Quick EAE question...

[Peter Florance]

^^^ Your numbers look good to me. It's interesting how non linear the voltage correction time is. It would be nice to accurately model an injector with MS.

What if I ran my car with variable power supply and did it by AFR?

[Black99rt]

If my math makes sense, it leads to the question of how the .2ms/V in the manual was found, are the corrections that much different for highZ injectors? Are there other factors not clear on the surface of ms hardware that require compensation through this parameter?

[Black99rt]

What if I ran my car with variable power supply and did it by AFR?

Interesting Idea, you could do it quick and dirty, check 12.2 13.2 and 14.2volts inputted raw and switched, with a flat VE section at idle and MAP multiply off, tune the correction to suit. You would need to have your opening time correct for 13.2v for this to work though, so you need a scope, and if you have a scope and a variable power supply, why not just do it on the bench? Unless your injectors have been scoped, and now you just want to tune the correction.

A related topic on increasing VBatt accuracy can be found here: http://www.msefi.com/viewtopic.php?f=64&t=33807

My voltage is now reading accurately. It's within .05 volts of actual voltage at my relay board.

The trick was making R3 exactly 10k ohm's, and R6 50K ohms. I used R52 (100k trim pot) for R6.

I'am trying to accurately model my injector characteristics. This is one more variable to remove from the list of problems.

Thanks for the help!

[KaPower]

I have a B&M power module that supplies a constant 17v (I may drop the voltage with a resistor). I will most likely hook it up until MS can model my 740cc high Z injectors. I'll have to see if it can supply a constant 17v during cranking!

I've bench tested (on the car! ;) and scoped my injectors, coming up with an actual opening time seems trivial. The numbers on the scope & bench do not match what I found switching between 4squirts and 2 squirts.

I used a battery charger and the injector test mode found in Megatune to determine opening times at different voltages.

[Peter Florance]

What if I ran my car with variable power supply and did it by AFR?

Interesting Idea, you could do it quick and dirty, check 12.2 13.2 and 14.2volts inputted raw and switched, with a flat VE section at idle and MAP multiply off, tune the correction to suit. You would need to have your opening time correct for 13.2v for this to work though, so you need a scope, and if you have a scope and a variable power supply, why not just do it on the bench? Unless your injectors have been scoped, and now you just want to tune the correction.

I think it has to happen on the car. I tried using the numbers from the accelerometer jig and they weren't helpful at all.

[Peter Florance]

Just on a thought, I tested from 1 sqt sim to 3 sqrt sim and came up with ~ .70 msecond
I was running .840 which seems to long.
This are small (220 cc/min) low impeadance injectors without ballast resistor

[Peter Florance]

raised from the dead
How's this look?

I pushed up AWC and SWC about 10% from 50 to 60 kpa to see if I could improve tip-in on stabs. (edit - I multiplied the existing entries by 1.1, I didn't add 10% to each value).

AFR's seem better but I think I feel the double TPSdot reading.
I can post datalog if there's any interest

eae_stab_2_12_09.png

edit: I think either I have a problem with the TPS or with the way the TPS is sampled

I'll do some datalogs parked of throttle stabs and see what I can come up with.

[Peter Florance]

I wonder if the TPSdot reading is a anomaly of lag compensation?

Even with that, the throttle response is impressive.

One feature that would be cool is to be able to bundle all the EAE coefficients in one file and save them (ala VEX files) so I could try different curves and still revert back easily. Hearing me, tuning software guys?

[Black99rt]

That is an excellent suggestion, and hopefully wouldn't be too hard to implement. I'm looking forward to tuning EAE, I'm just using the defualts for now so I can get a bit more of the aggressive areas of the map tuned.

[Peter Florance]

I've been working on rpm corrections. Also bumped up clt corrections just a little (doesn't seem to need much; I suspect a lot of engines might run with EAE CLT correction curves flat).

Here's the drive home from work if anyone is interested.
Note in MegalogViewer I created a custom field called AFR Target which I defined as [AFR Target 1]/100 and set the min and max of that field to the range of my wideband. That way I could overlay them and see how AFR was compared to the target.

