tuning programmable engine management systems

[dfink]

OK well I thought I would start an independant thread that has no attachement to anyone person or system to discuss programming engine management systems. There are several users doing this currently with either electromotive systems, stock systems, Haltec or probably even some megasquirt.
Some of the basic issues that have been discussed in other threads are
[EDIT] to prevent the need to read entire thread I will try to update this first post with accepted numbers. These may change as a general consensus is reached. There will always be differences.

idle A/F ITB 13.5-13.8
idle stock 14.2-14.6
cruise 13.5-14.7
cruise A/F ITB 16-17:1 said to be possible. May require 30+* advanced timing. (not confirmed as of 01-24-09)
WOT 12.5 - 13.1

timing idle 0 to 12 btdc (hot-rod would not start at 0 required 12)
cruise 30* to 40*
WOT 30* btdc at peak torque ramping up from 20*. Some say 24* max for twin plug dyno shows even with twin 30* if engine knocks back off timing and recheck. Power loss past 30*

basic a/f ratios
ignition curves
idle
various fuel enrichments, start, warmup, acceleration, air temp, blend

So to kick this off I will start with the A/F ratios. In searching threads back as far as they go the one thing that appears to fairly consistent is the WOT mixture. It appears that the target is 12.5:1 to make power. Some have strayed from this to as high as 13:1 but the majority still put it at 12.5:1 Now that leaves the entire rest of the fuel curve up for debate. I have seen anywhere from 17:1 at cruise to 13.5:1 . The stock system is going to try and get close to 14.7:1 as this is where the most efficient burning takes place. I have found that I get 14.7:1 and it will idle OK but I get a terrible stumble coming off idle. Now that is with individual throttle bodies. The stock system may handle the transistion much better. You will certainly get better mileage and have less emmissions if you can get the ratio up there. The general concensus I have come to is that for throttle bodies anyhow 13.5:1 appears to be about what is needed for a stable idle with no off idle stall. So now the question becomes where do you transition from 13.5 to 12.5. Well I can tell you don't do it right where you cruise or your system will be hunting all over the place. Right now I have 13.5 as a target from 3500rpm 90 kpa down to the minimum rpm and 25kpa. I also have 13.5 for all rpm ranges below 70 kpa. I then transition directy to 12.5 to fill in the upper rpm/ higher kpa area. I don't think the sudden transistion is the way I want it so will be working to smooth that just a bit. I do have wide band with mixture control turned on so if I happen to hit a spot where I have to hold it and that spot is on a transistion then the system really hunts. Anyhow have to go now.

[shbop]

signing-up!

[jpnovak]

I have recently spent a few hours tuning my new 3.2SS. Build spec is 98mm mahle on SC crank (70.4mm), 10:1 CR, Twin plug, DC40 cams on 108 deg lobe, 46mm TWM ITBs with 30#/hr injectors. . I am using Megasquirt II, V2.2 with 2.888 code. My interface is an Asus EEE 901 with TunerStudio MS software. Data analysis is done with Megalog Viewer.

I found that my car idles best with approximately 13.1:1. I have found that the car does not like anything leaner than 14.5:1 at cruise. Otherwise it will have a lean surge and pop through the intake occasionally. I have WOT bins set at 13.0:1 at 95kPa and 12.8:1 at 100 kPa. I have really crisp throttle response and found that going slightly richer might feel a tad stronger but felt sluggish on throttle transition.

Here is my target AFR.



I am completely sold on tunable EFI. Being able to set target values and then achieve these through proper manipulation of fuel levels is very fun and educational. Everyone talks about the improvement going from CIS to carbs. I did that transition and it was impressive. The switch from carbs to EFI was equally enlightening.

There are a few things to look out for when tuning that might cause problems for some. For example, too large of a throttle body in an ITB setup will cause the system to go very lean off-idle due to large amounts of air that dump into the system with small throttle inputs. Swept area is very important here. The cure is abnormally high acceleration enrichment values that cause the system to bog down at higher rpm. My method was to make the low rpm, high MAP values very high in the base map. This acts like enrichment and allows fine-tuning of accel events at higher RPMs.

[dfink]

I updated the first post to include some sort of consensus on settings to avoid having to read the entire post. I will try to update this as things progress.

