Had a fantastic few days in Iceland courtesy of Lucy for a Christmas present and saw what we went went for, which was a spectacular display of Northern Lights.
More pictures of Iceland that we took are HERE.
The 30 minute Auroral Activity Forecast which I followed, proved to be an accurate indicator, link HERE.
Link to WXForum excellent article on Anemometer and Vane – HERE
25 October 17 my wind speed sensor failed after 9 years service, the symptom being that the wind speed is always at zero after checking the connection to the ISS is tight and the cable to the anemometer is not damaged.
This is the blog is how to replace the reed switch and test its operation, also while it was in bits, I thought I’d take the opportunity and replace the bearings as well.
Online sources of information relating to replacing the Davies 6410 reed switch –
Magnetic Reed Switch 10mm MKA-10110 100v 0.5A Russia £3.50 for 10 from eBay.
Metal Shielded Bearing 0.125 x 0.375 x 0.156 Part R2zz £0.85 each from rcbearings.co.uk
Remove the vane and wind speed cups to avoid damage.
Undo allen screw, if tight, use penetrating oil first, the screw does not need to come out.
Once the allen screw is loose, the vane is an interference fit, and with a little gentle force, pulls up revealing the potentiometers (pot) shaft.
The shaft is not ‘keyed’ but will have a mark where the allen screw tightened against, when you reassemble, use this to align the vane so the screw tightens in the same place.
I marked the shaft showing the allen screw mark, I also marked the body of the pot in relation to the housing so that when I reassemble everything is in the same position and the wind direction reading will not be out.
Using the same allen key, undo the securing screw on the wind cups, once loose, the assembly slides off the shaft with little force.
Put the wind cups and vane in a safe place till later.
Removing the wind cups reveals the reed switch which can only be accessed by removing the pot.
Breaking down the arm and releasing the cable, this is important when we come to remove the pot.
Undo the machine screw and slide this out, once the cable has been released from the in-built grips of the mounting bracket, the tube should slide out of the housing.
The cable inside the tube has a crude cable grip made from an off-cut of plastic hose, either use pliers or a wood screw in the center of the hose, and gently tease it out.
The next part is to slacken the two allen screws which hold the wind and direction body to the arm, once done, slide the arm down the cable to give you working room.
Removing the pot, this is a push fit and held in place by hot melted indents, these need to be cut away with a sharp knife.
Once the indents have been cut away, the pot will pull out of the housing, NOTE – this is a tight fit, use pliers to hold onto the pot shaft and draw towards you, Warning – I pulled too hard and pulled wires off the pot as their is not much slack in the wires from the reed switch, it’s not the end of the earth if you do though as I cut them off anyway!
Remove the gunge from the rear of the pot and note which colour wires go where, the red and black fly lead go to the reed switch and these I cut, I also cut all wires to the pot and removed this so I could easily work on the reed switch and bearings.
The reed switch is held in place with the same gunge (hot glue?) as is on the rear of the pot, due to aging it had gone brittle, using a terminal screwdriver it was possible to break this up from inside the housing body.
Once glue is removed, the reed switch assemble will slide out.
Reed switch replacement, the reed didn’t appear to be visually damaged, only slight rusting. Testing with a meter and magnet, the reed flexed but no electrical contact was made, checks on the circuit board tracks and resistors were OK.
The zoomed pictures make things look easier to handle than they are, the picture below offers some scale.
New reed soldered into position with the contact leafs horizontal to the orientation of the PCB, I also replaced the Red and Black wires from the PCB with more flexible ones.
Once the reed switch excess wire was trimmed, sliding the PCB into the housing body and getting it flush broke the reed.
The cause was the reed needs sit,as flat as possible,inline with the PCB, I had used too much solder and this lifted the reed wires slightly off the PCB.
The picture below is the second attempt where I used minimal solder on the PCB pads and pressed the reed wires onto the pad before using a small amount of solder to connect to the pad. This seems to have worked and allows the PCB to slide into the housing and sit flush once the wires were trimmed.
After sliding the reed switch in the body, I connected a multi-meter on continuity buzzer setting, sliding on the wind cups and spinning them, this should cause the buzzer to sound once on each revolution.
Once correct operation is proved, remove the wind cups and reed switch assembly, I sprayed the reed switch with a protective coating and put them in a safe place until later.
Replace the wind shaft bearing, first the circlip needs removing and retaining for later use, once the clip is off, the shaft can be pushed into the housing.
The shaft and black cap can now be removed.
My model has the brass follower looking from the cup side, other pictures I’ve seen show this to be the bearing with the brass follower inside the body, I reassembled in the same order as I took apart.
