MAX9926 VR Conditioning question

[mtechmatt]

Hi guys,

I am having some issues conditioning a crank/cam signal from a Honda S2000 for one of our ECUs.

The Crank is a 24 tooth wheel (evenly spaced), the cam is a 3 tooth wheel (spaced like an even 4 tooth wheel with one of the teeth removed. The Honda sensors are normal VR type sensors that produce an AC waveform as the teeth pass it

We use the MAX9926 VR zero-crossing point detector chip in our ECU for triggering the MCU interrupts. The circuit we use for the chip is as below (1),

Yes, we have tied the -VE channels of the input (+ve on the device) to GND as that is what the factory ECU seems to do, although the + & - lines from the sensor do have their own pins into the ECU, so this could be changed. We did try it just running direct the -/+ from each sensor to the MAX9926 and received very strange interrupts from the chip - non sensible visually.

We get perfect signals which look like (2) (LOGIC PLOTS ARE FROM THE INT ROUTINE OF THE MCU AND JUST SHOW A SPIKE WHEN THE FALLING EDGE OF THE MAX9926 OUTPUTS ARE RECIEVED) up to around 6,000rpm then we lose sync (would be around 2400hz of signal on the crank line), and the logic analyser outputs show (3) (extra and missing teeth).

I plan to scope it today pre, and post MAX9926 to see what is happening and I welcome your input at solving this!

(1) INPUT CIRCUIT


(2) NORMAL TEETH SUB 6,000RPM


(3) SYNC LOSS AT/OVER 6K RPM

[kosmos]

Hello, your post is already very old, but I would still be interested to know if you have found a solution. I have the same problem that from around 2500 Hz the output only shows high. Despite the fact that I give the signal generator a clean sine wave with 20V peak to peak to the MX9926.