi-stage lc is active plus the knock cut feature. Haven't got around to adding in the o2 feeback disable yet but I've sorted out your limp home and lo octane ignition maps as well for you.

Enjoy

Andy

soldave
15-07-2009, 11:46 PM
Sweet. The O2 feedback disable (or closed loop disable) can wait for a little while at least. Will upload this onto my ECU before work today and see if I can have a play tonight.

What are your staged launch control figures at?

AndyF
15-07-2009, 11:57 PM
Mine are the same as in this rom but I haven't given them a proper test to tweak them yet.

Andy

soldave
16-07-2009, 12:50 AM
Right - well it looks like I'll guinea pig this one and report back with my findings!

merlin
16-07-2009, 03:46 AM
Andy, can you post up a pix of the limp home timing table/map please.
I have two unknown 3D maps in the VR4 ROM I am working on, maybe the same thing.

edit: nevermind, checked it via TP.

soldave
16-07-2009, 01:29 PM
Well I don't know if it was the staged launch control or something else (moving the rev limiter location?) but my car didn't like that new bin file. As soon as it started up you can hear relays clicking on and off behind the dash. Idle was all over the place - going to 1400rpm and then nearly cutting out at 600rpm. Put the old bin file back on and it was running smooth as silk.

Done a difference search between the bin that works and this new one, and I reckon it's got to be something either in the knock cut ignition retard (how many knock counts will cause it to retard with this feature, by the way?) or in the rev limiter/staged launch control code.

Any ideas? Have attached the file I was trying to upload to this post.

AndyF
16-07-2009, 03:43 PM
I'll have a look at it when I get home Dave and try it in my car as well to see if I can replicate the problem.

Andy

AndyF
16-07-2009, 05:40 PM
Right then Dave,
I've re-uploaded the two files to my post above and these will work.

There was one section in the knock retard code where I'd written it incorrectly and that was what was causing the problem.

Have fun :D

Andy

soldave
16-07-2009, 11:28 PM
Awesome. Will give it another go on my way to work again :)

palmeros
17-07-2009, 08:36 PM
hey guys i need help lol after jus recieved all the ostrich things to self map my evo plus another second hand ecu which i bought im having trouble with it..... as i hit around 4k revs it feels like it is hitting a brick wall and the engine light flashes like hell but when backed off the lights stop. the car drives fine when not on boost iv adjusted the boost to minimum too and is stil the same. iv got andyf basic map on there also. any advice would be great cheers andy

soldave
17-07-2009, 11:20 PM
Make your own thread!!! ;)

My first guess would be fuel cut. What is your fuel cut setting at, and what mods and boost are you running?

AndyF
18-07-2009, 08:31 AM
Dave,
Hopefully the final version for you now with a new xdf as well.

The initial one has the knock cut setup with the same value as the CEL on knock.

This one has the two seperated out independently so you can have the CEL flash set at a lower level than the ignition cut.

Cheers

Andy

soldave
18-07-2009, 08:50 AM
Nice work, sir. May all of your camels be bountiful!

Will give you a full recap of hot it performs after the drags tomorrow evening.

blackvr
24-07-2009, 03:46 PM
Andy, I used your sample bin file, wrote it to a chip and tried it out in our Evo2 gravel rally car. We will be running the E2 in a round of the Australian Rally Championship next week. I will be running it up on the dyno early next week to check things out and tinker with the tune.

Love the staged launch control.

Some questions...

1. In the fuel/spark tables, how do the LH axis numbers relate to load ie relatve to boost and vacuum. ie so I know what numbers to change when I am on the dyno. We will be running 20psi. Will this take it off the scale ?

2. Any reason why the cruise/closed loop fuel table areas are a nominal high 16 to low 17:1 AFR ??

3. To switch the ALS section off , do I just nominate say 10000rpm instead of your 4000rpm ?

Thanks for your good work !!

