Jump to content

SkySurveyBanner.jpg.21855908fce40597655603b6c9af720d.jpg

My rain gauge has died :(


JamesF

Recommended Posts

For many years I've had a 1-wire "tipping bucket" rain gauge sold by AAG.  I gave it a service a couple of years back and made a "crown of thorns" to stop birds perching on the edge of the funnel and, err, blocking it up.  Until the last month or so it has worked very nicely.  Now however it appears to have died completely :(

The 1-wire device still shows up in Linux and is readable, but the 1-wire counters never change.  I've checked by dribbling water into the funnel from a jug and the buckets tip, but there's still no change in the counters.  The circuit for the system is quite simple as far as I recall, being little more than the 1-wire counter chip and a reed switch in a housing next to the buckets, with a magnet fixed to the buckets tripping the reed switch each time the bucket tips.

It could be a failure of the 1-wire chip, or (as I suspect is more likely to be the case) the magnet is not activating the reed switch for some reason, perhaps because it passes the reed switch too fast.  I have thought about trying a different, perhaps stronger, magnet, but finding one of a suitable size and weight looks like a bit of a lost cause.

I could throw in the towel and buy a new rain gauge, but they all seem to want to come with displays or be wireless or other stuff I just don't want, with a price tag to match.

I'm about to disable it in the software and bring it inside to see if I can do anything, but at the moment I'm not feeling very hopeful :(

On the other hand, if anyone knows of a simple rain gauge (I don't need any other sensors, wireless or a display) that connects to, say, a Raspberry Pi, either complete or DIY, pointers would be gratefully received.

James

Link to comment
Share on other sites

My money would be on the reed switch. These can be replaced, assuming you can find the right one, but sometimes they stick in the closed position and tap on the outside body of the rain gauge body is enough to release them. I haven't had this on a rain gauge but I did on an anemometer. 

The only people I know of who still do 1-wire sensor etc is HomeChip so maybe worth contacting them to see if they can help. 

I also found a reed switch for a David VP2 rain gauge at Weather Spares for £2.00. ;)

I used to run a 1-Wire station years ago, but now it's all Davis VP2 equipment.

 

Link to comment
Share on other sites

Ah, yes, HomeChip might be worth a go.  I ran out of time to dismantle the rain gauge before it started to get dark, so I guess that will have to be a job for tomorrow.  I did at least find my spare DS9490R, so I don't have to disconnect everything else to be able to test it.

I assumed the reed switch would be normally open and therefore unlikely to fail closed, but I guess it's possible.

James

Link to comment
Share on other sites

I don't know if the reed switches get sticky over the years but, as I say, I've at least one reed switch go bad in an anemometer and thump on the pole was enough to free it for a while. This one kept sticking so was replaced in the end.

I've still got an iButton DS9490R  here, along with an old Hobby Boards Solar Sensor.

If they're of any use then PM me your address and I'll pop them in the post. No charge, I've not used them in years so not sure they still work. Although the DS9490R is recognised when I plug it in to a USB port. ;)

Link to comment
Share on other sites

16 hours ago, Budgie1 said:

I don't know if the reed switches get sticky over the years but, as I say, I've at least one reed switch go bad in an anemometer and thump on the pole was enough to free it for a while. This one kept sticking so was replaced in the end.

Now I have mine on the desk and opened up it looks as though the reed switch might also be stuck closed.  There are further complications though :)  I'll post some photos later.

16 hours ago, Budgie1 said:

I've still got an iButton DS9490R  here, along with an old Hobby Boards Solar Sensor.

If they're of any use then PM me your address and I'll pop them in the post. No charge, I've not used them in years so not sure they still work. Although the DS9490R is recognised when I plug it in to a USB port.

That's very kind of you to offer.  I can certainly put them to use as it would allow me to have two 1-wire channels and I always wanted to add a solar sensor.  I'll PM you.

James

Link to comment
Share on other sites

Ok, so this is the reed switch:

rain-gauge-06.jpg

The two-pin header is my later addition, as it allows the board to be disconnected from the cable, which was originally soldered straight into the PCB.  Most of the components are on the other side, buried under a load of hot melt :(

rain-gauge-07.jpg

Even the battery terminals are welded directly onto the battery itself.  No expense spent :)  Fortunately the battery is only there to preserve the counter values in the 1-wire chip over a loss of power.  I can live with that sort of failure I think, though replacing it with a proper battery socket does quite appeal.

The circuit actually looks quite simple and I'm half-tempted just to rebuild a new board from scratch but for the fact that I can't find anywhere selling the TAI-8575 any more and if it were surface-mount I'd still not be able to do it, so I guess the first job is to peel off the hot melt, and then try to work out if the reed switch is actually intended to be "normally open" but is stuck closed, or if it really should be "normally closed".  I think I need to find a magnet...

