I have the Map setup but the IAT is soo hot right now it will take a few months to get a complete table. Should I just start with 1.00 in this table before before starting to log? I think I should.
Bill
Printable View
I have the Map setup but the IAT is soo hot right now it will take a few months to get a complete table. Should I just start with 1.00 in this table before before starting to log? I think I should.
Bill
Yeah I think so. Then try and log at constant IAT and shape your VE on that. Then you can have that as your stake in the ground. Then you can do some logs with average BEN vs IAT and make your A00014 from that.
At least thats what I tried to do. Seems ok. Still a little tail chasing though.
Talk about chasing your tail. Try using a blower that "modulates" your IAT with every stomp of the pedal. LOL. Constant IAT? I've heard of it.
This whole thing has me a bit flummoxed. In general people comment that things go lean at low IAT's. Fair enough. However the A0014 tables people are coming up with indicate the opposite. OK, maybe something got reversed along the way. Then to add to the fun, DELCO comments that the tables being posted up are inconsistant with his experience, and that VE needs a push at both ends of the temperature range. Well, Dan has tuned more cars than any 10 of us put together, so I take what he says seriously.
It would seem to me, that for a given ECM type, the A0014 table required to get stable VE's over a wide temperature range wouldn't vary much from car to car. If that assumption is correct, then a "canned" setting should get us pretty close to the truth. Perhaps I'm full of hot air. (pun intended)
With my setup, the only hope for a useful temperature compensation would be to use someone elses A0014 table and remap my VE based on that, or plot BEN against temperature at the AIR INLET, because of the wild temperature swings caused by the blower and aluminum manifold, even with a constant ambient temperature.
prior to running a custom OS I see there was no A0014 style table. Ignoring my ignorance for a second is this just a fine tuning table to improve over stock calculations? If so, leaving at 1 should have a similar result to as if I was still using the factory OS?
Very good observation. {Just a little push} The stock OS Fuel Commanded in Open loop is based on ECT and MAP, since the COS goes off of RPM vs. MAP A0014 becomes the ECT mitigating factor.
This is the beauty of the COS...generally speaking once you have gotten up to full operating temp you will stay there and, gain further useful resolution based on RPM vs. MAP commanded fueling.
.. Also stock is closed loop :) Most guys with COS run open loop.
I also suspect the charge blending tables might need a touchup, the A00014 is somewhat easier to massage as its 1 dimensional, anyone had a go at fixing the VE vs IAT issue by messing with charge blending?
ringram, i've got a method where I optimize VE and charge bias at the same time, to yield the smallest errors across the board. it works--the AFR scatter is smaller, idle is smoother, yada yada yada. that's all on stock e40, no COS tricks. it's not perfect, as i'm suspecting the imperfect data for injectors dump different amount of fuel than they need, which makes some errors persist. tuning charge air bias is not the 'end all' solution, but it's a step in a right direction.
Just curious, what does your blend look like?Quote:
Originally Posted by redhardsupra
Mine is about 50~80% below stock:
LABELS Charge Temperature Blending (Factor)
Grams/Second {link: SAE.MAF} Value
0.00 0.145000
10.00 0.126000
20.00 0.112000
30.00 0.102000
40.00 0.095000
50.00 0.090000
60.00 0.086000
70.00 0.082000
80.00 0.078000
90.00 0.074000
100.00 0.070000
110.00 0.066000
120.00 0.062000
130.00 0.058000
140.00 0.054000
150.00 0.050000
This seems to work fairly well with the IAT relocated to capture true initial IAT temps prior to progressing through the intake tract and into the cylinders.
Kewl, do you have a script kiddie spread sheet or something Sir?Quote:
Originally Posted by redhardsupra
no not really, i've barely coming to grips with all the new math and ways of thinking about how to approach such problems. right now i have some rough 'engineering samples' with a lot of manual steps (aka hundred ways to do it wrong).
if you want a spreadsheet with just the bias optimization, i can probably do that. send me an email or something i'll send you a sample so you'll see what data you need to gather and how to go about the process.