MPS mystery, huh Options

[r3dplanet]

Last winter I had some fun rebuilding my MPS unit, which is for a '73 1.7 liter engine. The MPS part number ends with 049.

According to the calibration tests on the pbanders website in conjunction with my trusty Meterman LCR55 and MityVac, the unit is perfectly calibrated. It holds vacuum without issue. The readings at 15, 4, and 0 in.Hg all match perfectly. And I mean perfectly. No variation from what the LCR55 shows to the pbanders table.

The mystery is that when I hook it up to the car the engine won't start. It tries to start and I can sort of force to keep going by blipping the gas, but it won't idle on its own. However, when I install another verified trashed out unit, the car starts right up and idles and climbs smoothly.

Um. Huh. I'm hoping that there's some other adjustment elsewhere in the system that my tired brain won't remember. Somehow maybe my car is adjusted to work around the (leaky) old MPS? Or maybe I didn't rebuild the MPS correctly. But it was in and out of the thing for days and I'm convinced the physical operation is correct. But I'm always willing to be wrong.

Confused.

-m.

[stugray]

Jeff,

I just read the documentation that Paul Anders provides:

http://members.rennlist.com/pbanders/manif...sure_sensor.htm

The device IS a Linear Variable Transformer just as I suspected.
AND you can certainly have AC signals in a DC car.
However Bosch "cheated" and used a pulse as the stimulus signal instead of a sinusoid.

So the device works as I thought. I read the calibration procedure, and while I see how this works most of the time, I can also see how it can fail to work some of the time as well.

So "into a determinable and adjustable electrical signal (inductance)" is not 100% correct as the characteristic of the MPS that changes with pressure is the "transformer ratio" between the primary & secondary.

The "proper" way to calibrate such a device is as I stated above: Inject a known signal into the primary winding while measuring the amplitude of the secondary.

Paul's method obviously works for most of the units, but the variables in the primary (which is not checked other than resistance) could be causing some units to work while others do not.

Stu