fraction of pulse per gallon

[waterboy 7]

I have a flowmeter that provides 4-20 output for flow rate and an "odometer" for flow total. There is also a pulse output from this device whose spec is given as .26000 pulses per gallon. I will invert that spec to get 3.486153846 gallons per pulse, but why would a manfufacturer specify a value like that? In this context, how would anyone measure a fraction of a pulse?

[Tim Williams 1]

Obviously you can't measure fractional pulses. But if you were measuring a very large flow rate and didn't need precision then 3.5 gals/pulse would make sense. Or if you wanted to know the approximate total but didn't care about fine resolution it would work. If you used a 32-bit counter you could totalize up to 150 million gallons. If you needed to use this device to measure a small volume, then the proportional output and a little calculus would work nicely. Volume=Rate/Time. Edited 31 Oct 2011 by Tim Williams

[waterboy 7]

Turns out this .26 figure converts to a liter per pulse. Dont know why that didnt hit me.

[Tim Williams 1]

But it still doesn't quite add up, if it really is .26 pulses per gallon. You sure the spec isn't .26 gallons per pulse?

[waterboy 7]

Thats whats on the spec sheet. We just hooked it up and we find that it puts out one pulse every 1000 gallons. Thats what I needed to wind up with anyway so I'm good, but the sheet clearly says .26000 pulses per gallon (with the trailing zeros) . That spec sheet must pertain to a different part of the meter. Or was stuck in there as packing material :)

[Tim Williams 1]

Wow, then that's quite the instrument that can measure with a precision of 1 part in 26,000. Too bad the accuracy is off by a factor of around 3846. I've seen some out of whack spec sheets but that is crazy. Anyway I'm glad you got what you needed out of it and I hope you properly dispose of that speculation, I mean specification sheet.

[Ken Moore 23]

No need for high speed inputs, lol.

[waterboy 7]

Though we did spend a while trying to figure out why the pulses were so intermittent. Only got one every 90 seconds or so. I'll bet it would have been fun to watch our faces when we finally figured it out.

[slcman 9]

If the duty cycle is 50% (square wave 50% ON, 50% OFF) you may use rising and falling edge to increment your counter. By this way, you can double your precision.