PDA

View Full Version : Fuel remapping



Patryn999
13-04-2010, 02:42 AM
Now I know quite a few people have tweaked their upper quadrants in order to tune the AFR's and build a bit more power:

However I'm a little curious as to whether or not anyone's had a go at leaning out the low loads/mid rpm settings. Mainly around 2.5-3.5k rpm. Reason is I've been looking at the tables and we don't even run stoich at these speeds, so on the open road we're throwing gas out the back.

Now I assume this is to keep egt down, however I was wondering if anyone has had a play before with a probe in to see if we can save some fuel here. Also was curious about low-speed/low-load settings, as most modern cars run 16-15:1 at these settings to maximise fuel economy (IIRC, best fuel eco is 16.2:1) around town etc. Again I assume this is something to do with keeping the engine cooler but would like to hear ppls opinions. It'll be a month or 2 until I can afford wideband and egt sensor (as well as paying car off!) so I wont be able to give any personal results till a much latter date.

AderC
13-04-2010, 08:38 AM
At cruise you'll be running closed loop rather than reading off the maps, so it will be running stoich (14.7 AFR).

It is possible to force the car to run open loop and read from the fuel tables by altering the Open Loop Load tables (i.e. lower the load values at which open loop is triggered). You could then raise the AFR values to 15-16 if you wanted to run lean at cruise for fuel economy (but wouldn't recommend this unless you're monitoring with a wideband!).

Nick Mann
13-04-2010, 08:44 AM
3500rpm is not a value you can give an exact AFR to. It depends on the position of your right foot and the air flow at the time (boost pressure).

Cruising on the motorway/through town at 3500 I would want to have stoich AFRs if possible, but in 3rd gear at 1.0bar I would want a number beginning with 11.

You can improve economy with some kind of fuel control, but not enough to justify the cost of the fuel control itself, IMO. It can be seen as a small added bonus when combined with power and response improvements though.

Wodjno
13-04-2010, 08:58 AM
As above.. Buying equipment just for bringing down fuel consumption will not make financial sense. (well maybe over 5 years or so.
Also at those rpm you posted, if your on light throttle your fueling is governed by the lambda sensor, so you will be running at stoich. The maps u are seeing lower afr's at are open loop figures and are points on the map that are crossed under accelleration. therefore differrent loads are seen compared to cruising and thus lower afr's.
If you already have the kit to control and monitor AFR's in place for playing with the power. Then u can alter crusing AFR's via a wideband controller of piggyback ecu or standalone..
Quite a few have done this via a wideband, and I have managed it via a Piggyback..

Hope this helps..

Wodj

Patryn999
13-04-2010, 09:31 AM
I bought the equiptment (or am in the process of) to tune the higher up and just to learn as much as I can. Thinking if I have the equipment and the time then why not?

And Nick I do realise that, which is why the maps are 2D tables with dimensions of load and rpm. I have a programmable ECU (7202) and tactrix 2.0 cable so once I've purchased a wideband to replace the narrow (simulation out for ecu in) I'll see what I feel comfortable doing.

Thanks for your input guys :)

Turbo_Steve
13-04-2010, 04:00 PM
Of course, the alternative is to change the stoich target (14:1) for something a little leaner. Stoich is ONLY best for emissions - at cruise engine temps should remain stable at 15:1 or even 16:1, though I'd be getting nervous at how dry that's running on a longer run: running an EGT per cylinder becomes essential at that point!!! Which means you've just spent more than you would ever save :(

Wodjno
13-04-2010, 04:02 PM
Of course, the alternative is to change the stoich target (14:1) for something a little leaner. Stoich is ONLY best for emissions - at cruise engine temps should remain stable at 15:1 or even 16:1, though I'd be getting nervous at how dry that's running on a longer run: running an EGT per cylinder becomes essential at that point!!! Which means you've just spent more than you would ever save :(

What if you move it from 1 cylinder to another after collating data /Hmmm :P Thats 1/6th of the price /popcorn

Turbo_Steve
13-04-2010, 07:07 PM
Works great unless the hottest cylinder is the last one you try? Then it's quite expensive...........

In theory you could put one EGT in each manifold and see which bank runs hotter and then work on the basis that as long as the hotter bank is cool enough you're okay. Chances are there is one cylinder that runs hotter than the others, though, either because of marginally less fuel flow to the injector, or cavitation in the water jacket, or poorer oil-flow or whatever.

