My understanding of it is that the sensor does report absolute pressure, ergo "(M)anifold (A)bsolute (P)ressure". It uses this, plus TPS and RPM data to calculate mass airflow into the manifold (it also uses the IAT data to add a density factor). Corrections for altitude + weather variations are critical to get it right.
The other way to do this is with a Mass Airflow Sensor, measuring the intake airflow directly.
Gauge pressure is going to be working against whatever the local absolute air pressure is; it can never be higher than this value.
I didn't figure this out until I went to check mine about a week ago (see over in the CA NoX thread). All my data points were consistently low across the entire band, making me suspect a bad sensor. Then I realized that I'm not at sea level, and once I applied a correction for that, it lined up really closely. You can even see the errors in my measurements using a cheap hand pump and eyeballs.
Here's some excellent background on the Delphi sensors originally speced by GM, drawings, error band, and all: http://www.robietherobot.com/Storm/mapsensor.htm
Note the way that they work is by layering a silicon wafer over a near-perfect vacuum, creating a measurable distortion in the wafer that varies with pressure. Because it's over a vacuum, it is, by definition, absolute pressure. (I would also suspect that a common failure mode is loss of that vacuum due to leakage, which would cause the sensor to read lower than actual manifold pressure, likely causing a lean condition due to thinking there's less air mass and thus less oxygen available).
With the data they show for the 1-Bar sensor, and my own measurements, I came up with something like this (this table is hard to line up):
(inHg) (kPa) Measured Delphi Spec
29.53 100.00 4.75
28.76 97.39 4.68
28.35 96.00 4.5
26.87 91.00 4.25
26.26 88.93 4.06
25.40 86.00 4
23.92 81.00 3.75
23.76 80.46 3.72
22.76 77.07 3.6
22.74 77.00 3.5
21.76 73.69 3.36
21.26 72.00 3.25
21.76 73.69 3.36
20.76 70.30 3.19
19.79 67.00 3
19.76 66.92 3.05
18.76 63.53 2.86
18.31 62.00 2.75
17.76 60.14 2.7
17.13 58.00 2.5
16.76 56.76 2.5
15.76 53.37 2.4
15.65 53.00 2.25
14.76 49.98 2.1
14.17 48.00 2
13.76 46.60 1.9
12.76 43.21 1.76
12.70 43.00 1.75
11.76 39.82 1.55
11.52 39.00 1.5
10.76 36.44 1.28
10.04 34.00 1.25
9.76 33.05 1.15
8.76 29.66 1.02
8.56 29.00 1
7.76 26.28 0.85
7.09 24.00 0.75
6.76 22.89 0.742
5.91 20.00 0.5
5.76 19.51 0.42
4.76 16.12 0.3
4.43 15.00 0.25
3.76 12.73 0.1
2.95 10.00 0
It looks like it's pretty much right on it.
(I have some data from a table that XJMike posted in the other thread, but I found errors in the way I integrated it, and need to fix it)
If you're going to do the same with the meter and a vacuum pump, here's how you correct the data to fit it.
First, to correct Kilopascals (kPa) to inches of mercury (inHg), divide kPa by 3.3864. Multiply to go the other way. If you have a hand pump, I would assume it to be calibrated in inHg.
Then, to find the local air pressure, you need to know 2 items:
(1) Local elevation
(2) Local barometric pressure
A simple "rule of thumb" for air pressure corrected for elevation is 1"Hg per 1000' elevation (not perfect, because the change with altitude is not linear, but it's good enough for low elevations).
So, for my elevation of about 1390'. that's a drop of 1.39" over the local barometric pressure. On the day I took these measurements, that was 30.15". Subtract the change for my elevation, and I have a local absolute pressure of 28.76" to work against. Note that this is my highest reading in the above table, at 0" of suction.
So, if I'm starting at 28.76, then 2" of vacuum becomes 26.76", 5" becomes 23.76", 10" becomes 18.76", and so on.
Another example: Let's say I was in Prescott, AZ today. The local elevation is 5050', with a barometric pressure of 30.26". So... 30.26 - 5.05" = 25.21" absolute air pressure. If I were to check a MAP sensor sitting still (0" suction applied), then I should see an output of just under 4V.
Hope that helps, y'all...
