The pic project/ caravan water barrel measureing.

[Syd Rumpo]

Viscosity does affect them, but not very much. Fluid density change is the main issue, but this doesn't vary much for water. They're actually pretty good for what they are.

Cheers

[Andrew Gabriel]

If I needed one now, I'd make it from a raspberry pi. (Could use a PIC or arduino, but I know the Pi better.) Having a network connection would be particularly useful too.

[Simon Brown]

There must be some out there much cheaper than that given the cars that show fuel consumption in real time.

[Fredxxx]

OK I would have thought force on a ball would have been proportional to viscosity, where if nominally it halves between 25 and 50 degsC, I would have thought the indicated flow would also have halved? A density change of 1% would be negligible in comparison.

Where am I going wrong?

[Harry Bloomfield]

Simon Brown laid this down on his screen :

Yes, but they don't use a flow meter, the ECU know the pressure, it knows how long the injector is open for and the assumed flow through the injector. A relatively accurate way when compared with determining flow from the speed of spin of a propeller.

[Dave Liquorice]

I'd look at that as well, temperatures of the flows and returns is easy enough (1 wire temp senors) but the flow rates are a bit harder.

[Andy Burns]

I think it's more complicated than that, it's sold on eBay by Quantum Art Ltd (aka Canvas Discount), then delivered to an Argos store for you to collect

[damduck-egg]

I purchased an add on unit some 38 + years which was a DIY fit into the fuel line of a carburetor equipped car and that was a turbine which sent signals to a small box of electronics and a red led display. It was useless with the display reading from high to low figures so rapidly that no useful information could ascertained. The manufacturer which was actually based in the UK tried to be helpful and took it back and altered the sampling rate of the data from the turbine , it was still useless as the petrol flow varied so much. It needed more refinement and needed to be developed to show an average but was probably before it's time. About the last gadget I bought in my young driver period before I grew out of such things. But another application where turbine flow meters are used with accuracy because the liquid moves within a reasonably close flow speed range and only a total matters may well be in the cellar of the pub down the road. It has long been a common practice of the Pub owning companies to fit such things on all the individual product lines so they can remotely check on what the Landlords are actually selling compared to what they order through the approved supply chain and make sure they aren't buying cheaper stock from wholesalers or another pub with a cheaper deal.

G.Harman

[Harry Bloomfield]

[e-mail] was thinking very hard :

The averaging is still a problem remains a problem today to get right. No one solution will make everyone happy. I think my current car averages it over the last 30 miles and doesn't offer an instantaneous value at all, an instantaneous reading is completely meaningless.

[Adam Aglionby]

Beer measurement with Brulines now Vianet is of questionable accuracy

[Dave Liquorice]

Not completely meaningless, one can see how low you can get it.

[Chris J Dixon]

There is a simple pleasure in going on a long run and getting home with more "miles remaining" than when I started.

Chris