Initial development was just to get the map sensor working. References on the web reported that the MAP sensor pin outs were A,B,C equals 0v, signal ,5v. Trouble was mine didn't have any legends on it! There was a polarising nodule next to one of the pins, which I took a guess at being the 5v input. I used a bench PSU which allowed the current to be limited to a few milliamps to test this assumption. All was well and the little MAP outputted 4.82V at atmospheric pressure. This dropped to 0.3V close to vacuum (a few seconds on an Edwards vacuum pump! (A few seconds later this would have probably gone to 0V as the guts were sucked into the turbo molecular blades!). At least this behaviour matched one of the profiles that had been posted to the diy-efi list; well it was close enough for me to say it was working!

 Input   model 460   

So I knocked a little 5vPSU made from a 9v PP3 and a 7805. 

I plumbed into the car's inlet manifold (I chose the vacuum pipe which goes to the fuel pressure regulator). The sensor is seen below securely attached to the cruise control unit!

With the original AFM and the new MAP wired into my Pico scope ADC100 oscilloscope and data logger I get off down the road for a few test drives.

A few of the runs are shown below.

One of the problems was the MAP would full range far to quickly, the inlet manifold vacuum would equalise to atmosphere with relatively small throttle openings. The idle signal was also completely opposite to what I would like (high manifold vacuum at idle). The long and short of it was there is no simple relationship even taking into account a rpm signal (and doing something cleaver at idle) and even that wouldn't increase the 'dynamic' range of measurement so I would probably have to have a throttle opening input in their as well. 

Adding a freer flowing exhaust or inlet manifold would require a re-calibration of these relationships as well. All in all too much work for the ultimate in restriction free inlet tract! Time to have a look at mass air flow meters, hot wire or hot film...

Just of of interest if you zoom in on data logged at idle you can view the pulses in the manifold pressure caused by the piston strokes!

Before the MAP was removed. I thought I would have a bit of a speed trial in my car. As the MAP output records the the load of the engine and the data-logger records this with respect to time it became apparent that time 'vs' load plots would clearly record gear changes. Do the gear change on the rev-limiter and you know what speed you are travelling at that point!

Here's a bit of a laugh on a private stretch of duel carriageway near my house: Basically a fast pull off (no wheel spin or clutch destroying antics here!) followed by WOT (wide open throttle) acceleration to the rev limit (7000rpm) in the first three gears. Speeds are calculated from rpm, gear ratio and tyre size, not speedometer reading!

So 0 to 68mph takes my car 8.4 seconds, with a more brutal take off and faster gear change I think that the book figure of 7.2 seconds 0-60 would be obtainable, if not beatable. 7000rpm is probably not the best change point for this experiment but it was fun trying!

The "kit" required for this experiment cost less than 100quid; a data logger from pico technologies and a second hand MAP sensor (assuming you have a laptop!)