Need fuel management advice: megasquirt

[bruce]

Originally posted by mycarsucks
You are fast again bruce
This is from the MS Features page"

I assume that the acceleration enrichment would be equivalent to PE. I could be mistaken

AE accleration enrichment..
PE power enrichment.

AE is a temporay condition to supply fuel for accleration.
Power enrichment allows for changing AFRs due to TPS.

It's IMO, and seemly GM's, to be able to cruise in a near stoich condition, and then change to a power mode without having to make a drastic change in MAP to get the fuel enrichment for real power. This is especially beneficial, again in MO, with a turbo application. Since you have have entirely different fuel needs at at given air flow, ie N/A vs Boosted.

Now if you look at like the Syclone code, they have the VE table for stoich, then a commanded PE AFR based on the VE table, and then a boost multiplier. A very compact, effiecent way of handling fuel. Toss in a AFR vs MAT, and you have about the best there is. The qualifier to about the best is how much resolution you devote to the about corrections. The Sy stuff is close, IMO, just needed a bit more resolution, again IMO.

Now remembering that GM does routines in loops, they can really do a good job with them ole slow processors.

 

[mycarsucks]

Really appreciating the info bruce.

Ok, some interesting things here. From what I read, MS doesn't use the MAT sensor after initial warmup, However, they specifically state that you need to use a fast MAT because of the turbo boost temp increase. I must be reading something wrong here.

This is what I get for their method of boost enrichment.

Generally, VE table numbers above 100% are used only to richen mixtures. Even a turbocharged engine capable of 20 lbs/in of boost will generally not have extremely large VE numbers. The addition of fuel for boost comes through the MAP term in the fuel equation:
PW = REQ_FUEL * VE * MAP * E + accel + Injector_open_time

req fuel is described as:

Required Fuel – (Req_Fuel) The injector pulse width, in milliseconds, required to supply the fuel for a single injection event at stoichiometric combustion, 100% volumetric efficiency and standard temperature.

How do you think this setup compares to the GM setup? AE is dependant on tps and not MAP to the best of my knowledge. Would the combination of AE based on TPS for quick enrich, and boost enrichment be acceptable?


I also found some more stuff for the rest of you guys out there to look over. First if a car craft article about MS.
http://www.carcraft.com/techarticles/116_0403_mega/index.html

Next is the Cool logging/tuning software info and screenshots that I was talking about. (keep in mind this is one many options and is BETA)
http://bellsouthpwp2.net/s/_/s_moseley/megasquirt/msgui/index.html

 

[jastrckl]

MCS: seems like w/ the MS you could use Megajolt Lite Jr (MJLJ) or megasquirt 'n edis to work with some ford ignition crud to get spark control.

http://picasso.org/mjlj/
"Megajolt Lite Jr (MJLJ) is an experimental ignition controller designed to control a Ford EDIS 4, 6 or 8 distributorless ignition module. It can run stand alone, or in conjunction with a fuel injection computer such as Megasquirt EFI." Looking at this picture, you'd be replacing the EEC-IV ecu w/ the MS/MJLJ. http://picasso.org/mjlj/?q=node/view/2


http://www.jsm-net.demon.co.uk/megasquirtnedis/
seems to work the same way. It needs this kind of crank sensor to work: http://www.jsm-net.demon.co.uk/megasquirtnedis/finding-nam.html
here's a picture of it all thrown together.
http://www.jsm-net.demon.co.uk/megasquirtnedis/whatyouneed.html

-Josh S

 

[bruce]

Originally posted by mycarsucks
Really appreciating the info bruce.
Ok, some interesting things here. From what I read, MS doesn't use the MAT sensor after initial warmup, However, they specifically state that you need to use a fast MAT because of the turbo boost temp increase. I must be reading something wrong here.
This is what I get for their method of boost enrichment.
req fuel is described as:
How do you think this setup compares to the GM setup? AE is dependant on tps and not MAP to the best of my knowledge. Would the combination of AE based on TPS for quick enrich, and boost enrichment be acceptable?
I also found some more stuff for the rest of you guys out there to look over. First if a car craft article about MS.

The Packhard Catalog gives all the response times of the various MAT, CTS, HAC, and Tranny sensors. They're all about the same with the exception of what seems to be two tranny sensors that never made preduction, or that can't for some reason be ordered by mortals. Any of the exposed sensor MATs seem to be the fastest. If you want to dampen the readings then use a CTS, or change the sampling rate in the code. IMO, the stock code/open sensor MATs are just fine.

