[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 2-bar MAP sensor




| I am intrigued by your '730 2 bar setup.  Did you simply substitute a 2 bar
| sensor in place of the 1 bar, and then rescale everything that it touched
| (ie: VE, spark, AE, EGR, BARO,  dwell, flow, etc)?  Or, did you create a "1
| bar representation" of the 2 bar signal, store both to RAM,  and
| subsequently use each to drive different functions (ie: use the 2 bar for VE
| and spark lookup, and use the 1 bar to drive everything else)?   Have you
| had good luck so far?
| 
| 
| Walt.


Hi Walt,

yes, the later.

I "intercepted" the A/D read routine and test for the MAP sensor
channel. If it's a MAP sensor read, I read the 2 bar sensor then
convert that into a 1 bar MAP value (i.e., scaling and limiting) and
an additional boost value that I save in memory. This part is working;
however, the slight change in MAP reading seems to have made blm
variations larger. After 2 1/2 months of data collection I was within
+/- 4 blm counts worst case in every blm cell. After changing sensors and
using the "calculated" conversion, I'm at about +/- 12 counts (from a
limited data set). I made a side-by-side measurements last night
comparing the two sensors. The least mean squared relationship of the
form y=mx+b based upon my measurement is Map1= 2.1078 Map2 - 2.3438
(in counts, and from memory as it's at home). The two equations differ
by about 2-3%. I may switch to this equation or I may just continue
with the VE "learning".

My intent is to switch the spark and VE tables when under boost and
use a form of the above boost value as the load parameter. I use a
"fuel management device" (i.e., boost fuel pressure under boost for
those that wonder what an FMD is) so I don't intend to alter the
fueling equations; just to be able to tweak them through the VE table.

I'd put the code on my ftp site now, but it's at home. I will latter.

john gwynne
----------------------------------------------------------------------------
To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
in the body of a message (not the subject) to majordomo@lists.diy-efi.org