I'm really happy with how it runs though.
http://www.firstfives.org/peter/ms/mega ... 162108.msq
http://www.firstfives.org/peter/ms/data ... 162059.xls

[Peter Florance]

It was 30F (-1C) this morning. With the above settings, I was able to start the car, immediately put it in gear and drive away smoothly. I had my MS1Extra installation able to do that for a short period of time but with this code I had to wait 10-15 seconds.
This is 20W50 Castrol GTX oil (pours like grease at that temperature)

This is really great code!

Thanks, (in no particular order) Ken and James !!

[Zaphod]

I've been working on rpm corrections. Also bumped up clt corrections just a little (doesn't seem to need much; I suspect a lot of engines might run with EAE CLT correction curves flat).

Here's the drive home from work if anyone is interested.
Note in MegalogViewer I created a custom field called AFR Target which I defined as [AFR Target 1]/100 and set the min and max of that field to the range of my wideband. That way I could overlay them and see how AFR was compared to the target.

I'm really happy with how it runs though.
http://www.firstfives.org/peter/ms/mega ... 162108.msq
http://www.firstfives.org/peter/ms/data ... 162059.xls

As I was also looking into tuning my AE or EAE I had a look at my last logs with Megalogviewer and also created this custom field [AFR Target 1]/100 - but this doesn't really seem to correlate with my afr table - The values in my afr taböe go down to 11.9 where the newly created field only shows down to 14.0... any idea?

[Peter Florance]

As I was also looking into tuning my AE or EAE I had a look at my last logs with Megalogviewer and also created this custom field [AFR Target 1]/100 - but this doesn't really seem to correlate with my afr table - The values in my afr taböe go down to 11.9 where the newly created field only shows down to 14.0... any idea?

Did you define min and max for your new field?

[Keithg]

I am trying to get my EAE curves set, finally. I am pretty reasonable in the idle to 100kpa range. I have a theoretical question, though. Under boost, what should the curves look like? I am running the sequential code and am seeing huge PW transitions at a certain point under accel (140-160 kpa) where it just drowns the motor. I assume this means my curves are not good in that range, but I do not know what they should look like much of the stuff here is with NA engines. I do know that with the latest (sequential) code, my standard AE is way screwed up as it stalls the car.

Any comments appreciated.

KeithG

[Peter Florance]

I haven't thought about boost but why couldn't the curve just keep in the same general shape above 100kpa?

[muythaibxr]

I agree with Peter, I don't see why it would change shape.

[Keithg]

I agree with Peter, I don't see why it would change shape.

OK, I'll try that, then. I think I'll need higher values, though. I can edit them in the ini, right?

KeithG

[Keithg]

Ok, so now I have more normal EAE curves, but I am getting results which make no sense to me.

The MSQ and the snapshot of the log is here:
http://www.msextra.com/viewtopic.php?f= ... 81#p206681

the log is here:
http://ys3al35l.googlepages.com/datalog200904231933.xls

Note at the clip I sent that the wall fuel rises rapidly and I max out my injectors! second and 3rd show this at WOT, but not in 4th gear at a similar map point?!? The point that this happens is when wall fuel rises. You feel it as though you hit a wall in accel.

Any pointers appreciated.

KeithG

[Peter Florance]

I also posted this in the other thread

My guess is the trend of AWC and SWC curve to flatten out continues even more than you (or maybe I) think (it's probably a geometric relationship rather than arithmetic).

I think I would reduce both above boost quite a bit. You can tell by the wall fuel that it thinks a lot of fuel is stored on the walls under boost; which it isn't based on your AFR's

From what I can tell, it seems to be working correctly with the curves it has and is probably not anything with Jean's code.

[Peter Florance]

I thought EAE was:

a) screwing with my idle control

b) keeping my city MPG down to about 17 mpg


Turns out neither was true and turned it back on. It was nice to autocross again with EAE

http://www.vimeo.com/5234833

For your amusement and critique