[Bart_dood]

This is an interesting thread; currently I'm aiming towards building in an "expert" mode on my VFM to MAF system.
The idea being that the user would be able to record areas of low or high A/F ratios and then using my spreadsheets they would re-program the microcontroller in my system with some new curve values, this would tune the A/F ratio in some more. It would just require a PC and a USB cable.

However I am unsure if this would really work at WOT? as far as I know the motronic system goes into "dumb" mode and simply reads in rpms and little else, resorting to its lookup table for values and basically ignores the VFM air volume input, even if it did at 5200rpm the VFM pegs and its flying blind.

[psalt]

Dfink wrote,

The stock system is going to try and get close to 14.7:1 as this is where the most efficient burning takes place.


Hello Dfink,

OEM's target stoich for dithering the engine management system to preserve the life of the catalytic converter and maintain the best balance for lowering the three designated "evil" pollutants. 14.7:1 is not the most efficient burn, it is simply the perfect ratio of reactants (14.64:1 by weight), and has little to do with power or economy in a running engine. When tuning a programmable EFI system without a cat, 14.7:1 is really irrelevent. In test engines, max efficiency is around 1.2 lambda (over 16) with 50 degrees ignition advance. I have been tuning EFI for over 10 years (mostly Haltech) and on single throttle applications, there is usually no problem running over 15:1 at part throttle cruise. You just keep lowering the target voltage for the closed loop control until drivability deteriorates. Sloppy idle control and bucking at lean part throttle cruise may be an ITB trait, but you did not go there for idle or cruise performance. With the Haltech ECU's, you have an acceleration enrichment amount and a sustain (time) and it is usually no problem to fix a transition by adding fuel or extending the time. The fuel cut off on closed throttle makes a big mpg difference, but it takes a while to find the best compromise on settings. If you want OEM drivability and not spilling fuel like a drunken sailor, it takes a lot of effort to get it right under all conditions. Hemi head engines (like 911's) with large chambers do not like idling at 14.7:1 and if you are not running a cat, it is an unrealistic goal. On cold starts, you only get one a day and everyday and every engine seems to be different to me.

[safe]

I have a feeling this thread is gonna be gold!
I'm mapping my engine in a few weeks...

[dfink]

psalt - You are certainly not the first to mention 16-17:1 ratios for cruise so it may be an acceptable range. I found other entries suggesting this is possible. Would like to get some more input before changing the table. I think mine would have trouble at 16:1 any trouble with temps under those conditions or if you just get too lean it doesn't run correctly.
What does a lean condition cause under load. Does it just get hot or does it cause detonation why is it bad to run lean? Why will it kill the engine?

[psalt]

dfink,

You have to separate what happens at light load, part throttle cruise and what happens at WOT under load because they are very different states. One of the hardest thing for a first time tuner to understand is that idle conditions, where most of us started messing with engines, setting timing and twiddling idle screws, is a very small part of the picture and almost meaningless. All you want from an idle is the minimum rpm to prevent stalling. A road engine spends 85% of it's time under 30% load, most of the time is spent accelerating, decelerating and at a steady cruise. AFR's are not static and even the most sophisticated system can only dither the mixture around a target. Detonation only occurs at under high load, usually WOT around peak VE and is meaningless at idle or part throttle cruise. It is a pressure spike from spontaneous combustion of the fuel and cylinder pressure is a function of filling and throttle angle. You are not worried about detonation at 15% throttle. If you are looking for max efficiency, you want to run leaner than 14.7 : 1 with much more timing than 30BTDC under light load. Are you running a cat and do you have closed loop control ?

[dfink]

Well I was hoping to keep this to a general tuning thread. I know every engine combination is going to behave differently. For the sake of arguement let take two engine combinations.

First one stock 9.5:1 with motronic fuel injection. Perhaps first hand information can be provided by scarceller as I know he can proprgam his own stock system.

Second senerio hot-rod 3.4 with DC21 cams 10.5:1 headers full flow exhaust 46mm PMO ITBs, twin plug and full aftermarket engine management

This should about cover anything except turbos as most will fall somewhere in between or both are common.