The best way to do it would be to go to a proper tuning place, and either put the engine on a bench dyno, affixing 6 EGT probes, Wideband lambdas and 4 or 5 fuel pressure monitors, feed it warm air (heat up the intercooler) at high load for 10mins, lean it out, repeat, lean it, repeat, and see which cylinder heats up.

It's worthwhile knowledge that I suppose we could club together to find out, as every engine I've ever worked on has one "hotter" cylinder.

ninjadaniel
08-08-2010, 12:32 AM
You will find that our engines do run stoich above 2500 rpm at low load, you need to swap the axis on your def file. You should see a line of 14.1 almost to 100 load, across the first rpm row, if you don't then the axis aren't swapped.

Compare with stock evo map, these are all swapped axis :)

lateshow
23-08-2010, 07:52 PM
So is it actually aiming for 14.1 ? Because my wideband shows that my car is doing a loop: 14.1-14.7 So the average must be something other than 14.7. Is this normal. I'm using the same wideband to control the ecu also.




Of course, the alternative is to change the stoich target (14:1) for something a little leaner. Stoich is ONLY best for emissions - at cruise engine temps should remain stable at 15:1 or even 16:1, though I'd be getting nervous at how dry that's running on a longer run: running an EGT per cylinder becomes essential at that point!!! Which means you've just spent more than you would ever save :(

Wodjno
23-08-2010, 07:56 PM
Stoich Target is 14.7. I think Steve made a typo.. Aim for 15.1-2, or leaner. Depends how your car feels. Unless u have knock monitor or EGT's. Then it'll be by the seat of ur pants tuning.
15.2 would be best if you haven't any other monitoring except AFR.

scientist
24-08-2010, 07:32 PM
Have any of you guys tried disabling Closed loop yet?

Kenneth
24-08-2010, 10:08 PM
My car didn't start running rough until close to 20:1 AFR.

Wodjno
24-08-2010, 10:10 PM
My car didn't start running rough until close to 20:1 AFR.

But was it running efficiently /juggle

Kenneth
25-08-2010, 12:31 AM
Efficiently enough for me drive the short distance to work and back :P

Do you mean power efficiency, or cost efficiency? Not that it matters, as I didn't run that way long enough to find out.

Shtiv
19-10-2010, 01:02 AM
Daniel's comments above aare correct, the axis need to be swapped, seeing as this is about fueling, has anyone found and tested the lean spool tables yet?

phosty
26-11-2010, 06:07 PM
So Daniel/Shtiv, should the Hi_octane map look like the lower map in this screenshot?

39344

If so then it makes sense as it generally shows the mixture getting richer as the load increases for each rpm line. But anybody any idea why they lean it off slightly aove 6500rpm?

I thought that regardless of the ecu operating in open or closed loop the AFR setpoint was always taken from the Octane maps (not sure why I think this - possibly mentioned previously on this forum?). The open loop tables identified by AderC here (http://www.geekmapped.com/forums/showpost.php?p=4703&postcount=68) just dictate if the ecu operates in Open or Closed loop depending on load.

If so then just raising the AFRs in the octane maps should lean you out as desired.

Or am I sorely mistaken :huh2:

Nutter_John
26-11-2010, 07:30 PM
The top map looks correct but the bottom looks like you need to flip the axis over as the 9.8 values are at the bottom middle like the top one

and yes it is just a case of upping the values to lean out the mixture

Patryn999
26-11-2010, 10:16 PM
Phosty, at a guess I would say thats because our peak boost is around that area so the air will be hotter == more fuel to avoid det. But that'd be shear guesswork.

Current opinion seems to be that the bottom configuration is correct (esp. in OzVR4). I had a play when tuning mine and got a result where I expected it to using the swaped values (lower) and no change when altering the map in same load/rpm location when not swapped, but then I popped my intercooler hosing and lost a hose clamp; such was the end of my tuning day.

However I believe Shtiv has done some dyno work with the tables and seen specific single-cell altering results so it'd probably be best to wait for him to reply.

AderC
26-11-2010, 10:50 PM
I thought that regardless of the ecu operating in open or closed loop the AFR setpoint was always taken from the Octane maps (not sure why I think this - possibly mentioned previously on this forum?). The open loop tables identified by AderC here (http://www.geekmapped.com/forums/showpost.php?p=4703&postcount=68) just dictate if the ecu operates in Open or Closed loop depending on load.

