Topic: Better calibration procedure for the rain sensor

[ebweather]

Update: A better procedure is available in one of the comments! (link)
Below is my original procedure, now obsolete.  Read on to judge how much more complicated it was than needed to be.   

After struggling to calibrate using official procedures, I started using math. Would love to know what you guys think of this.

What did not work: I've tried official procedure of poking small hole in a reference cup, measuring 8 oz of water and trying to adjust the screws based on percentage of error, had big difficulties getting any meaningful accuracy in adjustment and tests would take a ton of time. After that I've tried getting a dropper and made sure there were equal number of drops between the bucket (aka collector cup; aka self emptying rain collector) tips, followed by the 8 oz cup calibration. Again that took far too long and did not bring me closer to meaningful calibration. Here is what I ended up doing in order to get perfect 1.06" out of 8 oz.

What worked: I ended up realizing that in order to get 1.06", 8 oz of water needed to tip the rain bucket 106 times, thus each "tip" should carry (8 oz / 106) = 2.232 milliliters of liquid. Next I went to my local walgreens and asked the pharmacy to give me an accurate syringe with milliliter (ml) or cubic centimeter (cc) [1ml = 1cc] markings (they are free if you ask them nicely). Syringe I got from walgreens had 1ml major (larger) markings, each separated by four 0.2 ml minor (smaller) markings.  So here is what I did to accurately calibrate the unit:

Before we get started, I would like to recommend that you try to calibrate the sensor where it is installed, do not take it elsewhere as recommended by official procedure.  Sometimes this will not be possible, due to the height at which the sensor is installed, but you will get better accuracy if you calibrate the sensor right where it will be used. This calibration will take care of any slight deviation, if any,  from perfect level, so calibrating it while the unit is leveled elsewhere will not be as effective.

What you will need: A phillips head screwdriver (to adjust the calibration screws), a cup of water, a syringe with ml or cc markings (usually available from a local pharmacy) and something narrow which could fit into the drain slots (a drinking straw can work).

What is NOT needed: Connection to the display / internet bridge. This calibration will use visual observation and information from the syringe, I recommend you remove at least two bottom batteries to stop the sensor from transmitting while you are doing calibration. This will prevent the weather data from being contaminated with false rainfall readings.

• Understand that calibration of each side is independent of the other, don't try to calibrate both sides at the same time. Do one side first and than the other.
  
• Don't worry about screws not being adjusted by the same amount at this time, this procedure does not require it.
  
• My final calibration ended up being 25 quarter turns for the screw nearer the pole mount and 15 quarter turns for the screw farther the pole mount, you may want to start there and make adjustments as necessary. I counted quarter turns from the screws being screwed in all the way (my 0 point).
  
• Pick a screw to calibrate, as an example we'll start with the screw nearest to the pole mount. Start by carefully inserting a straw or a similar slim object into the drain by the other (far) screw. This will tip the cup onto the screw you are calibrating and make sure you are starting with no water in the cup.
  
• Fill a syringe with about 5ml of liquid and start slowly pouring the liquid into a rain funnel. Your goal is to get to the point where it tips after you poured in 2.2 ml (with 2.8 ml remaining in the syringe if you have started with 5 ml). You can start quick and pour in 1.2 ml a bit faster, than you should slow down and start dripping in the last 1 ml carefully, stopping when the cup tips.
  
• If you have started with 5ml, your goal should be (5 - 2.2) = 2.8 ml to be left in the syringe. If the cup tipped too early (you have more than 2.8 ml remaining) unscrew the screw you are calibrating a bit further (you need to decrease the amount left in the syringe, conveniently this is the same as decreasing the rainfall, so turn in the negative direction as indicated). If the opposite is true (you have less than 2.8 ml remaining), screw in the calibration screw (you need to increase the amount remaining, which is the same as increasing the rainfall).
  
• Wait a few seconds to make sure all the water drained out of the collection bowl (which you may have already been doing while making adjustment in step 6) and reset the cup by inserting a straw into a far drain to manually tip the cup back onto calibration screw and make sure you are starting with empty cup.
  