Mine happens to be in the later category. It has been found by most that in senerio 2 of the hot-rod that an idle mixture leaner than 13.5 tends to cause a stumble coming off idle. There is alot of throttle plate opening at once. It has also been suggested that in the stock system you can go leaner and still have adequate acceleration characteristics. I am open to running leaner at cruise but attempts at idle don't work even when using acceleration enrichments. It just stalls then goes really rich. Cruise is still a mystery to me and is generally not discussed much. Lots of info on idle and WOT.

I have read that not having enough timing advance at idle will cause headers to get really hot. The initial programming of my unit had an idle advance of 12* btdc and I have read where others use that setting. I have also read the stock setting of 0* +- 5* so again conflicting information. Is one good for stock and one good for headers. I agree no knock at idle but in theory you could probably get knock at cruise on single plug when pulling a hill. The twin plug pretty much takes knock out of the equation and it becomes about how the engine runs.

If knock is not the problem with running an engine lean then what is the lean killer everyone talks about. I will assume super high head temps. What does 16 or 17:1 do to head temps during cruise. I don't have a true head temp (spark plug ring) I only have the factory cylinder temp and it does not appear on my system to work well under high load conditions.

[psalt]

Dfink wrote,

If knock is not the problem with running an engine lean then what is the lean killer everyone talks about



I don't know who you are talking to, but if they don't separate cruise from WOT, they are talking nonsense. More lean only means more heat rich of stoich, then temps drop as the AFR goes up. Running lean under high load is a problem and running stoich under load creates the highest temperature. The only issue running lean of stoich during cruise is emissions. If you want to run leaner at cruise, you also have to run more ignition advance or you can get bucking, making you think your too lean.

If you have full control over spark tables, idle timing should be close to TDC. 12 BTDC and 5 BTDC are the result of distributor limitations and the need to meet a curb idle emission test. My advice is to forget AFR and timing at idle, establish a stable idle and work on your tip in stumble. A combination of timing and fuel adjustments should make it disappear.

[dfink]

Only thing I can use for real information myself is an airplane. With an adjustable A/F and cylinder head temps you can really see the head temps go up when running to lean. Of course in an airplane you are under pretty much full load constantly. We then richen the mixture to bring head temps into range. I am not agreeing or disagreeing with the lean cruise just don't know what is safe. Could certainly use a few more opinions. I wish I had a head temp sensor would certainly help the situation. But that leads to another question. If you have spark plug ring head temp sensors or some other type. What is the maximum safe head temp.

[psalt]

Dfink wrote,

Of course in an airplane you are under pretty much full load constantly

Yes, and a 65 mph cruise at 15% throttle in a car is a much lower load. If you are a pilot and understand LOP ( lean of peak) fueling, it should be easy to understand that a road car under lower load can run a lean of stoich mixture.

There are some good articles on fueling on this site:

http://www.avweb.com/news/pelican/pelicans_perch_84_mixture_cht_194816-1.html

which recommends against using CHT valvues for fueling.

There are also good threads on the LM-1 Innovate site about lean cruise fueling using +15 AFR's

[dfink]

Well it has been awhile since I flew but I don't remember it quite like the article. In the plane I was taught in we didn't have head temp or a/f guages only the throttle and a mixture adjustement. We took off full rich at cruise leaned to peak the backed off to the rich side. Certainly didn't go lean of peak. Then went full rich again for landing. Problem is I have no idea what a/f mixture was peak as we didn't have a gauge only the tach. The article you posted suggests going way past lean of peak and eventually the cylinder would cool down. Sorry don't have enough balls to lean out an aircraft engine that far. The first step out is a *****.

Anyhow tried a couple of things last night.
First I set my initial timing to 5 degrees (car would not start cold it was very unhappy even knocking during tring to turn over) didn't try that one long. Went back to 12* and it popped right off.

Second tried leaning out the idle. I currently run around 13.5:1 I tried going to 14.5:1 and the idle speed of the car dropped significantly idle was no longer smooth and stalled on acceleration. I was able to take out some of the stuble with acceleration enrichment but not all and mostly the enrichment just made it go rich after the stall. Went back to 13.5:1 and idle speed picked right up also tried going richer mostly due to other enrichments car would run down to even 10:1 but not very well however better than lean. At least for my car and confirmed by several others the target for at least the ITB setup is 13.4-13.7:1 at idle.
This is also the setting I use for cruise at this point. I need an assistant to slowly lean out the cruise and see where the best is. Certainly don't want to go richer at cruise. Even going to 12.5:1 at cruise will show a significant reduction in speed.

