Best way or ways?

[Archangel]

Hi Ed,
Point to consider, when that rotating arm generates a vortex around it, what's going to happen to the pressure?
These might be helpful . . .

www.freescale.com/files/sensors/doc/app_note/AN1536.pdf

www.v3instruments.com/trouble_shooting_guides/speedometer_pitot__ts.pdf

[rsocor01]

Point to consider, when that rotating arm generates a vortex around it, what's going to happen to the pressure?

Very good point!!!

I have seen several water pressure sensors for sale in mouser.com. I think that would be the way to go.

Robert

[Ramius]

Thanks Archangel and rsocor01!
I would imagine that if the sensor is about 2-3" in front of the arm that there should not be any vortex at the point of measurement. The biggest question is how to measure travel time.

Best, Ed

[Demon]

Unless you use a large pool, the water will ever so slowly gradually rotate. I would think you get a true reading on 1st rotation and then it would decrease slowly with each rotation.

Robert

[Ramius]

Thank Robert!
I will certainly watch for this.

Best, Ed

[rsocor01]

The biggest question is how to measure travel time.

Best, Ed

What about a water flow sensor?
http://futurlec.com/Flow_Sensor.shtml

[Ramius]

Thanks Rsocor01!
If pressure does not work then this would be an excellent choice.

Ed

[Art]

Usually for counting pulses in time, such as speedometers/tachometers or the like,
you can either go for counting the number of pulses that occur over a constant defined duration,
or timing the space between each individual pulse with an interrupt.

The former is easier to implement from a BASIC point of view,
but when it gets to something like counting the number of pulses from your
transmission in one second, and calculating speed on that,
you end up with a display that can only be updated once per second,
and might find there aren't enough pulses to increase it's resolution.

There are some tricks to improve it such as using multiple intervals that
overlap each other, so a single pulse count might apply to a calculation
for multiple time intervals.

The former method, calculating based on the duration between each pulse
is more prone to error in some situations such as a heart rate monitor
where the hardware might throw in random pulses because our interface
to the Human heart isn't good, and although I haven't used one,
it sounds like a problem you might face with your sensor.

In that case I had the program look at a number of the previous times,
so that it could either average them out, or disqualify silly looking readings.
I was happy with that, although for a speedometer, it would mean that if
you crash it into a wall, the speedo would drop to zero over a number of readings.

I would be worried about the projects influence on the performance of the vessel,
but since you're the one doing it, I'm sure you've taken it into consideration.

[Ramius]

Hi Art!
Thank you and it all makes sense. When reading the pressure sensor there are several "pauses" in the software so that one idea isn't going to work. When the pressure sensor is installed in the sub there should not be any noticeable influence in performance as the sub is 8" in diameter and 6' long! So a 1/4" hole in the bow to read pressure should not be a problem. Thank you for everything you wrote you gave me a lot to think about!

Best, Ed