[Ether]

Originally Posted by Mark McLeod

P.S. Joe reminded me that the IFI signal voltage was lower, and I'd assume that the signal driver cable would mirror those lower voltages that Joe listed.

I think the measured IFI signal voltage is lower (than the signal from the cRIO) lower because the IFI period is longer. They both have 5v pulses.

If you look at the chart Joe posted, the following formula holds true for all 3 columns:

S = W/L*H

where:

S is the measured (with a voltmeter) output signal voltage

W is the ontime (pulse width)

L is the period length

H = 5v is the height of the pulse (pulse voltage)

the PWM voltage I measure off an old pBasic IFI controller has neutral at .32v, rather than .375, so some variation between controlling devices should be expected.

That seems reasonable. The measured (with a voltmeter) output signal voltage "H" would be affected by tolerances off all three terms on the right, as follows:

∆S = (H/L)∙∆W + (W/L)∙∆H - (H∙W/L²)∙∆L ¹

Since you already have a testbed setup there, if you have the time and inclination an oscilloscope trace would reveal which of the three terms is making the largest contribution to the difference.



¹ This is derived from the above formula S=W/L*H for the algebraic average of the signal. I wonder if all voltmeters respond precisely to this metric. If not, that could be an additional reason why the measured voltage differs.