tuning programmable engine management systems

[WERK I]

Originally Posted by scarceller

Thanks for getting me the # of teeth on the 993 flywheel.

I figured out the timing formula and found the right maps, here they are:



Looks much better, I check my numbers (ballpark) against this site:
http://911chips.com/ignition.htm
And I trust this stite

Sal,
Thanks for your contributions on this great thread. Reading the ignition map above, it seems this map would work very well in a twin-plug turbo engine as well. Other than adding the additional values for boost (positive) MAP, this appears a great starting point.

[911st]

If you want it to feel quicker just above cruse load, would you want to get closer to max TQ settings quicker on the PT map?

[scarceller]

Originally Posted by 911st

If you want it to feel quicker just above cruse load, would you want to get closer to max TQ settings quicker on the PT map?

That's the idea but the problem is that at PT loads close to WOT like 75% load I have found that WOT AFRs are to rich, for example my car produces best power at WOT around 13.0AFR but this mixture at 75% load is simply to rich, the car likes 13.3 or so at 75%. The only way to really tune and know for sure is on a Load Dyno.

In the end the idea is to richen as load increases but you need to know how much to richen.

[911st]

Interesting.

I moved my WOT switch to trigger as soon as it can. Do you see a down side to having done so. I have a SW chip and did not notice any differance to tell the truth.

Almost sounds like with a properly tuned PT map, there dose not seem much need for the WOT map. Having both may be more about allowing the factory to achieve gas and smog needs while still maintaining acces to performance but only under WOT.

[scarceller]

Originally Posted by 911st

Interesting.

I moved my WOT switch to trigger as soon as it can. Do you see a down side to having done so. I have a SW chip and did not notice any differance to tell the truth.

Almost sounds like with a properly tuned PT map, there dose not seem much need for the WOT map. Having both may be more about allowing the factory to achieve gas and smog needs while still maintaining acces to performance but only under WOT.

Very good observation, I have built some chip images that have the PT hi-load maps set like WOT and then simply unplug the WOT switch. This does work but has 2 issues:

1- The WOT switch cuts out the O2 closed loop mode and puts you in open loop, but if you are not using O2 you should be OK.
2- If you look at the RPM points of the PT map you will find there are not as many as in the WOT maps, I had to completely change the RPM points in the PT maps I built some trial 16rpmX12load point maps, in the end way to much work for what you achieve.

Simply I'd stick to the stock system of using 3 sets of maps (idle, PT, WOT)

As for moving the WOT switch to trigger earlier I did this also and can't really see any difference.

[jpnovak]

IN reality each engine is different. You should run leaner at cruise and PT positions on the map. How lean and how much spark advance will need to be tuned by the user (or mechanic). What I did was have a friend drive the car. While cruising on the highway on level ground I moved the fuel trim up and down. When you hear that lean surge and occasional intake pops you know you have gone too far. For me the WBO2 was in the mid- to upper 14 range. I then took note of the MAP reading on level ground at different highway speeds and set these target AFR bins accordingly. Then make a smooth transition down to the closed throttle and WOT bins. The same can be said for spark. I do not currently have integrated spark but that is coming in the next few months.

I agree. I believe that a properly tuned map does not need a separate WOT table. In reality it just ends up being an extra row of bins on top of the map. As long as their is a smooth transition between the PT and WOT areas this should not be a problem.

From a hardware perspective the issue is one of recognizing "load". What does the Motronic use? The Air meter position? In this case, the load sensing will be different than speed density (MAP), MAF or TPS based systems. It will be closer to alpha-n (TPS) based. Modern hardware has much improved since the early 80s when Motronic was first developed. Even the chip technology allows for more sensors to give a more accurate representation of the engine "load".

My engine runs really strong with EFI, better than it did on webers. Now I need dyno time to tweak that last few % of performance.

[sjf911]

IIRC, the Carrera Motronic TPS is not a potentiometer and cannot give relative throttle position therefore the need for a WOT switch to deactivate the closed-loop O2 feedback.

[scarceller]

Jamie,

The 84-89 Carrera does use a AirFlowMeter (not MAF) with IntakeAirTemp sensor to help calculate MassOfAir.

Then the actual Load formula is complex - but the idea (and theoretical formula) is:

Tl (tee sub ell) = Q/(n*Ki)

Where Q = airmass, n = rpm, and Ki = injector constant

Load is the basic injection time w/o any corrections,. which is used to index the corrections.

[dwightp]

Wow, great discussions.

I've got a similar setup as dfink but run open loop everywhere and throttle pos vs. RPM (alpha-n) in lieu of MAP vs RPM for primary fuel and ign maps. Anyone similar care to share your experiences?

Thanx

[JohnJL]

Here is a fuel and an ignition map as a reference point. Twin-plugged NA 3.0 with DR20 cams and carrera intake.

Figures in the ignition map are degrees before TDC. Numbers across the top are RPM.




Figures in the fuel map are miliseconds of injector opening time per rpm with 26lb injectors.

[mppickett]

Last weekend I set up my 930 turbo twin plug Megasquirt II setup advance table with data from the OEM 993 spark profile posted by Scarceller.

Seems to work well. Here's the link to the setup thread and tables.

