[qldbureau]

Is your replacement another FO or clone thereof or another type and make altogether?

I am not certain about my replacement, though I suspect they are clones. Both were listed as replacement parts for the FO series. My first set arrived in small boxes with no markings, yet my replacement gauge I sourced from another supplier; which I paid marginally more for arrived without its own box. This one appears to be a better build quality inside, particularly with how the magnet tip is placed and exposed.

It will be interesting to see your test results.

[JT118]

Well here are the results;

The FO Rain Gauge is definitely built on the following numbers:

Collector area 5500mm2 or 55cm2;
Each tip of the bucket REPRESENTS 0.3 mm of rain falling into our 55cm2 collector.

To do a check calibration proceed as follows:
Accurately and carefully measure 55ml for a quick check or 110ml for a better check. This must be an exact measure using a laboratory measure.

55ml of water represents 10mm of rain. Therefore if you very carefully and slowly drip (don't pour )(and not directly into the small hole above the bucket, let it run down the side of the collector), that water into the gauge you should observe two things; one is you should hear the gauge tip 33 times. The second is you should see 10mm of rain register on your readout. If these things do not happen there is something wrong.

I have done this test 20 times now with both 55ml of water and 110ml of water. The gauge has consistently tipped 33 + or - 1 (or 66 + or - 2) times and read 10mm (or 20mm) on a readout. So this gauge at any rate is reasonably accurate.

If this does not happen check: a) wire connections; b) that the gauge is absolutely level, c) and that the tipping bucket is moving freely.

That is the accuracy check. Fairly good really.

Please remember the volume of water held by the tipping bucket is NOT the same figure as the amount of rain it represents. The figures are related arithmetically but are not the same numbers. (In the case of the FO Gauge each tip requires 1.66ml of water (55ml water / 33 tips = 1.66ml water) and this represents 0.3mm of rain which is the figure that really matters.)

Precision is another matter altogether. You will find this very difficult to achieve with this gauge. There is always going to be splash out given the design of the collector, and that will vary according to the intensity of the rainfall, wind etc etc. If you are tempted to put an extension on to reduce this problem be very careful how you do it. With such a tiny collector area (relative to professional gauges) a tiny increase (ie a few mm extra in any cross sectional dimension) will easily make the gauge over read by 10% or more. If you put a funnel or similar as your collector you will completely alter the recorded rain to the point that your figures mean nothing at all, unless your funnel has a cross sectional area of exactly 55cm2.

In summary the FO gauge will give a vague representation of rainfall over time. So long as this is accepted and the rainfall figures it produces not taken as a accurate, only representative; it is good value for money (@£4.99).

But if you want more accurate data consider an alternative such as Davis Rain Gauge or similar, remembering that each tip only on these better quality gauges represents 0.2mm (or 0.01" if it is an imperial gauge; NB: they are different and are setup differently mainly by the screw adjusters under each "Bucket" See attached) and one will need to alter the arduino sketch and / or the multipliers in Cumulus to get a correct rainfall figure. Davis or Texas Electronics ( and others are either calibrated as Metric or Imperial versions this is changed by altering the adjusting screws. Imperial gauges are calibrated to 0.01" or 0.254mm and metric to 0.2mm or 0.00787", as you can see there is a substantial 27% difference between the two and careful calibration is needed.

       

[JT118]

I think one needs to be careful of assuming there is some equivalence between a rain gauge calibrated in inches, which is most gauges supplied to the USA market and some (particularly older gauges) supplied to the UK market, and the metric calibrations supplied to the rest of the world.
I am not sure if this should go into the FO gauge section but is of relevance given the discussion on accuracy of FO gauges and how small errors can build to point where data generated is of no value.

Most gauges which are of reasonable quality and accuracy ( Davis, (http://www.davisnet.com/product_document...852_SS.pdf) Texas Instruments who supply Hobo and others, and various other tipping bucket gauges), are manufactured using the same moulds and dimensions for both imperial and metric versions.
The gauges are then factory calibrated either metric or imperial versions by altering the adjusting screws under the buckets.( see pictures in post 22) They are then sent out as either imperial or metric version with an appropriate code or stamp. For example in the case Texas with either and “M” or an “I” in the model code, or with Davis 7852 a sticker inside the gauge stating the calibration 0.2mm or 0.01”
This calibration is described in the previous posts. (see posts 7, 14 & 22 in this thread). Though when done in a factory the accuracy is even better.
Let us consider a 1000mm / annum rainfall area ( approximately the annual rainfall in Ayrshire, West of Scotland.)
In a tipping bucket gauge calibrated to:
a) 0.2mm gauge this would take 5000 tips.
b) 0.254 (or 0.01”) gauge would take 3937 tips
c) 0.2794 ( or 0.011”) would take 3579 tips
d) 0.3mm gauge would take 3333 tips
If one had an imperial gauge and assumed it was metric and 1000 mm of rain fell it would only register 3937 tips but the software if it had been set as a metric 0.2 gauge would only register 787 mm of rain that year. A very large discrepancy of 26%.
Assume software is set to 0.2mm / tip and 1000 mm of rain fell.
i) 0.2mm gauge would correctly register 1000mm of rain that year after 5000 tips.
ii) 0.254 (or 0.01”) gauge would incorrectly register 787mm of rain that year after 3937 tips.
iii) 0.2794 ( or 0.011”) gauge would incorrectly register 716mm of rain that year after 3579 tips
iv) 0.3mm gauge would incorrectly register 666mm of rain that year after 3333 tips.
These are large discrepancies and would mean that any data generated would be of little value when gauge calibration does not agree with software input.

   

[tyntop]

Hi,

Can you tell. Me weather the 0.2 and 0.3 tipping bucket corresponds to the amount of water it takes to tip the bucket.

 IE  0.2= 2mls of water
     0.3= 3mls of water.

I was just wondering about this as I have Timex TX6700 which was originally a wireless unit which I have converted to a simple reed switch setup. It has a 4" 100mm funnel and it takes 3.5mls of water to tip the bucket. I have used a syringe to measure and check this 35mls = 10 tips. This test was done several times to check that it was consistent

This would make this particular bucket setup a 0.35= 3.5mls and therefore, I imagine, would require a slight adjustment to the software

I have worked out for 20mm of rain that 157mls of water is required and at 3.5mls per tip it would take 44 tips  to register this and correspond to 20mm of rain fall.

Can any one tell me if I am correct with these assessments

Regards

Tyntop

[JT118]

Your 157mls of water is correct. However your assumption that 0.2=2mls of water is not.

A= pi r^2 =  3.143 * 50^2= 78.57 ml represents 10mm rain etc. Therefore your are correct 20mm rain = 157mls water.

As a check I would very carefully and slowly drip (don't pour )(and not directly into the small hole above the bucket, let it run down the side of the collector), the 157mls water into the gauge. If your syringe method is correct then it should tip nearer to 45 tips. 157/3.5=44.86 times.

If it does, 20mm rain/45  tips = 0.444mm of rain represented by each tip and it is this figure you need to put into your TX Config sketch also change your Cumulus multiplier accordingly to 2.2.  (0.444/0.20=2.2)

http://www.meteocercal.info/forum/Thread...figuration

[tyntop]

Great.

Thank you for your reply JT118, I think I have enough info now to be getting on with.

Much appreciated

Tyntop