Dropping link to external sensors

[DavidG]

I have a Smartweather WH2080. It arrived last Christmas and worked perfectly until June (so about 6 months of operation).

It suddenly started to loose communication with its external sensors. This has nothing to do with the link to the PC or Cumulus as it happens regardless of whether the base unit is linked to the PC or not.

Initially I assumed batteries needed to be replaced, but this made no difference, nor did cleaning the battery contacts.

Experimentation has found that placing the base unit on a hallway windowcill in direct line of sight to the external sensors gives 100% successful communication. When in this position the distance from the base unit to the sensors is 15 metres. If I pick up the base unit and carry it into my office where it used to be for the first 6 months I am moving it an additional 5 metres away from the sensors and through one internal wall. It does not loose communication immediately and I can see it gathering additional readings but within about 5 minutes all of the external sensors readings on the base unit go blank. Once this happens it never reestablishes communication until I move it back into the hallway.

If it had never worked I would have assumed I was simply too far away but it worked so well for 6 months! I assume that either the strength of the signal being transmitted or capability of the base unit to receive the signal has changed in some way?

Can anyone suggest anything I can try to restore things to how they used to be? I would like to keep the base unit permanently linked to the PC as I used to do.

I apologies for posting something that has nothing to do with Cumulus which I think is an excellent piece of software, but I was hoping one of the many users might know something I could try.

Many thanks

David

[AllyCat]

It does not loose communication immediately and I can see it gathering additional readings but within about 5 minutes all of the external sensors readings on the base unit go blank.

Hi David,

Are you sure about that? My experience is that it takes about 5 - 6 minutes after communications are lost before the "-- --" appear? Are you sure it's updating the external data (barometric pressure and "internal" data are measured within the console)?

The voltage from Alkaline (and the cheaper "Carbon") cells gradually falls from 1.5 volts down to about 1 volt over their life so the transmitter power (and thus maximum range) is likely to gradually reduce over the months. There is also the possibility that a "neighbour" has acquired a system which uses the same radio frequency and is causing interference.

In general, the quoted "open field" range of systems of this type can be very misleading, and it's largely a matter of "luck" whether you have a system (and location) which works well, even at moderately short range. Also, the transmitting frequency is such that "standing waves" can be set up, so that moving (or rotating) the transmitter or receiver by just a few inches can make the dfference between success and failure.

Cheers, Alan.

[DavidG]

Hi Alan

Thanks for replying. I have 'studied' this behaviour for the last few months before posting here.

In reply to your thoughts:

The base unit has small icon on the display (bottom right between the humidity figures) that lights momentarily when it is communicating with the external sensors. It comes on roughly once every 40 seconds and then the display values update. So frequently changing things such as wind speed and direction do alter even after I have carried it through into my office - until the comms fails and I get the dreaded "---" display on the wind, rain, ext temp and ext humidity. I am aware what comes from an external sensor and what does not...

Next to this icon is another which is supposed to light up if there is a weak signal from the transmitter - indicating that new transmitter batteries are required - this has never lit under any circumstances.

I have fiddled about with several sets of brand new batteries (and cleaning the contacts) in both base unit and transmitter but much to my surprise I have never been able to restore the original behavior.

I live in a rural area with no close neighbours to interfere with my signal. But you have set me thinking, is there any possibility that a change in the make of the wireless internet router that I use could cause interference where the previous one did not?

Thanks for your interest

David

[AllyCat]

Hi David,

Sorry, I don't know much about the 2080, my experience is with the 1080/1 and 3080. I had believed that the 2080 is more similar to the 1080 than the 3080, but that might not be the case. To add to the confusion, FO (branded Maplin in the UK) appear to be selling a "1081" which uses a wireless protocol (more) similar to the 3080.

AFAIK the "communicating" icon (on the 3080, there is no such icon on the 1080/1) indicates that the Console is "listening", i.e. the receiver is switched on (it's switched off most of the time to conserve battery power), NOT that it is receiving data. However, if the displayed external data changes then obviously data IS being received (unless the Console is doing some local averaging, but I don't believe that's the case).