Best,

[scarceller]

Originally Posted by mppickett

Last weekend I set up my 930 turbo twin plug Megasquirt II setup advance table with data from the OEM 993 spark profile posted by Scarceller.

Seems to work well. Here's the link to the setup thread and tables.

Best,

Mike, do you run a knock sensor? Just keep in mind those 993 timing values are twin plug but NOT turbo conditions. So with higher intake pressure how does it affect flame propagation? Does flame front move quicker in a turbo cyl? If so you may want to back off on those ign values. The best way to know for sure is dyno testing.

Good luck and glad the 993 data helped.

[911st]

Just a thought, the 993 is a twin plug with a compression ratio of about 11.3.

A motor that is about 7/1 compression ratio and pushing one bar boost has an effective compression ratio of almost the same. Plus, a Turbo should be running a much richer mix (12.2/1 ?) to protect against pre ignition and slow down combustion.

I am wondering if they might be close under WOT.

Off boost might be different. Might want to look at what the total advance is at WOT near TQ peak (most air/fuel in and highest combustion pressure) and be sure it is not over about 24 or 25 deg total advance. But double check that is the goal as I am rusty on this.

Still, as Sal says, caution is warranted.

[sjf911]

Mike, the msq file advance table does not seem to match your vertex file. Is that an older file?

[Trog]

I've had some great success recently using the Auto-VE tune feature in the TECgt. First I used the TPS/MAP blend feature to get IOT and TOG setting in the ballpark.



It didn't take much TPS blend to achieve a 14:1 AFR across the RPM range at no load.

The interface is a bit clunky at first, but once you've got it figured out, the AUTO-VE tune feature performs exactly as advertised. It's amazing to watch the application add or delete fuel in the VE table in real time. The performance feedback is instantaneous.

A very nice feature indeed.

[scarceller]

Trog, it's interesting you mention Auto-VE feature. I just finished reading a book "How to Tune and Modify Engine Management Systems" and it states that the very first thing to get setup properly is the VE table. It even goes on to explain that systems with Auto-Tune feature for VE-table should set the target AFR to 12.8 across the board, this way you will always be rich enough even for hi-load conditions. Then with 12.8 as the target (for every RPM/Load combo) simply enable auto-VE and let the system build the VE-Table. Then you will fill in VE cells that are hard to reach or where not obtained by hand. The basic idea is to shoot for a safe rich mixture that will work under most all conditions (12.5 to 12.8) then let auto tune go to work and tune/build the VE table for you. Once the VE-table is built you simply pull fuel out via the fuel trim table for the lighter load areas.

After the VE is built then proceed to other tuning, like starting to tune no-load/lo-load for 14.5AFR and things like this. Also Acceleration enrich should also be done only after the VE table is correctly tuned. The book states to tune the VE-Table first and only when the VE-table is correct should you move to other tuning.

[Trog]

Sal,

Your tuning description is pretty much spot-on with the advise I've received from Electromotive.

One deviation is that I've set my AFR target such that at idle and low loads the Auto-VE makes adjustments to achieve an AFR=14:1. My AFR target is currently configured such that it gradually increases from 14:1 at low loads to 12.6 at full load.

I'm not sure why your "How to Tune...." reference book suggests a flat 12.8:1 across the board when the TECgt Auto-VE tune can accommodate a load-based sliding AFR target and make the necessary adjustments so fast that there's virtually no risk of running lean?

Rgs,

T...

[mppickett]

Sorry. This is now fixed. The upload failed and I didn't catch it. Thanks.

Originally Posted by sjf911

Mike, the msq file advance table does not seem to match your vertex file. Is that an older file?

[mppickett]

Scarceller, No knock sensor but I only altered my idle and partial throttle (up to 0 psi boost) values. Anything above .05 bar runs on my normal boost map (that has been to the dyno). The 993 values at idle and partial throttle are actually more conservative than what I've been running for about 3 years with no knock. I'm running 8.0 to 1 compression before the boost cuts in so no worries there. I'm also running a full bay intercooler and haven't run into any knock problems at the advance levels and plug heat ranges I've been running. By the way, many thanks for finding and uploading the 993 values. You're local to the Providence area, right? We should meet for coffee when the ice thaws (July?).

Originally Posted by scarceller

Mike, do you run a knock sensor? Just keep in mind those 993 timing values are twin plug but NOT turbo conditions. So with higher intake pressure how does it affect flame propagation? Does flame front move quicker in a turbo cyl? If so you may want to back off on those ign values. The best way to know for sure is dyno testing.

Good luck and glad the 993 data helped.

[scarceller]

Originally Posted by mppickett

Scarceller, No knock sensor but I only altered my idle and partial throttle (up to 0 psi boost) values. Anything above .05 bar runs on my normal boost map (that has been to the dyno). The 993 values at idle and partial throttle are actually more conservative than what I've been running for about 3 years with no knock. I'm running 8.0 to 1 compression before the boost cuts in so no worries there. I'm also running a full bay intercooler and haven't run into any knock problems at the advance levels and plug heat ranges I've been running. By the way, many thanks for finding and uploading the 993 values. You're local to the Providence area, right? We should meet for coffee when the ice thaws (July?).

Mike,

Yes, I'm in Portsmouth, RI. Would be nice to chat over coffee sometime this Summer. Glad you found the 993 info useful.

BTW - I really like the look of your car.