Sandaysoft

I'm afraid not! You'd need to change the programming in the chip.

_________________
Gina

Sorry, no banner - weather station out of action. Hoping to be up and running with a new home-made one soon.

Probably not. The issue is that the sensor TX is clocked to send at 48 second intervals, and also the RX in the display may well only be active for a few seconds before an expected next transmission. That may be done to save battery power at the display. It would also explain why the stations take time to sync up. To speculate further, it may be that the RX waits continuously at initial power up for a data transmission, then clocks further RX module power-ups to say 47 seconds later, just in time for the next transmission.
That's probably how I'd do it, to use as little battery power as possible in the display.

So if you follow my logic and speculation, in wireless mode it may require both ends to be modified for faster updates to be possible. Or, they missed an opportunity to save power...

All speculation of course at this time.

Martin
PS. Had another thought (again). If somebody runs two complete same systems near each other but initialized separately, do they ever get an issue where two displays seem to lock to one TX eventually ?
I'm thinking it's possible to have timing drift between two systems cause the TX with the slightly faster updates (like 47.999 seconds instead of 48.000) to 'capture' the second systems display...
If you don't follow me don't worry about it, it's not important right now

I've not had a problem with that when running for several daysIt might eventually but since the transmitter is crystal controlled I think it would be a long time.

_________________
Gina

Sorry, no banner - weather station out of action. Hoping to be up and running with a new home-made one soon.

Hi,

The crystals may differ in frequency by perhaps one part per million which represents 1 second in about 11 days. For this case, on average the messages from two transmitters might drift into collision in about nine months. The problem is that they may then continue colliding for many days or even weeks, totally blocking the transmissions for one or both systems. The duration of interference would depend on the length of the data messages and how long it takes the receiver to recover from a "wrong" message (e.g. AGC recovery or clock resynchronisation) etc..

If the transmitted data contains some "address" information then the receiver may recognise the "wrong" data and ultimately be able to lock onto the correct transmission again. But if there is no address (i.e. the receiver just locks onto any message within the expected timing window), then the receiver could permanently lock onto the wrong signal.

When I designed a similar system some years ago, I introduced a short, pseudo-random delay/offset (implemented with a "maximal length feedback shift register" emulated in firmware) for each message, so that never more than a few consecutive collisions could occur. If anybody discovers how (or if) the FO devices reliably accommodate multiple transmitters, then it's probably worth starting a dedicated thread.
_____

But back to the original topic, here is an internal photo of the wind vane head. I wanted to see the arrangement of the magnetic circuit with a view to improving the "equality" of the 16 compass sectors (and maybe damping the rotation as well). I will start a dedicated thread in due course, but for now it's sufficient to say that the magnet is the small vertical silver rod at the far right and the two empty recesses nearer the centre are perhaps for additional magnets if the relays fail to operate adequately on the production line. The bearing appears identical to that in the anemometer, i.e. with external and internal diameters of 10 and 5 mm, and the hole in cental "shaft" is 2.5mm.

The method I used to dismantle the head was to insert a thin, wide strip of steel (I used a 6" engineers' ruler, but a table knife or even a thin coin might work) into the gap between the rotating and stationary parts, twisting it gently and moving continuously around the circumference. After a minute or two, the gap was noticeably larger, so I continued until the head popped off. I cannot say if I was "lucky", therefore you attempt this method entirely at your own risk.

Cheers, Alan.

Hi to one and all

About a month ago I relocated my system to the middle of a field where I fly from. I took the opportunity to do some tests on the wind vane. I measured the resistance on the output and slowly moved the vane round the full 360 deg. It never switched on two of the reed switches at once; I only ever got the resistance readings of each of the single resistors.

I then reconnected the system and repeated the 360 deg. rotation with the same results; only the eight points were ever displayed. However, if during the 48 second period of non transmision I moved the vane plus and minus a few degrees from it's position I sometimes got a display reading in between the eight points.

Now that the system is installed in its new home all apears to be working normally but the prime points are recorded many more time than the other eight. From this I concluded that the other eight are software generated.

Any thoughts?

With hindsight I should have replaced the wind vane with fixed resistors to the transmitter to simulate two switches on at the same time and see what was displayed.

Light winds and blue skies

Philip

When I tested mine, the reed switch unit, I did get in-between direction resistance values but only over a very small range - the 8 main points covered a much bigger direction range.

_________________
Gina

Sorry, no banner - weather station out of action. Hoping to be up and running with a new home-made one soon.

Hi,

Sorry for a very late reply, but perhaps better late than never.

