self tuning efi

what is there to keep an efi system from tuning it self. I mean like a turbo motor running with efi that would see a cyl getting leaner than the others ( by reading the egts )and adding fuel to that cyl. and even adjusting timing using an adjustable knock sensor. All done with software, right?

 

Is this BS3? I didn't know any system made closed loop changes based on EGT.

 

http://www.snydermotorsports.com/ComSync/home.htm

 

The aftermarket EFI systems I've worked with a couple things come to mind. I don't think you're talking about the auto tune features - those you can turn off if that's what you want. I think you're talking about closed or open loop operation. If I have that correct you could limit the correction factors to minimal in closed loop operation, or you could run the system in open loop - just have to be sure your tables are what you need, because at that point the technology you paid for is no longer working for you.

 

I would use Holley or Halltech. You still need to fine tune the timing map and fuel. But if you can run it on a dyno that will help get a great baseline.

 

Still need to read plugs even with all the gizmo's !!!!!

 

The self learning is more for street car applications that utilize smaller camshafts. Basically you input a desired A/F ratio and when the system enters closed loop it will try to adjust the fuel map to target the desired the A/F ratio. It really sucks for idle tuning and is not that great overall for max effort type race engines. Once it creates a new fuel map in the learn table you can transfer it to the base table. I do not use the function for 2 other reasons.
It simply tunes off of the wideband O2 sensor. If the sensor fails you are screwed. You can set it to only adjust up and down fuel at a certain percentage but will constantly try to bring the tune out of range. Once we got the base fuel table down on the dyno I just use the closed loop compensation set at 10% correction. That means it can only correct up to 10% either way from the base fuel table. I can use that information to go in and make small adjustments to my tune. Currently the system only makes 3% changes at most and is very consistent. If the O2 sensor ever fails the system will revert to base fuel table and will still run safely.

 

Haha that would take the fun out of it!

Obviously I did misread the original post and thought he was asking about self learn.

OP if you run any EFI system in open loop it will run solely off of the base fuel map plus any modifiers that are programmed. The O2 sensor will not have any effect on the fuel delivery.
If you have a really good map and don't need to account for weather changes etc you can certainly run the system in open lop.

 

,

Yup, never run a competition (drag) EFI motor with self learn capability in a short duration competition. Use it with limited authority as part of the initial tune and R&D (If you must....) and then turn it off. For race you want a predictable foundation to work from. Any sort of learning algorithm is too slow, and EGT's are way too slow other than to tell you which pistons you will have to replace post run. You want repeatability, reliability, and reproducibility at all costs. Make a run, log your data, and adjust your base map/parameters to suit, if you have to. Make a weather factor global table to adjust for weather (if not already compensated for by your sensors and algorithm). There are self learn systems out there for sure, but they are all targeted either at: 1)street use, or 2)offshore marine where the engine hangs at WOT for extended periods of time.

One more thing. The base hardware selection and planning is the key underpinning of the whole deal. You have to get the base hardware right, in order to be able to later control it, irrespective of who's ECU you use. The underlying hardware design and theory has to first be correct. That is the most common shortcoming (mistake) that I see in EFI systems that I have been asked to assess.

 

I was thinking more along the idea that if I miss the tuneup that it would keep me from getting to trouble. Maybe it could learn but also that it would self-adjust rich/lean. I guess that if the egts are that slow then it wouldn't work. In pulling we may run wot twice as long or more than drag racing

 

Best thing to do is get it running and go do some dyno pulls and you can get some good data. If you try it at the track you are going to waste a lot of time. Its worth 500.00 bucks or pay somebody 2500-5000 at the track.

 

The nice thing about a EFI system is that through the use of wideband O2 sensor you can see how effectively your engine is actually burning fuel. The fuel portion of a EFI system uses a few things to decide how long to leave a injector on . The sensors tell the ECU the RPM , manifold air pressure ,and it corresponds to a set value in the VE table ( volumetric efficiency table ) and air fuel table . Those numbers can fluctuate with conditions and during closed loop operation ( wideband is correcting to target desired AFR ) , the ECU will adjust fuel to accordingly .During a datalog you can see where the ECU was making changes and adjust your VE table to decrease the correction .

 

Excellent post ....

 

FT500 corrects, amongst about 100 other things. This technology has been out for a year or two in Pro Modified, I am trying to get it into Top Alcohol. I encourage you to go to Fuel Tech's website and download/play with their software. If you'd like a junk file to manipulate (fuel tables, closed loop, o2, etc) please let me know and either myself or Anderson (Fuel Tech CEO) can get something to you. One of the neat new features in the newest Beta software is "tuner" passwords. In essence, if you're an ignition expert but I don't want you seeing what I'm doing in the fuel system or other areas, I could write protect those areas and you could only look at what I want you to see. What a time to be alive, folks!

Brandon Booher
NHRA TAD #34
Torque MGMT - Owner
World's Quickest and Fastest Converter Driven Vehicle
3rd Fastest Blown Alcohol Car in NHRA History

 

Agree 100%......

 

Hello, I am the creator of the Comsync EFI system. I saw this thread and thought I would chime in. I too would like to see Top Alcohol go to EFI and I have been working to develop a system capable of handling the needs of a Blown Top Alcohol engine. I would like to promote my system but, it is more important that if someone goes to EFI they do it correctly and not implement the system incorrectly. From the above mention. Yes, I use the EGT's as a safety to keep from damaging the engine. Comsync also has 8 wideband O2s that you can use to fine tune each cylinder with limits you program. I have been doing fuel injection since 1984 and started an engine in 1987. The Comsync box is my 5th generation EFI system. It can run Coil packs on a blown engine up to at least 12,000RPM maybe up to 14,000RPM. I have tested to 16,000RPM on the bench but, that doesn't mean much until proven. 10,500RPM is the highest I have had a blown engine on coil packs. The system also allows for a separate smaller fuel table for the hat with 2, 4 or 8 injectors up top. Besides that it can support 4 injectors per cylinder but, with the larger fuel injectors our now, I would suggest 2 per cylinder as this will give much better performance off the line. The system has individual cylinder control for both Fuel and Ignition. I did not incorporate a self learning function because, I believe that you need to know what is going on in the engine. You can do a VE table or you can use pounds per hour to set up the tables. Ryan Micke and myself can help with the fuel tables and dialing in the motor. Anyway, enough with the sales pitch. I am hear to answer any questions you might have.
Blake

 

Hi Blake,
Welcome to the site. My question is what kind of Widebands O2 sensors are using? I thought they weren't capable of handling alky very well in a blown alcohol race application?
Thanks, Joe