Not referencing AE to MAP is ignoring an important, again IMO tuning component. In the stock GN code it's turned off, but the reaction seems to be to changing converters. Lots of inertia on that topic. Anyway, enabling it made my car nicer to drive.

VE is such a misused term it's hard to imagine which way they mean. A turbo motor thou, can reach TRUE VE's of 125% to 140%, according to Vizard, and Lockheed.

I explained what I thought was the best fueling strategy when I talked about the Syclone, earlier. Folks seem to get hung up on terms, and AFRs and lose sight that it's about resolution and giving the motor what it wants that makes em fast.

Oh, and not to mention the wastegate PW cut based on knock the GM code has. I've not played with it, but that seems worth while. And the Time in PE fuel enrichment is something few folks use that is really kinda trick in my book.

 

[bruce]

Originally posted by jastrckl
MCS: seems like w/ the MS you could use Megajolt Lite Jr (MJLJ) or megasquirt 'n edis to work with some ford ignition crud to get spark control.

But, for alot less work, thou, it'd cost more, the eDist and CNP set up is still the best set up I know of and is alot less work.
Or changing the ignition over to the FAST START, that GM started using with the 3x18 shuttered crank sensors.

The interesting thing here is that with the FAST START you can use a later PCM, and do tranny management, ie run a 4L80E or 60E if you wanted. That's what I'm working towards. The PCMs with Torque management, and full tranny function management are the unltimate answer, IMO.

 

[jastrckl]

the LS1 CNP is definitely trick bruce, but it's bank too. he wanted something cheap. I've been eyeballing your EDIST + LS1 coils for a little while now. is it possible to fire those smart coils off in waste spark from the stock module or do they require CDI or something else that's more fancy than stock?

 

[mycarsucks]

That is true, I am looking for inexpensive. I would really like to try this out on my elcamino first, It's got no ignition issues to deal with. But it's starting to sound like ignition costs for the GN might limit or prohibit this system for it.

 

[bruce]

Originally posted by mycarsucks
That is true, I am looking for inexpensive. I would really like to try this out on my elcamino first, It's got no ignition issues to deal with. But it's starting to sound like ignition costs for the GN might limit or prohibit this system for it.

Part of going with the 3x18 shutter wheels is then going with the DIS firing strategy but using the CNPs. That would simply the wiring alot, compared to the eDis, and be cheaper. The LS1 coils come up resonably cheap from time to time on eBay.

 

[jastrckl]

bruce, that sounds like what I want to do. can you elaborate?

 

[bruce]

Originally posted by jastrckl
bruce, that sounds like what I want to do. can you elaborate?

Not any further then the concept.
I've got about 3 dozen irons in the fire right now, but, hopefully in the not too distant future, will have more then just a few ideas about it.

 

[mycarsucks]

Further exploration on the megasquirt archives shows alot of non buick people using the stock ECM to handle the ignition side of things, and MS for the fuel side.

For some reason I just don't feel like this should work for our cars, but can't really think of any reason that it wouldn't. Based on this article:

http://www.turbotweak.com/Starting_the_Engine.pdf

It looks like it would at least start. If this would work, I could custom burn my own chips for timing with no loss to me. Although, I don't know if I want to, I could even make one of those boost referenced timing chips.

 

[bruce]

Originally posted by mycarsucks


http://www.turbotweak.com/Starting_the_Engine.pdf

It looks like it would at least start. If this would work, I could custom burn my own chips for timing with no loss to me. Although, I don't know if I want to, I could even make one of those boost referenced timing chips.

While fairly accurate, he doesn't mention there are timers for how long the choke AFR is applied, or the decay rates from choke AFR to Stoich. And there is no mention of the IAC.

Good code needs a certain amount of sophistication to run well.

Some years ago, I just used a 5556, diodes and a couple power MOSFETs, and a push button for the prime, first run logic. The one 555 was the timer for the injectors, and the other as a debouncer.

An ecm bench might be a real good investment. Beig able to run the code on the bench is nothing less then totally neseccary of your going to get serious about this. And before thinking about custom code a good understanding of the GN code will save you untold hours of trying to figure out how to handle routines. Not to mention possibly being a real education in how to run code as a series of interuptions, and loops.

Again, I just mean to helpful in pointing the way. I'm far from being a code guy.