Cheers Mike

soldave
25-07-2009, 12:08 AM
1. Think 20psi on the timing table will be about row 14.
2. I had it set that high when trying to disable closed loop so I could idle and cruise more lean.
3. Think that would work.

blackvr
25-07-2009, 10:09 AM
Thanks for your reply...Can you suggest what row numbers roughly line up with some common load points eg 10" vacuum, 0 " vacuum, 5 psi and 10 psi ?

Thanks, Mike

blackvr
27-07-2009, 01:32 PM
I ran the Evo 2 gravel rally car on the Dyno Dynamics AWD dyno tonight... and had a good tuning session with soldaves xdf definition. It took a bit of fiddling around and trial and error to work out just what load point I was at... but got a good result with a high 11's to low 12's AFR at 22psi on 98 ULP fuel.

Was a bit puzzled when I gave it an initial run on soldaves base tuning bin, but quickly worked out that engine was using bigger 730cc injectors ! A quick resize back to the 510cc standard injectors and all was good ! The fuel and timing map was pretty well spot on for our engine.

So at 22 psi boost , and a 34 mm restrictor as required by gravel rally regs, it made 150kw , and power peaking at 4200 rpm, and the boost dropping to 14 psi at 5500 rpm ( bloody resrictor ).

Thanks guys for the tuning files. Only one thing, is there a guide available for figuring out what some of the tables are for, and what the axis values actually relate to ?

Cheers Mike ( Adelaide , Australia).

soldave
28-07-2009, 02:24 AM
Glad to hear you had some success. The only thing you'd have to watch out for is that the top load levels on my maps are made for my car which has meth injection and which will affect AFRs.

As for my tuning thread, it will keep on going but might make a slight detour for a little while. I've been asked by ECM Tuning to beta test their new ECU setup. If anyone hangs around the DSM forums you're bound to have heard of DSM Link. Well I'll be testing out the new version of it which is going to be available for early Evos. Speed density and narrowband simulation using wideband O2 sensor are just 2 of the features I'll be reporting on for a little while.

blackvr
28-07-2009, 02:55 AM
Glad to hear you had some success. The only thing you'd have to watch out for is that the top load levels on my maps are made for my car which has meth injection and which will affect AFRs.



No problems... 10 chip burns during tuning and after modifying the top row and I am very happy with the AFR's !

Mike

AndyF
28-07-2009, 08:09 AM
Mike,
What tables are you needing to know what they do etc.?
Also, I'll dig out my info on roughly what 'load' level corresponds to what column on the tables this evening for you.

Cheers

Andy

blackvr
28-07-2009, 11:35 AM
Andy, I have mainly axis values questions

1. Scales for the 2 cold start enrichment tables
2.Load axis values corresponding to approx vacuum/boost?

and what and when do these tables come into play...

3. Closed loop TPS threshold ( what does this mean?)
4. Alternate igntion map - when does this and the limp home work
5. Boost cut - how do values respond in real life ie is 255 high boost

A bit unrelated... I would like to put the 3 stage launch into my JSpec VR4 race car. Any clues where the code needs to go ? I can give you a bin if needed.

Thanks for your help.Mike

soldave
28-07-2009, 02:15 PM
I might be wrong, but isn't limp home mode when a CEL is fired up or something like that?

AndyF
28-07-2009, 02:52 PM
I can answer a couple now but the rest will have to wait till I trawl through my disassembly tonight.

3. Closed loop TPS threshold is the values of throttle position vs rpm IIRC where the ECU switches back into closed loop operation. Basically, if the TPS is below a set value (given in this table) and the airflow is also below a set value (given in another table) then the ECu activates closed loop operation.

4. Alternate ignition map is used when running swtichable maps but you also need to wire up a switch to a pin on the ECU to activate it. Likewise the idle map (auto swtich enabled) is the idle table when running on the alternate map. The limp home map (of which there are actually 2 more than the 1 shown in the xdf) is there for when a number of different sensors fail. Again though, I'll tell you which ones in particular tonight.

5. The boost cut and boost limiter are 2 seperate trigger points based on airflow values over a certain time period. maxxing them out effectively removes the cuts completely.


As for the VR4 then I have no idea as to even if the code is the same. Do you know what processor it uses etc?

Cheers

Andy