• After you get a good 2.2ml used (or 2.8 ml remaining, however you wish to think of it) reference point, I recommend you screw in the calibration screw (turning clockwise)  all the way in, counting each quarter turn as you do so. After that unscrew the screw again the same number of quarter turns to establish your calibration number. Verify that the cup still tips after 2.2ml and save the number somewhere in order to get back into calibration quicker next time you need to do it all again.
  
• Repeat steps 7 through 12 for the other screw
  
• After both sides tip after 2.2 ml used, you can reinsert the batteries and verify the calibration by using 8 oz of water in a dripping cup to get 1.06 inches of water method.
  

Remember to save your calibration number of quarter turns somewhere you have a chance to reference them. Share them here so others may have a better starting point.
Whenever testing 8 oz drip measurement, I recommend you use two cups. Make one cup with smaller hole which works well for high water level and another cup with bigger holes to use with low water level. Simply start the test using the cup with smaller holes, and switch to another cup (and pour any amount remaining in the original cup) when the drip rate gets too low.

For those who wish to use this procedure for 00875 (or similar) rain gauge, it is mostly the same. The main difference is that 8 oz of water should tip the cup 220 times, instead of 106, thus your target should be (8 oz / 220) = 1.075 milliliters (ml) or cubic centimeters (CC).  Another difference is that the screw will need to be raised (unscrewed) in order to decrease the tipping amount (such as to go from 1.3 ml down to 1.0 ml) and screwed in (lowered) to increase the dipping amount (such as from 0.8 ml to 1ml). You may need to invest into a more accurate syringe in order to achieve a higher accuracy.

Well, this is what I did to get a perfect (equal) rain calibration without having to do long 8 oz cup tests or tedious drop counting. Remember to do one screw at a time and not to worry about having to adjust both screws the same amount. Hope this procedure will help you as well.

ELI5 version:
Idea is to use syringe which can be obtained for free from your local pharmacy. You should measure how much water it takes to tip the internal rain collector (as can be seen on this image).

Since we know that AcuRite counts 0.01" of rainfall each time a sensor is tilting and we know that 8 oz of water should result in 1.06" of rainfall measurement or 106 tilts, we can plug this into google: "8oz / 106 in ml" to find your target (in milliliters or cubic centimeters, which is what syringes tend to have markings for) which would result in perfect calibration.

To make things easier, you should consider instead how much water should be left in the syringe. For example if you fill up your syringe with 5 ml of water and you need to dispense 2.2ml, you should be left with 5 - 2.2 = 2.8 ml of liquid in the syringe. Furthermore decreasing the amount of liquid in the syringe (such as to go from 3ml remaining, down to 2.8ml) is the same as decreasing the rainfall (so turn the screw in the negative direction), while increasing the amount of liquid left (such as to go from 2.5 ml remaining, up to 2.8ml) is the same as increasing the rainfall (so turn the screw in the positive direction).

Now that we know what to aim for, keep in mind that calibration of each screw is independent of the other. When screw is active (being used by the system), the measurement sensor is leaning against it and when enough water (ideally 2.23 ml) gets into the sensor, it will tip away from the active screw and dump the water out of the drain next to inactive screw. So when you see the water pour out of the drain next to one of the calibration screws, remember that it was the other calibration screw that was responsible for how much water it took to tilt and adjust accordingly.

I recommend you calibrate one screw at a time (do NOT adjust both at the same time as officially advised), for this you should use a drinking straw or use original narrow Rain Gauge Stabilizer (a bit of plastic that you were told to remove and throw away) in order to manually tip the measurement cup back onto the screw you are calibrating. You do so by carefully inserting a straw (which can be pinched narrow) into the drain next to the screw you are NOT calibrating until you feel some resistance (about 2 inches or so). Repeat this as necessary until it takes 2.2 ml of liquid to tilt the measurement cup.

Repeat the previous step with the other screw and the other drain.

Good luck, cheers

P.S. I've made a measurement, as accurately as I could, of the rain funnel opening area and determined it to be at most 13.6041 in² or 8776.8 mm². This means that 8 oz of liquid is actually at least 1.0613 inches of rainfall and that the perfect calibration is somewhere between 2.22 and 2.23 ml per tip.

[daman]

I voted to complex...

Nice wright up but to time consuming, I find the factory settings to be pretty accurate compared to my manual tube type gage.   