James

  • Like 1
Link to comment
Share on other sites

Oh dear :(

I just attempted to gently straighten the leads for the reed switch a little so I could get a better view of the mechanism and the glass of the switch came apart as soon as I touched it.  I suspect it may have already failed and my attempting to move it was the last straw.  If I can't find a circuit diagram online then I think I'm going to assume it must be a "normally open" switch.  I'd guess that if it were normally closed then current would be flowing somewhere all the time which doesn't seem like a good decision for a 1-wire device.

Having had a better look over the PCB with a magnifying glass I think "TAI-8575" might be the part number for the entire thing.  The 1-wire counter itself looks like it is probably a DS2423 which I don't believe has been manufactured for years.  If I can't get it working with a replacement reed switch then it looks like I might have to try something along the lines of replacing all the electronics (perhaps using an Arduino?) in the existing hardware and find a different way to get the data onto a PC.

James

Link to comment
Share on other sites

Hmmm.  I've just found a datasheet for the DS2423 which says that the counters increment on a low-going edge, suggesting that the reed switch may be "normally closed" after all, or that somewhere there's an inverter in the circuit.  At this point I think a circuit diagram would be helpful.

James

Link to comment
Share on other sites

I found this with a wiring diagram, not sure if it helps at all: http://www.midondesign.com/Documents/1wire_raingauge.PDF

It says: "A magnet on the tipping bucket mechanism momentarily closes the magnetic reed switch each time a
bucket tips and empties."
Although this is for a Dallas system, rather than AAG, so I'm not sure. Have a look and see if it makes any sense. ;)

Link to comment
Share on other sites

I definitely have a different circuit as there's no DS2407 on mine.  I think one is based on the other though, so it may well be reasonable to assume that the reed switches work the same way.

James

Link to comment
Share on other sites

James, most reed switches are normally open, being closed when affected by the magnetic field.  I think the negative edge trigger is a read herring, as a normally open reed switch, having been triggered will go form high to low when the magnet releases the switch.  In the sea-saw arrangement (if the module is similar to the Maplin stations) rain water fills the small bowl on the sea-saw, causing it to tip.  If the magnet and reed switch is on the opposite side this would result in the negative going "logic" triggering a counter, as the magnet would be moved away form the switch allowing the reed to open.  If however if the magnet and reed are under the bowl the reed will be closed when the sea-saw tips under the weight of the water, but then as the water is emptied, the reed is opened again as the sea-saw gets reset once again.  

I'm sure I have a selection of reed switches left over from the days of using them with model railways for train detection if you need some.

Edited by malc-c
Link to comment
Share on other sites

Another kind offer, thank you.  I've got a few on order now though, so as long as I don't break loads trying to replace the existing one then I should be good :)

In the case of this rain gauge, the magnet is at the top of the wall between the two buckets and the reed switch placed such that the magnet passes it as the bucket tips from one side to the other which I think is the last of your options.  I guess that makes sense in that when either of the buckets is filling, the switch is inactive because there's no way to know how long it will be until the next tip.

I don't particularly like that the magnet is inside a plastic housing on the bucket see-saw and the reed switch is inside another plastic housing sealed away from the water.  It's not a very powerful magnet (about 3mm in diameter and 5mm long) and putting possibly as much as 3mm of plastic between it and the reed switch doesn't seem like it would help reliability.  I have toyed with the idea of putting a neodymium magnet in instead, but of course the risk then is that the switch is always activated :)

For completeness, I'll add my attempt at the full circuit diagram for the board here as well (may not be entirely accurate as I've drawn it by following the traces on the board which are not always clear.

TAI-8575.png

James

Link to comment
Share on other sites

Googling the part number the DS2423 is a one wire EPROM with a function. 

According to the datasheet, the counters are triggered by inputs A and B, respectively, increment with every low-going pulse on their input.  I'm not an electronics expert, but from your schematic it would seem that in order to do that pins 5 and 6 are pulled down to GND via the 1M resistor, but in order for the pins to register a pulse that must mean that they are held high through the transistor, which as there is no current flowing to its base will be turned on (its a PNP type).  So when turned on, current / voltage flows through the transistor to ground with the 1M resistor forming the "load" inputs A/B are held high.  When the reed switch is closed by the magnet voltage flows through R4 and R3 to ground.  With current momentarily flowing through R3 and R4 the transistor is turned off,  so inputs A/B are grounded via the 1M resistor.  