To remove the bearing, I left the brass follower in and used a terminal driver to go through the hole and using it at a slight angle, gently tap against the underside of the bearing, moving around the inside of the bearing and tapping to tease the bearing out of the housing.
The old bearing is in the bag and it is in good condition with only slight signs of rusting.
The new bearing simple pushed into the housing.
Reassembly:
I used a Glaxio Telecom Wall Box RJ11 which cost £2.69 from eBay as a breakout box for testing that the reed and directing pot are working:
My old 6410 is now refurbished and will be stored in the loft until the one in use packs up.
If you need any further information, please contact me.
25 October 17 @ 01:24 – Wind speed stopped recording, when I noticed I checked all the accessible connections, but still no reading, can’t complain, after 9 years its not done to bad!
Ordered a new one from Scaled Instruments, Gainesville, Florida, also I added a couple of other nice to have goodies:
| Product | Quantity | Price |
| Davis 6462 – AeroCone w/Bird Spikes and Debris Screen With Davis logo | 1 | $24.50 |
| Davis 7120.031 – Reed Switch for Tipping Bucket | 1 | $2.70 |
| Davis 6410 – Anemometer for Vantage Pro2 & Vantage Pro | 1 | $105.00 |
| Subtotal: | $132.20 | |
Delivery & Installation Progress
Updated 16 October 2024
Link to original Davis Fan Motor change
When the original Davis FARS motor failed I installed a new motor and kept back some spare motors, waiting for the inevitable motor failure to occur, reading online, changing from the Davis Fan to a PC varient made a lot of sense and not only is the life of motor excellent, but it is possible to monitor the motors output for operation.
I was interested to know the existing Davis fans air flow, mine was running at 2.4vDC and showed 9.7m/s:
The PC fan I bought was a Noctura NF-R8 redux -1800 PWM for £9.99 of eBay.
With the David FARS removed, the existing fan assembly slides out of the housing as a complete fan & surround, the hole left will take the 80 x 80mm fan with only a very minor filing of the fan case body to make it slide into the body of the FARS.
No modifications are made to the existing FARS body allowing reversion back to the existing fan if I wish (can’t think why, but you never know!).
To form a seal around the gaps, I used self adhesive door/window strip seal.
The finished fan after the seal is applied and checking that the fan is sucking, rather than blowing:
The oringinal fan ran on 2.4v, I replaced the voltage regulators fixed value resistor for the correct one to give 11.54vDC output and this feeds the motor and the motor monitoring Tacho.
The air flow of the PC fan when installed and connected was 5.5m/s, this is less than the original davis fan, but this is still better than static air within the sensing chamber.
The Davis weather station I have is the wired version, and I used Cat5e cable rather than the supplied 4 core cable from the ISS to the console, using the Cat5e unused cores, I fed the supply voltage to the fan and the Tacho pulse into the house where they are connected to a 52mm (2″) counter tachometer guage RPM, this was bought off eBay for £6.70 + £1.69 postage.
Tacho installed in the equipment cabinet:
The PC fan connections are:
The Tacho is set for a 4 cyclinder engine using a switch on the back of the unit, with the fan running the backlit RPM display is showing just under 1000RPM (reading slightly low due to voltage drop introducd by the distance from the ISS to the end point), to the left of the Tacho is my Meteobridge Pro weather station server to the internet.
The motor is guaranteed for 6 years, replacement will be very easy and I’m able to remotley monitor that the fan is functioning, all for less than £20.
16 September 2018, after 14 months continuous use, the motor failed, not even close to 6 years!!
I wrote to Noctua who are based in Austria and after asking me to perform some basic fault finding and requesting the fans data markings, agreed to send me a replacement fan, I in return was asked to send then a picture of the faulty fan with some of the blades broken off when the new one arrives.
While this correspondence was going on I bought a new fan off eBay for £12.50 which was installed on 22 Sept 18.
12 April 2020, whilst we are in virus lockdown I thought I’d clean and service the weather station and I noticed the fan had stopped working, so replaced it with the one sent to me by Noctura a couple of years ago, looks like the average continuous running life is two years for these fans.
30 March 2021, Nocura fan gave up, so much for a 6 year warranty, decided to go with a cheap and cheerful PC fan off eBay for £5.83.
16 October 2024, Reverted back to a standard unmonitored 12vDC PC fan (80 x 80 x 25mm) – Amazon £8.98.
Updated 16 October 2024
I decided to give the weather station a revamp, the two mini projects are the replacement of the Fan used to asperate the temperture/humidity sensor and the replacement of the original Davis temp/hum sensor with the more accurate chipset SHT15.