I'm not going to take the credit for finding the Open Loop tables, only confirming that they were correct :-)

I've leaned my fuel map out slightly on boost:

Stock:

39347

Leaned:

39346

I daren't go any further as I don't have a wideband, but this makes a very noticeable difference. I've had to tune out knock in some areas to compensate, but I'm not getting more than 1 or 2 knocksum values anywhere.

Paul C
29-11-2010, 03:48 AM
Aderc what does the 240 cell represent in terms of boost?

Nutter_John
29-11-2010, 08:52 AM
roughly 1.4 bar

AderC
29-11-2010, 09:42 AM
roughly 1.4 bar

I haven't actually got the boost set that high - I'm running ~12psi

Nutter_John
29-11-2010, 09:56 AM
I assumed Paul is talking about the load line on your map , which the last one is at 240 , this is roughly 1.4 bar (1 bar normal + 1.4 bar ) .
You may not be running more than 12 psi but your maps having the fuel to 1.4 where as normal maps goto 1.0 bar (200)

Shtiv
29-11-2010, 01:14 PM
So Daniel/Shtiv, should the Hi_octane map look like the lower map in this screenshot?

39344

If so then it makes sense as it generally shows the mixture getting richer as the load increases for each rpm line. But anybody any idea why they lean it off slightly aove 6500rpm?

I thought that regardless of the ecu operating in open or closed loop the AFR setpoint was always taken from the Octane maps (not sure why I think this - possibly mentioned previously on this forum?). The open loop tables identified by AderC here (http://www.geekmapped.com/forums/showpost.php?p=4703&postcount=68) just dictate if the ecu operates in Open or Closed loop depending on load.

If so then just raising the AFRs in the octane maps should lean you out as desired.

Or am I sorely mistaken :huh2:

The bottom map is correct, trust me I tested it to death, most people have been using the axis incorrectly and taking it as gospel. You need a statement saying flipxy = true.

And if you think i am wrong ask yourself these questions about the top map.

Why does the map go to 14.1 at low load when it runs closed loop up to 4000RPM (again at varying loads)?

Why does the map start enriching at 4000RPM when this is still closed loop?

And if you still do not believe me, try enriching all of the map above say 80% load and then change that to being rich above 4000RPM instead and see for yourself which map works.

phosty
29-11-2010, 03:36 PM
Thanks for confirming that Shtiv - I just wanted to be sure. I also crossed checked against a few evo maps and they are the same too with flipped xy. Maybe this explains why Benh was having problems when tuning the first few 7202s - I recall him mentioning that when they changed values the effects happened elsewhere?

Thinking back to what I suggested earlier about the ecu using the Octane maps as setpoints for closed loop - how can this be correct, because the stock lambda is only narrow band and can only detect stoich? It may use the Octane maps for a first guess, but then use the trims to actually home in on the correct mixture for 14.7 AFR.

AderC - when you describe a very noticeable difference, what do mean? Improved economy? Or more power? (or both)

On a related fueling note, has anybody confirmed the correct address for Injector Scaling - is it 10306 (from Merlins guide) or 10318 or something else? I seem to get no effect changing the value at 10318.

Paul C
29-11-2010, 10:56 PM
I assumed Paul is talking about the load line on your map , which the last one is at 240 , this is roughly 1.4 bar (1 bar normal + 1.4 bar ) .
You may not be running more than 12 psi but your maps having the fuel to 1.4 where as normal maps goto 1.0 bar (200)

Correct, things have progressed alot since i was last on here regarding the standard ecu Map. I was interested to see what it topped out at.

Shtiv
29-11-2010, 11:31 PM
I haven't tried injector scaling but I had 10318 and at the risk of sounding whingy I'm gonna ask again, does anyone have lean spool tables that work?

lateshow
04-12-2010, 10:02 AM
So if I got this right: When I get the ROM from my ECU to ecuflash it shows like the original upper photo posted. And that gives me false information because the axis should be flipped? We have been altering the tables a bit by leaning the mixture on the lower right corner to somewhere around 11.0-11.5. So how should one operate when trying to alter the mixture?

Should I somehow flip xy when starting to alter the table and the flip xy again before flashing it back to ECU? And how does one flip xy?

Shtiv
04-12-2010, 11:11 AM
In the xml file you are using, the fuel map code should say as below.

If you downloaded of your car and its the other way round it means your xml file is wrong. The map itself is not wrong you are just reading it with the scales in the wrong spots so if you adjust high RPM values you would in reality be adjusting high load values and vice versa. You'll find that the diagonal line from the top left to the bottom right is still correct as flipping the axis makes no difference but if you made all values above 100% load rich as hell (say 9:1) and then measure it, you'd find that you have actually modified all load points above 5000RPM irrespective of load.