Still pretty convinced that 12.5-13:1 is the number for WOT would really like to have more opinions that just psalt and myself don't know psalt but I am not very authoritative. Untill confirmed by some others I will edit the first post to include the lean values but with a disclaimer to not confirmed.

[psalt]

Dfink wrote,

Problem is I have no idea what a/f mixture was peak

14.7:1 is peak. Lean= higher temp is only true when you are rich of stoich.

Some of your comments do not make sense to me. Are you adjusting injector pulsewidth or a closed loop AFR target ? No system can hold an AFR under cruise, only dither back and forth over a target. You usually have to look at the correction table and dial in the pulsewidth to get the correction factor below 5%. Have you verified with a timing light that what the computer knows as TDC actually is TDC ? The best AFR for WOT under load can only be determined by max output on a dyno or with a stopwatch. Low 13's is usually LBT, anything in the 12's is a safety factor for cooling, nothing below 14.1 actually burns. You seem to be searching for a set of "ideal" specific numbers and it just doesn't work that way in my experience. If you are tuning a road car exempt from emissions, you want a steady idle, good throttle response, a lean cruise with a lot of load based ignition timing (like vacuum advance) and a safe AFR under WOT. The reason full load AFR is discussed more, it that it is really the only time it matters.

[scarceller]

dfink,

As you know I've spent a lot of time tunning stock Motronic 3.2L system. I have solid understanding of how the stock system works and it's maps. I think you can learn a lot by understanding the stock motronic system before you start tuning an aftermarket EFI.

I will tell you first hand that the real trick to power and throttle response is in the timing maps and not so much the fuel maps. Of course the fuel maps also matter but they are 2nd to the timing. You should first decide on your timing map(s) before you touch fuel. A flat timing map will not get the most out of these motors.

The 89 chip is the most aggressive chip Porsche produced for the stock 3.2L and if you inspect the maps you'll learn a lot from it. So here's the 89 timing maps, the stock chip has 3 of these: 1 for idle, 1 for PartThrottle and the last for WOT:



Idle ignition, not really that interesting just note that they set idle at 3deg BTDC or so. The thing to know is you can get idle speed control by simply advancing idle ignition if the RPMs drop below the target (like if AC comes on). You can do the inverse to bring idle down if it goes above target. In my car I have idle at 920RPMs with ign at 5BTDC if idle drops below 850rpm I start to crank timing up to as much as 10BTDC. you get the idea.

Then the WOT ign map: this map is very important and you should start by matching these 89 settings more or less. For starters simply set the WOT ign to about 25BTDC from 1000RPMs all the way to red line, this will give you decent power at WOT. Now once you have your AFR set at WOT to around 12.8AFR then start cranking the WOT ign map closer to 30BTDC, I would not go over 30! these cars have NO knock sensors. But at 30BTDC you should be using 93 octane fuel. For street driving and easy on the engine I'd use 27 or 28BTDC.

Now for the PT ign map: this is the hidden gem of tuning and many tuners simply get this worng! A flat PT ign map will have poor throttle response. Look closely at what the Porsche engineers did here, you will learn a lot. The load on the left 8% to 58% is kind of not right. Really think of the 3% as low-load and the 58% max load but not WOT, so just before the WOT ign map kicks in. So if you look at the last 58% row you find it matches very well with the WOT ign map and this is good to help with transition from PT to WOT. Same holds true for the 8% load row and the idle ign map (not exact but close enough). So the PT ign map starts like the idle map and ends like the WOT map and this makes sense. But look at the middle of the PT ign map, WOW! not linear at all. You can even see it in 3D graph form on the right. So the idea here is that at low load first 3 rows they sort of stay conservative. But in the center rows (16-21%) they really crank the ign timing into the 40BTDC range. Why do they do this? Simple, Throttle Response. This is so when you jab on the gas the engine delivers results. I'll tell you that this part of the map and these ideas is where throttle response comes from, but it's also coupled with Fuel Maps (topic for a future post). In my car this center area has timing in the low 50BTDC range, I don't suggest you do that for starters but I do suggest you try to match what you see in the stock 89 chip graph above.