No, the eight intermediate points are not software-generated they occur when two reeds are activated together. The transmitter software does not even "remember" the last position, so if no reed switch is closed a "17th state" (no direction) is reported. But, as many have said, the intermediate positions are generally far smaller than the prime compass positions

In principle, the operation of two reeds at intermediate positions is almost an essential requirement. If one reed opened exactly when another closed, then minor tolerance changes could cause NO reed to be closed at certain positions and thus NO wind direction report . In theory, the reeds are part of a magnetic circuit so the actual opening and closing of the reeds may slightly "steer" the magnetic field (causing another to close/open), or the reeds might even "cog" the vane to preferred directions, but I suspect that this effect is very small.

Ideally, the sensor reeds require a sector-shaped radial field of about 67 degrees at the bearing/axis. Two bar magnets in the shape of a "V" (maybe even with the angle adjustable) could produce far better results, but are not easy to retro-fit. I did try shaping the field with soft-iron "poles", but it didn't work. Perhaps my magnetic material wasn't suitable, or the FO magnet is too short (and I don't think can be removed without damage to the head).

The solution I have now adopted is to attach a small "button" magnet * onto the original rod magnet, which makes the intermediate directions far more prominent. My intention was/is to measure all the reported directions (either by rotating the head very slowly, or by spinning it fast with a 'scope monitoring the resistances) and push on the rotating head until they are reasonably equal.

* The magnet which I used was removed from the "lid" fastening of the box (IMHO ridiculously over-designed) in which Krusell Phone/PDA leather cases are supplied. So sadly, I cannot recommend an economical source.

Cheers, Alan.

Amazon sell a range of small powerful magnets. I use a 6mm diameter by 3mm thick one in my home made anemometer. After a lot of experimenting, I gave up on magnets and reed switches for the wind vane and went for optical. I also tried "improving" the FO wind vane but gave up. Getting the magnetic field right is very tricky and the reed switches are not close tolerance on activation field strength. Also, they open at a lower magnetic field strength than they close at. With such a light vane on the FO they do show a slight sign of "cogging".

_________________
Gina

Sorry, no banner - weather station out of action. Hoping to be up and running with a new home-made one soon.

I also added a small super magnet on top of the one on the pointer. It made an improvement and has stayed in position for over a year now. I bought it (along with several others that are holding pictures on my 'fridge) here:
http://www.leevalley.com/en/wood/page.a ... 2363,42348
Great store for high quality, hard to find items!

Is the circuit card just a means to hold the reed switch, or does it have other components to send the signal to the transmitter?

I've had my FO for just a few weeks, and have learned a lot from reading this forum, most especially from Gina's many informative posts. Thank you.

I have a nice wireless rain gauge that I bought some years ago. It cost almost as much as the full FO station. The transmitter died and I never go around to fixing it. It has high edges and a 20 cm diameter funnel, and when it worked I found it corresponded very well to a manual gauge. I'm hoping to transplant the reed switch from the FO rain gauge to this one, which also uses a magnet on the tipper.

While the FO design is indeed cheap and cheery, I'm impressed by how much they pack into a very inexpensive package. I like to tinker and have all manner of plans, but this is a great value. I had looked at other PWS but wasn't willing to part with $500 or more. I have terrible experience with Oregon Scientific and won't buy their products again. When I found the FO on Amazon, and followed some links here, I had to try it.

The Cumulus software is outstanding as well. It's going to take me a while to learn how to use its many features. Thanks, Steve, for the effort and for hosting this forum.

Hi Aaron,

Yes AFAIK the PCB carries nothing more than the reed switch (and connector). The reed switch closes as the see-saw tips. But a few extra components (e.g. a resistor and capacitor to form a low-pass filter) might be beneficial, particularly if you have any plans to extend the cables.

Cheers, Alan.

Yes, nothing but a reed switch, and the cable is hard wired too. I recently replaced my cable with a longer one, using a phone extension lead with the connection cut off one end. Tried soldering the frizzy copper wire with embedded nylon(?) fibre to the board without success, and ended up wrapping and soldering it to short pieces of solid wire.
No problems subsequently with the data, only a longer wind cable is liable to induce spikes in the temp readings.

_________________

These images show the inside of my Maplin N96GY / WH-1081 transmitter circuit (433MHz) The red wire soldered to the board was added by me.

I can't quite figure out where the antenna is on this circuit. My guess is it's the blue cylindrical part that looks a bit like an inductor. I've never seen an antenna that looks like this (certainly not for low-power 433MHz that's meant to range 100m), so I'd be grateful if anyone could correct me on this.

As it happens, the 17.5cm wire I added does *seem* to increase the range slightly, but not by much.

the track that goes around the "L" cutout is the twig.

That was my initial thought, but I can't figure out why it would continue into the ground plane.

Also, that blue component that looks like an inductor actually has infinite resistance across it, whereas an inductor would have close to zero.