Also, the "TX within a battery" icon reports a status flag which is contained within the transmission when the Transmitter battery is measured as low (below about 2.5 volts). With the 1080 Console the icon appears to be updated only at midnight, but the 3080 updates with each transmission. AFAIK this TX icon is not set if a "weak" signal is received (I don't believe the receiver has this capability) nor if no signal is received for longer than a "timeout" period (which is indicated by -- in the display and an error message within Cumulus).

Certainly a new router might cause more/different interference, but don't discount possible changes in signal strength (due to "multipath") that can be caused by a very small (inches) alteration of the position/orientation of the Console or the transmitter.

Cheers, Alan.

[ntinos]

Hi David

One solution is if you have soldering iron to add more antenna at console receiver.

You have this weather station
http://www.foshk.com/Weather_Professional/WH2080.htm

I see is 433 Μhz ,because inside usually it have for antenna a small cable,will be better you add one-ply 0,5 mm cable for example one wire from cat 5-6 ethernet cable ,lenght 34,6 cm L /2 tuning wave 433 Mhz inside console all around .

PS: You will replace all the small antenna cable where it have console before ,and you put a new with total lenght 34,6 cm

[DavidG]

Hi Alan and ntinos

Thank you both for your comments.

You both know more about the way the hardware works than I do. I am tempted to try ntinos suggestion of improving the antenna in the receiver if I can be brave enough to take it apart!

If I do fit a longer antenna wire - exactly 34.6cm long as you suggest - will it be ok to put it all inside the case or will it need to hang out, perhaps in a straight line?

Thanks again

David

[ntinos]

Hi Alan and ntinos
If I do fit a longer antenna wire - exactly 34.6cm long as you suggest - will it be ok to put it all inside the case or will it need to hang out, perhaps in a straight line?

Thanks again

David

Ηi David

It will be better is outside at straight line ,i said inside for you not open hole at plastic for passed the antenna cable out console.But when you make this mod try you see first inside results,and if you do not have good results,try outside at straight line.
For sure you read your manual if your weather station work at 433 mhz ,if it work at 868 mhz the total lenght must is the half 17,3 cm.

[DavidG]

I have looked in the manual and found the Specifications for Outdoor data where it says:

Frequency: 433MHz/868MHz

This is not very helpful !

But on the back of the plastic case of the base console where it says WH2080 it also says 433MHz.
So I must believe this I think.

There is a sticky thread on this forum called "Photos of the insides of Fine Offset sensors"
If you scroll down to pictures 7 and 9 of the main console PC board I can see a red wire about 5-10 cm long, is this the antenna you suggest I replace?

David

[ntinos]

I have looked in the manual and found the Specifications for Outdoor data where it says:

Frequency: 433MHz/868MHz

This is not very helpful !

But on the back of the plastic case of the base console where it says WH2080 it also says 433MHz.
So I must believe this I think.

There is a sticky thread on this forum called "Photos of the insides of Fine Offset sensors"
If you scroll down to pictures 7 and 9 of the main console PC board I can see a red wire about 5-10 cm long, is this the antenna you suggest I replace?

David

Υes is 433 Μhz
Yes this is RX module at pictures 7-9 and the red cable is antenna it is very small L/4 wave ,and is better replace it L/2 wave.
But this weather station is NG265,your is another type,if you like take a capture inside your console ,maybe your console circuit the RX module is onboard.

[AllyCat]

Hi,

I'm relectant to say much about improving the wireless reception (particularly of FO stations) because it's very much a "black art" which doesn't always behave as "theory" suggests.

However, the picture showing the "insides" of the Console was posted by a Canadian, so it probably uses the North American 915 MHz band, therefore the aerial (antenna) will indeed be less than half the length of a 434 MHz UK antenna.

The theoretical "optimum" length of a "monople" antenna is a quarter wavelength, not a half (which normally refers to dipole antennae - two monopoles joined at the centre), so the length "should" be about 17cms. Ideally it would be vertical rod (and a greater diameter than a flexible wire may help), but folding it within the Console might not make too much difference in practice. However, all this assumes that the input impedance of the receiver is reasonably correct (resistive), in practice it is often (intentionally) made inductive (e.g. with a "loading coil") so that an antenna shorter than a quarter-wave works best.