[psalt]

Looking at the factory fuel and spark maps is a good starting point. It also helps to understand the emission control strategies and not copy them. All the OEM's target the EPA emission and mileage test for obvious reasons and sometimes they get pretty sneaky retarding timing at certain points, even using different maps when the A/C is on. The reason they are using 40 BTDC at 15-20% load is fuel economy at cruise. An engine under low load can handle a leaner mixture, this requires much more ignition advance because of the burn time, and it is not a problem at low load. This was the concept behind vaccum advance, add 5-10 degrees of additional timing under high vaccum, low load cruise and take it away as soon as the throttle is opened. You can also learn a lot by looking at the distributor spec on pre EFI motors, but with Porsche after 1968, timing was determined by emissions more than optimum performance. The idea of setting idle timing to 12 BTDC came from having to live with a 20 degree mechanical advance and wanting 32 BTDC at full advance. The later 5 BTDC setting was all about getting to TDC with vaccum retard to lower HC emissions for the curb idle test. With a programmable system these are not issues.

[TimT]

The AVWeb "Pelicans Perch" series of articles has some real gems. Alot of the information is applicable to the auto world..

The articles describe in lay terms a/f ratio, egt BSFC etc..

This one is good!

http://www.avweb.com/news/pelican/182084-1.html

[scarceller]

Originally Posted by psalt

Looking at the factory fuel and spark maps is a good starting point. It also helps to understand the emission control strategies and not copy them. All the OEM's target the EPA emission and mileage test for obvious reasons and sometimes they get pretty sneaky retarding timing at certain points, even using different maps when the A/C is on. The reason they are using 40 BTDC at 15-20% load is fuel economy at cruise. An engine under low load can handle a leaner mixture, this requires much more ignition advance because of the burn time, and it is not a problem at low load. This was the concept behind vaccum advance, add 5-10 degrees of additional timing under high vaccum, low load cruise and take it away as soon as the throttle is opened. You can also learn a lot by looking at the distributor spec on pre EFI motors, but with Porsche after 1968, timing was determined by emissions more than optimum performance. The idea of setting idle timing to 12 BTDC came from having to live with a 20 degree mechanical advance and wanting 32 BTDC at full advance. The later 5 BTDC setting was all about getting to TDC with vaccum retard to lower HC emissions for the curb idle test. With a programmable system these are not issues.

Paul,

I also have Factory Euro Chip maps, do you think these where tuned for Emissions? Maybe it's a better idea to look at the Euro chip? I have inspect the USA and Euro and yes they are somewhat different but not by much. Even the Euro seems to dig deep for fuel and Ignition when you increase load demand for sharp acceleration. In the end it's just a good idea to understand the stock setup.

Thanks for the feedback.

[psalt]

Sal,

I don't know the Euro Carrera chip map, but if they didn't have a cat and the IM240 test, it would be closer to optimum. I think it had higher compression and a different fuel spec. Many of the US 911 road cars were detuned for low octane fuel after the 60's because it is such a knock sensitive design. Copying the factory timing from these models may not be so good if you have a modified engine, twin plugs or good fuel.

Is that 8%-58% load map, total timing or an offset from idle (static) ? I don't understand a $1336 offset, sounds like it belongs on Martha's tax return. Does it say that timing goes negative (-1.4 and -4.2), ATDC, when suddenly accelerating of idle ? I wonder if this is a "burst Knock" or throttle response issue ? I agree that timing changes can do a lot for drivability. Idle timing should be close to TDC, there is really no need for advance at idle on a properly fueled engine and it is bad for cranking. OEM's are now retarding the idle timing after start up to intentionally overheat the exhaust, in an attempt to get the cat up to temperature to meet the hardest part of the emissions test. Not something to copy. For part throttle cruise on a road car, you are just wasting fuel if you do not lean above 14.7 and have a load based advance above the 40's BTDC. There have been some interesting articles about the choices we have made about emissions and how they have increased overall fuel consumption. With a different emission strategy, fuel consumption could be much lower. I am not familar with ECU's with maps that target AFR's. It has to be irrelevent at cold start and during transitions.