The only bit that confuses me is the Diode arrangement.  Normally one wire data lines are pulled high via a 4.7K - 10K resistor.  In the schematic, D1 will take the 3v form the battery and allow current / voltage to flow to the  transistor and vBat terminal.  I think there would be a 0.7v  voltage drop as this is normally associated with diodes.  D2 blocks power getting to the data line, but I have no idea how it would effect the data line.

Like I said, I'm no electronics engineer, but that's how I follow the schematic....

If you get stuck with the reed switches  I have these two which I can send you.  One is very small and sensitive (well it needed to detect a small magnet on a passing n gauge train)

P1019633.thumb.JPG.d3f5908c371de474e782839b606cc8fe.JPG

Link to comment
Share on other sites

My reed switches arrived today.  They're the same size as the original to within fractions of a millimetre, but after connecting one to a multimeter it's clear that the existing magnet struggles to trigger it.  I've ordered some tiny neodymium magnets to see if they can do any better.

James

Link to comment
Share on other sites

Well, I ran out of time this weekend.  For the best of reasons though, as we spent much of Sunday catching up with friends we've not seen in a fair while.

I ordered some tiny round neodymium magnets (1mm deep by 2mm diameter) and I think my first experiment once everything is reassembled will be to add a few to the end of the existing magnet to see if that helps things at all.  Having had little experience with reed switches it's interesting that the magnet has far more effect at the ends of the switch than it does in the middle.

I couldn't find any holders for a 3V battery (something like a CR2032 for instance, which we usually have plenty of) that looked suitable and weren't going to take six weeks to arrive from China, so in the end I bought a holder for some AAA batteries instead.  I reckon there should be room for that inside the sealed compartment that contains all the electronics.

That's my evening sorted then :)

James

Link to comment
Share on other sites

On 19/11/2021 at 00:04, malc-c said:

...but in order for the pins to register a pulse that must mean that they are held high through the transistor, which as there is no current flowing to its base will be turned on (its a PNP type)...

It'll be turned OFF by lack of base current when SW1 is "open". There is no bias until SW1 closes, so the transistor is effectively not there. It's a bipolar transistor not a FET, so no base current = no collector current. PNP and NPN are opposite polarity but work the same way.

Link to comment
Share on other sites

It never rains but it pours, ho ho.  It seems that this evening my air pressure sensor has started throwing a fit as well :(

Up to about 4:30pm I had a pressure reading of about 1034 mb and temperature of 7.18C which is a bit high, but then it always reads about 1.5C high.  Then it started going crazy and producing figures like -152.52C and 717.4 mb or 314.67C and 1643.4 mb interspersed with values that look fairly normal.

It could be that for some reason it's not happy about the temperature which is currently down below 2.5C for the first time in ages, but that wouldn't be helpful.  Still, I'll worry about it after I have the rain gauge sorted.

James

Link to comment
Share on other sites

Woohoo!  Rain gauge reassembled to the point of being able to test it with the new reed switch and battery pack, and it appears to count "tips" without a problem.  When I unplug and re-plug it, the existing counter values are retained as well.  Excellent news :)

So, tomorrow I shall completely reassemble it and make sure the calibration is correct by dripping water in from a syringe.  Once that is done I reckon it's ready to go back up.

James

  • Like 1
Link to comment
Share on other sites

And to calibration...

My gauge appears to measure 112.5mm in diameter (just under 4.5") which is a bit on the small side I think.  I guess it wouldn't be that hard to fit a larger funnel if I wished though.  A bit of research suggests that a diameter of 8"/200mm is considered reasonable.

Anyhow, I calculate the collection area of my gauge at 9940mm².  Dripping water into it from a syringe suggests that the bucket volume is something very close to 2.5ml, meaning that each tip would represent 0.25mm (ish) of rain.  Within the level of accuracy I'm able to achieve it therefore does look as though each bucket tip probably measures 0.01" of rain (I think most of this stuff was intended to be used for imperial measurements).

And now it's time to get the step ladder out :)

James

Link to comment
Share on other sites

1 hour ago, JamesF said:

Anyhow, I calculate the collection area of my gauge at 9940mm².  Dripping water into it from a syringe suggests that the bucket volume is something very close to 2.5ml, meaning that each tip would represent 0.25mm (ish) of rain.  Within the level of accuracy I'm able to achieve it therefore does look as though each bucket tip probably measures 0.01" of rain (I think most of this stuff was intended to be used for imperial measurements).

Yep, when AAG was manufacturing these they were mostly for the U.S market, so 0.01" would be correct.

When I got my first Davis Vantage Pro station, you had to change the weight under the tipping buckets to change it from 0.01" to 0.2mm per tip. Now they sell the rain gauge already set for either inches or mm. ;)

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.