Matrix Orbital Corp

Hi All,

Here is a question for the hardware experts.
What is the upper limit is on the RPM sensor (fan header) on an MX2..

How many pulses per second is the sensor capable of reading.

I would like to use one of the fan header to monitor a Gem Sensor FT-110 flow sensor.
http://www.gemssensors.com/SpecTemplate ... roupID=115

As far as I have been able to determine I should be able to connect the flow sensor directly to a fan header (with a current limiting resistor) as the flow sensor appears to use the same Hall Effect sensing method that fans use for rpm sensing (dc pulses).


The only question is the fan header capable of reading really high equivalent rpm readings.
I figure that the RPM sense should be able to read at least 9999 RPMs (333 Hz) but can it read beyond that?

Depending on the model of flow sensor I use the reading can be as high as 1150Hz (pulses per second) which would be equal to a fan running at 34500 RPMs.

I would like to use either the 173933 (1150Hz > 34500 RPMs) or the 173934 (550Hz > 16500 RPMs)

I would appreciate any help

Thanks
Elliyanna

Well in theory, the RPM sensing should work upto 312500Hz In practice the error would be too much up in that range to work (+/- 50%). But down in the range you're talking about, the RPM sensing should be fine. In the range of 1500Hz (90,000RPM) the error would be about 8Hz.

James,

Thank you very much that is whatI wanted to know.

The flow meter is not going to be that accurate and this will give me a good idea of what the flow rate is so able to monitor it for changes and performace effects.

Elliyanna

Assuming that you are writing your own software to monitor this, I do have some suggestions:

If you are using the PWM output on the plug that you are using the RPM sensing on, then only check the RPM about every 2-3 seconds. If you can, avoid using the PWM output on that particular plug. If you don't use the PWM output, you could easily poll the RPM about 1 to 5 times a second.

Average the last 'N' values you get back. This will help eliminate the measurement noise. If you average a large number of previous measurements, then you will have very low noise, but your average won't react to rapid changes very well. I'd try to average around 2 to 4 previous samples and see if that reduces the measurement noise.