Also, bear in mind that any change that you make to the antenna (or the Console) is effectively changing the "position of the antenna in space", and what you may actually be doing is altering its precise position within a "standing wave" radiation field.

Cheers, Alan.

PS: Sorry, no definitive web links - IMHO even Wikipedia is in need of some authoritive/user-friendly editing.

[ntinos]

Ηi Alan
You are correct The theoretical "optimum" length of a "monople" antenna is a quarter wavelength, not a half.

But at check i have see for this frequency the console (receiver) with this small power mw external sensor,it have better results the monopole at half wave lenght at receiver.

Anyway David can try it ,and after he can tell us the results.

[DavidG]

Thank you for your continued interest. I have plucked up the courage to dismantle my console and hope I have successfully attached two pictures to this post for you to look at.

The first shows the rear of the circuit board and the way the antenna is mounted inside the console and the second the front side of the circuit board after I removed it from the case.

Careful measurement of the antenna wire when stretched out gives a length of 22.3cm which does not fit with the half wave requirement (34.64cm) or the quarter wave (17.32cm).

I might try removing it and replacing it with a different length but Alan's comments suggest to me that the console might have been 'tuned' to suit the length they have fitted. I can see a small 'coil' of three turns of copper wire next to the antenna attachment point. Is this the 'loading coil' you refer to?

In the light of what we can now see in the pictures do we feel I should:

a) Leave it alone
b) Shorten the antenna to 17.32cm
c) Lengthen it to 34.64cm

To go back to my original post, I should repeat that it worked successfully for 6 months before I started to get the communication problems. Do we have any thoughts about what might have suddenly changed?

David

IMG_1141.JPG

IMG_1142.JPG

[ntinos]

Hi David,thanks for the capture
Look,i suggest you make antenna cable one 34,6 cm
And another one 44,6 cm and you tried who is the better for you.

I know your problem,all this happen some time the trimmer capacitor at remote thermo/hygro sensor change little capacity and it is necessary to tuning again for more power better LC tuning .But it difficult because must you have
RF field strength meter ,try to change the antenna receiver if you want,you do not have lose anything if you make this modification.

And pass it like here all arround ,because at middle is the batteries.

[AllyCat]

Hi David,

IMHO you will probably waste a lot of time and are unlikely to get any consistent, repeatable results by trying various lenths of wire as an antenna. The one test I might try is a properly straight, vertical, monopole perhaps 15 cms long. Preferably a "rod" not just a wire because the greater diameter "softens" the criticality of the length/tuning. Yes, theoretically that coil is a "loading coil", but of such small size that I doubt if it makes much practical difference.

As for why your system "changed", well crystals (or probably SWIFs in this case) can slightly change their frequency characteristic over life (they literally shake themselves apart), and most electronic components can "drift" with age and temperature. However, another worthwhile test could be to switch off your new router for a few minutes whilst repeating the reception test.

Also beware that the reception of some models of the FO Console can be severely affected by interference radiated by the USB cable from the computer. Just this morning I was downloading data from a 3080 and for much of the time that a computer was connected (not just one but two different computers) then the dreaded "-- --" appeared on the Console, to vanish as soon as the cable was disconnected again.

Cheers, Alan.

[DavidG]

Sorry I have not replied sooner, been away from the computer.

I have replaced the antenna wire as ntinos suggested. I went for the longer option as that makes it easier to cut down later if necessary without re-soldering.
So I replaced the original 22.3cm with a 44.6cm wire taken from a CAT 5 network cable. I installed it exactly as suggested.

Initially I thought that I had fixed the problem. It quickly picked up a signel from the external sensors and carried on worked when I placed it in the original location in my office where I could connect it to the PC. But after about 15-20 minutes the connection failed and I got the "---" display again.

I don't understand why it never restores communication once it has failed.

So it looks like something has altered inside the console like Alan has suggested which makes it only able to communicate under perfect conditions - in my hallway in line of sight of the sensors.

I am worried that it might get even worse over time and end up without any communication at all - so - does anyone have a recommendation for a new weather station that I might buy that might be slightly better quality - but not very expensive?

I suppose I am thinking around £100 - £200.

David