So how long would it take for you to get the MAF AirCounts and SD AirCounts thing patched in? I'd be willing to beta test a version of it if it comes out around early March. I plan to finally get my IAT sensor installed by then.

I have fun because I'm using the 2g MAF on a DSM 4g63 engine. I think those two logging parameters would really help me get the tune close, quicker.

It wouldn't be hard to possibly write and Excel spreadsheet or even a executable program that quickly takes a look through the file and give the user good SD settings to try next.

 

I imagine it will be done by the end of Feb. We'll probably send it to folks with SD 1.0 experience to beta test.

 

Yeah. That probably wouldn't be a horrible idea now that I think about it.

 

I don't know how feasible it would be, but it would be pretty convenient if there was some way to filter the bad data out. Kind of how there are various flags to say when the car is in different conditions. Would it be possible to have a flag that just said "hey, all I'm using right now is the main fuel map."

The intention being so that you could log AFRMAP, Actual AFR, and this variable while in open loop and then easily filter out the data where accel, decel, fuel cut, etc. is taking place. After that, it would be very easy to just make actual AFR match predicted AFR, and then the main fuel map becomes a true AFR target map.

 

It would be nice if a flag would trip that we could log, for now all we can do is log:

STFT: When STFT=0, you're in open loop and going off the map.
IPW: When IPW=0, you're in fuel cut.
TPS: You could calculate the delta(TPS) and if it exceeds a certain value, don't use that data.
Speed: If delta(Speed) exceeds a certain value, don't use the data.

Also you could,
Disable Lean spool - you'll run richer, but at least you can dial in your SD.
Set all decel fuel cut tables so they don't do anything.
Set all accel tables so they don't do anything.

If you can hit most of the cells a few times without completely smashing the gas, you could at least get close.

But at the same time, if I'm hitting some of the higher load and higher RPM cells, I'm usually accelerating. Under those conditions, I want to hit a target. You'd just change 1 map instead of many. Not the greatest solution.

The auto Ve tuner for Megasquirt has triggers that signify accel, decel, fuel cut, etc. When it auto tunes the VE tables, it doesn't use values when one of those flags are tripped. (At least on the setup my friend has in his Dakota).

 

I know some of you are awaiting the patch already.
I just wanted to update you with latest status, I have finished the coding almost a week ago, now the implementation/move is on mrfred (updating the logging capability and bigger MAP axis for 4-5 bar sensors).

 

I just noticed this particular line-

I havent noticed this to be an issue, J can you explain maybe in more detail? I set my fuel table at 16psi on any car (MAF or SD) and because of the nature of the calculations it "boost comps" like an AEM until you either run out of turbo or fuel pump.

 

I think he's talking about the point where you run out of turbo.

As boost goes up, your exhaust pressure increases almost exponentially as you reach the max flow of the compressor which drops VE.

 

gotcha...I thought that might be what it was. I ran into that to day actually, I was going to start another post about making map changes when VE changes like that.

 

i love that idea about logging the Maf... anything to make tuning the Ve easier is ideal. Im still having the issue with cruising in 4th gear at 4k or 4500 and my afrs drop into the 12's and hold around there. I have tried tuning the RPM scale and that did pretty much nothing. I was up around 110 thru the 4500 to 5500 range, brought it down to 100, but my problem is that if i actually do a full run, then i start playing with hitting that fuel cut area. I suppose i could change that to a higher value. Its just frustrating watching my fuel gauge when cruising at that point. JB or any of the other SD geniuses have any suggestions ?

Another question or suggestion... are you guys logging kPa or just calculating it from the map sensor PSI reading ?

I added a value into evoscan that i now log that reads kPa from the omni... not sure if anyone else has yet, but more than likely im sure.

 

I like the idea of the 3D VE table just for the fact of where I live has a lot of hills and mountains when driving through towns or on highways. Every time I go out and log my trims and make corrections to the VE table, the trim values will always be different on every log I do depending on where I'm driving and how many hills there are lol.

As an example, I will log a 45 minute commute to work on one day, and make any changes to the VE to try and bring my trims closer to perfect. The next day I will log the changes and they will be spot-on. The day after that I may drive a different route, and now because of the up and down hills and my throttle usage, now my trims are all over trying to compensate. Where I live doesn't have extreme elevations, but it's not uncommon to go from +800ft to +2200ft within 4 miles and then back down to +1200ft and down as far as +500ft and back to +1800ft. A lot of hills and valleys where I live that kills VE lol.

 

Just as an update, retuning the car with baro set to 85kpa took care of a lot of little issues I was seeing. I was at least able to start the car and have it idle on it's own at 8500'. It did idle considerably leaner then normal though.

Is the fuel controlled at idle separately from the normal fuel map or something? It weird that as soon as you get above 1500-2000 RPM, the AFR goes right to how it was tuned 4000' lower but below that point, it wants to run very lean.

Having baro correction for the secondary functions I think will help quite a bit. When implemented it should not affect main fueling though. Driving AFR is great while going up in the mountains.

 

03whitegsr, I've noticed that I can either tune the car to idle extremely well and not drive well in the higher rpm ranges or tune the car to drive well in the higher rpm ranges and idle poorly. For instance my one car idles at 30 load however I can achieve that same 30 load doing a light cruise down a hill in 5th gear but this time it is at 2500rpm's. I think 3d tables may help solve this for you.

EDIT:
When I say poorly, I meant less than desirable. It is still better than a MAF based car.

 

Sooo..... how is this coming along??? I now have SD running on my car and this is pertinent to my interests again.

 

Ok. That works for me. My car is running pretty well with SD v1.0 but I was curious where SD v2.0 was in development. If it's in testing, sweet. The switch to SD has definitely smoothed out the way my car runs.