I bought the Davis 7346-174 upgrade from Scaled Instruments in Florida for $67.50 delivered (£51.94), the unit arrived very quickly as expected as I have used Scaled Instruments before and the service is exceptional.
Disassembly was quite straightforward after putting the station in install mode, what I was suprised by was the amount of dirt that had been drawn into to the fan guard and other parts of the Stevensons Sheild which all need a good wash with soapy water:
The original sensor is secured with two machine screws and the cable by a ‘P’ clip:
The replacement sensor was a direct fit as you wpould expects apart from the fact that instead of a ‘P’ clip, the environmental coating was used to form a cable clip, thit need a stand-off and additional machine screw to enable the sensor wire to be secured.
Once the housing was reassembled, sensor plugged up, fan reconnected and the station taken out of install mode everything worked just fine.
I’m awaiting the new fan to be deivered, so the second part of this will be blogged soon.
The LD-250 Lighting Detector from Boltek has an internal output for a relay interface, the manufactures units are quite expensive, so I decided to make my own.
Inside the LD-250 is a 14 way header which connects via ribbon cable to the RLO-10, off eBay I bought the 14 way ribbon cable and IDC cable mount socket for £5.00.
Opening the LD-250 the header JP1 is immediately obvious:
Using my multimeter, the header output pins linked to the front panel LED’s and the operating voltage was quickly found.
Using a spare strip of veroboard I mounted a magnetically shielded reed relay 5v, with flywheel diode across the coil, and the switched Normally Open reed output to a 2.54mm x 2 pitch connector, I also put veropins in the board so I can select which function I want the relay to operate on, should it be needed in the future.
The reed switch is used to switch 24v DC to an indicating LED and a a PLC input, the total load was measured at 21.49mA, well within the 500mA rating of the reed switch.
The module was placed in a small enclosure:
The ribbon cable was then plugged into JP1 inside the LD-250:
Switching on the Boltek performs a self test of the front LED’s and internal buzzer, as I have used the output from the ‘Close’ LED, the reed relay operated and the mast which was raised, automatically retracted.
All in all the project performs as expected and cost me £7 (enclosure was £2) saving me £58.95 on a factory unit.
Chatteris Weather has been publishing weather data to UK Weather Network (UKWX) for a number of years when Martin from UKWX Admin mailed to let me know my data was no longer being shown.
Originaly Weather Display was configured to send data to UKWX and during the changeover to Meteobridge Pro this got missed, fortunatly in Martins mail alerting me to the problem was a link on how to configure Meteobridge.
I followed the above steps but couln’t get it to work, Martin kindly gave me the details of Bob Montgomery owner of Bishop Sutton Weather Station who also uses a Meteobridge, and he was very willing to offer advice in order to help me get the data created in a format which UKWX can injest.
This is what I learned with help from Bob:
All done and time for a coffee!
Added Weather Station for WordPress to my blog pages, the application needs the following to work:
My native setup 4.6 provided by GoDaddy had 3 out of the 4, with Internationalization support missing, and thefore the app would not run.
Talking to GoDaddy techincial support, this was resolved very easily.
My main domain name is Chatteris. biz, Chatteris Weather and M0HTA.uk are linked to this domain name.
In order to give users confidence that the site they are linking to is secure, I have upgraded to SSL.
SSL (Secure Sockets Layer) is the standard security technology for establishing an encrypted link between a web server and a browser. This link ensures that all data passed between the web server and browsers remain private and integral.
During the transition it was found that some of the existing information displayed broke the security integrity of SSL, and therefore, I have either changed the menu to the remote link directly or removed the link completely, this has been unavoidable.
Tidied up the installation of my Meteobridge Pro as originally it was inside a metal patch cabinet and I wanted to try using it on wi-fi.
The unit has been on test since it was returned from repair and has performed really well with no issues requireing a reboot, dropped uploads have been attributed to network problems, hence moving it outside of ther cabinet and off the wired LAN.
I monitor uptime from the Status Page of my weather web site, cumulative downtime is recorded in a rolling 7 day period within the MORE setting HERE.
The front USB port has a micro 16Gb thumb drive for saving scheduled backups.
Power for the Meteobridge Pro and the Davis Vantage 2 Pro are both fed from a UPS in order to filter the mains to the adapters and to keep the units active during short duration ‘blips’ in power.
The screenshot below shows all the services the MB Pro is running perfectly with the bonus of significantly reduced power consumption.
| Power Data: | 5.47V 413mA 2.26W (Box Climate: 51.5°C 11%) |