Also, you may know this, but to be sure, the values are target values, you then need to measure them after you have changed them to see what they actually are so an 11:1 target may give an actual AFR or 11.8 or something like that....

That all make sense?


<table name="Fuel Map 1 - HI OCTANE #1" category="Fuel and Ignition Maps" address="110a3" type="3D" swapxy="true" scaling="AFR">
<table name="Load" address="2cf54" type="X Axis" elements="15" scaling="Load"/>
<table name="RPM" address="2cf30" type="Y Axis" elements="15" scaling="RPM"/>
</table>

<table name="Fuel Map 2 - HI OCTANE #2" category="Fuel and Ignition Maps" address="1118b" type="3D" swapxy="true" scaling="AFR">
<table name="Load" address="2cf54" type="X Axis" elements="15" scaling="Load"/>
<table name="RPM" address="2cf30" type="Y Axis" elements="15" scaling="RPM"/>
</table>

<table name="Fuel Map 3 - LO OCTANE #1" category="Fuel and Ignition Maps" address="11273" type="3D" swapxy="true" scaling="AFR">
<table name="Load" address="2d1ce" type="X Axis" elements="15" scaling="Load"/>
<table name="RPM" address="2cf30" type="Y Axis" elements="15" scaling="RPM"/>
</table>

<table name="Fuel Map 4 - LO OCTANE #2" category="Fuel and Ignition Maps" address="1135b" type="3D" swapxy="true" scaling="AFR">
<table name="Load" address="2d1ce" type="X Axis" elements="15" scaling="Load"/>
<table name="RPM" address="2cf30" type="Y Axis" elements="15" scaling="RPM"/>
</table>

lateshow
05-12-2010, 11:04 AM
Ok, thanks got that. I didn't have that swapxy part in my XML, I can write that there but is there available a newer, better XML file?

I have a wideband O2-sensor, it says I'm doing about 12.4 full throttle and high rpm and the target value is 11.5 (lean spool propably) My fuel map is done so that the lower right corner is about in the same value. Should I try to get the richest part in the same place where it is in the original one?

Shtiv
07-12-2010, 07:33 AM
Target values are exactly that, targets. If you want to see what a load cell is doing without any lean spool influence you need to get to that load cell and hold it there or keep the acceleration slower than (from memory) 1 sec per 500RPM.

As for where you need to get rich and lean parts that's a whole story on how to tune a car. There is a good starting point thread on evolution m. search for "guide to tuning an evo" or something like that.

We are all trying to get a common xml file happening but I am flat out and haven't learned how to use bitbucket and everyone else seems busy as well

lateshow
09-12-2010, 10:23 PM
Now the AFR values seem correct, time to make some adjustments. Everyone here has been using that table the wrong way.

Shtiv
11-12-2010, 12:12 PM
I kept saying it but nobody listened....

lateshow
06-01-2011, 01:02 PM
What do you think of my fuel and ignition maps? Using 99 octane now and doesn't knock. Is my fuel se too lean, wideband says low 12 when full throttle (0,9 bar) and usually about 13 when accelrating with medium boost (0.5-0,7)

What about my ignition? I advanced it on lower load and had to do the other way on bigger loads.

A link to my rom
http://users.tkk.fi/~lylirost/vr4/lauri%20modi2.bin

kinkyafro
06-01-2011, 09:01 PM
Thats leaner then I'd choose to go how are you testing for knock - Evoscan and a hard pull in 3rd?

lateshow
06-01-2011, 09:46 PM
Yes, and then with "knock-detecting" earphones or how does one say that? I'll be changing to richer fuel now. What about this one? http://users.tkk.fi/~lylirost/vr4/lauri%20modi3.bin

Anderz
02-02-2011, 06:27 PM
Just got a 7202 ECU for my car:) It is from an automatic (EM2005, 20030011) but my car is manual. Works just fine but CEL turns on after abouth 30 sec or running, probably beacuse it can't contact the AT-ECU. Anyone willing to share a Manual stock rom for that ECU, flashing that to the Ecu will probably solve the CEL problem.

Shtiv
08-02-2011, 10:32 AM
PM me you email addy and I'll send you a stock manual ROM

Anderz
08-02-2011, 01:23 PM
Email address sent
Thanks for the help