[DaleReid]

I can understand the urge to be as accurate as possible.  I have gotten my kitchen floor soaked more than one weekend in my life trying to do the minute dribble rate vs the calibrated tip volume and even an old IV from a hospitalization with the tube adjuster for a tiny amount slow enough to take into account the slop with each tip.

In the long run most of the runs lead me to believe that things were pretty close to start.

A friend had the similar urge.  One summer he had something like six raingauges all within a 10' circle in the clear.  The ranged from a rectangular throat cheapie, to a very expensive tipping bucket to a glass vial like the old implement companies gave to their customers at Christmas and new years to an ancient brass Taylor with a knife edge collecting throat.  After many rains some gentle to a few real frog stranglers, the results the gauges gave were surprisingly similar, with one exception.  During one particularly heavy storm, the Taylor had almost 3/4" more in it than the others indicated, despite him being very sure it had been dumped correctly, and no bird had 'added' to it nor the sprinkler on the garden, etc.

I have two new-to-me high end gauges 'salvaged' from eBay so they would have a good home and be cleaned and serviced with love and care that I undoubtedly will set up and run a few test volumes through to see just how close they are to accurate.  That obsessive behavior is part of trying to do the best we can, especially at our hobby.  And then if there are differences between the amounts reported, I'll blame it on microclimate variations from one collector to the next.

PS I was just reminded about the effort to try to be the best one can and the effect it had on some part of our history. I got the story from the daughter of the engineer involved, and since she isn't prone to magnifying the stories she tells, I have a feeling there may be some kernel of truth to it.  Her father worked in Houston for NASA during the space program. His contribution to the project was summed up by replacing a potentiometer with a couple of high grade resistors.  It turned out that as the circuit board for whatever function it was, passed through numerous calibrations and attempts at being dead on for a received downlink telemetry during manned launches (after all astronauts' lives and health were at stake) the thing got adjusted so much back and forth to find the center zone, that things got loose and it jumped about much more than expected.  He traced to over-adjustment and either a much more expensive part would need to be substituted that could withstand all that racking back and forth, or a fixed value was substituted once the optimal value was determined and was never adjusted again. I guess it worked ok, and saved a lot of money and nervous technicians overcalibrating the receivers. 

[ebweather]

Thanks for the feedback, I did make it too complicated didn't I?  I'll see if I can put together a demo video sometime in the future, procedure itself takes about 5 to 10 minutes to accurately calibrate the rain sensor. This is actually far less than doing even one 20 minute 8 oz drip test. If you ever consider calibrating the rain sensor in the future, check back to see if a better instructions are available. Cheers 

[CW2274]

After jacking up my Rainwise, I merely used my Cocorahs, a digital kitchen scale, and a milk jug to get it right again. After half a dozen rain events, nothing torrential thou, they've matched exactly. I quite pleased...so far, but see no reason it wouldn't continue to be acceptable.

[DaleReid]

I would love a concise video to show how you approached it.  Thanks.

[CW2274]

I would love a concise video to show how you approached it.  Thanks.

Me?

[ebweather]

How about this for simplified procedure?

1) Get a 5ml syringe from a local pharmacy (usually free of charge) and remove at least two batteries from your 5-in-1 (in order to prevent false rain-data being recorded)
2) Unscrew both calibration screws a few turns (in the -ve direction), to make sure they underestimate the rain.
3) Use a drinking straw to tilt a rocking cup onto one of the screws (insert in the vent opposite to screw you wish to calibrate).
4) Intake 2.2 ml worth of water into a syringe (see attachment for example)
5) Pour the liquid into the rain funnel, cup should not tip at this point. If it did: Unscrew the screw you are calibrating further, and repeat steps 3 and 4.
6) Slowly screw in the screw (in the +ve direction) until the cup tilts.
7) Repeat steps 3, 4 and 5, this time being very careful to drip at the end. The cup should tilt on the last drop.
8) If the cup did not tilt at the last drop, repeat steps 6 and 7. If it did, congratulations, you are done with this screw, repeat the process (3+) with the other screw.

No need to remember you calibration number, it's actually pretty easy to calibrate each time you need it.
Video is coming this week, stay tuned.

Cheers.

[daman]

Nice, yea that's seems more simpler  , a video would be great also.

[Weatherdude24]

Nice, yea that's seems more simpler  , a video would be great also.

Did you ever make a video?