Sandaysoft

After struggling with "Lost Sensor Contact" errors every few hours, I bit the bullet and opened up the transmitter unit in order to examine the antenna design used, as my gut feeling is the errors might be caused by poor gain and/or interference from all the other wireless gadgets, including a LaCrosse 2355 weatherstation.

Here are pictures of its internals.
The question remains as to how this loop antenna can be made more efficient without messing up its tuning.
(Pity they didn't use screws to hold the case together, but then I didn't really expect to make any successful warranty claims)

_________________
Graeme.
http://weather.gktnet.com/cumulus/index.htm

Quite a bit different from the WH1081 transmitter. Maybe they've made some improvements

_________________
Gina

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

hi guys.

the most obvious way to improve gain would be an external aerial.

By running a coaxial cable from inside to a 'stub' outside the gain could be increased greatly.

A simple and very cheap 'co-ax /co-linear' aerial can be made from coaxial cable. See this article ...

http://www.rason.org/Projects/collant/collant.htm

Other aerial types could be 'cantenna' , dipole, beam, etc.

I have seen a neat idea for 'free gain' using a cardboard/foil 'reflector' sited behind and around a monopole aerial as used on wireless routers .....

http://www.binarywolf.com/249/diy-parab ... lector.htm

Neat !

Sounds like a 'blue peter' moment

_________________
Phil - G0DOR

That's cute Phil Thank you. If you've got a big enough Stevenson screen box you could put the transmitter unit at the focus and get 3dB gain Bring the aerial wire out of the console and at the focus of another parabolic reflector and get another 3dB extra signal. On the receiver, it also gives more directionality, reducing possibility of interference pickup. I think 3dB might be a bit optimistic but there should certainly be some improvement, I think.

_________________
Gina

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

These links are worth a look
http://www.rfm.com/corp/appdata/antenna.pdf
http://www.hoperf.com/upfile/ANTENNAS_0102.pdf

_________________
Roy
CWOP: DW6272
WUnderground
RadstockWeather on Twitter
Met Office Wow Site

The are a few problems to overcome in fitting an external antenna that I have yet to overcome.

The signal levels are very low and so attenuation in the coaxial link 'twixt transmitter and antenna undoes any gain theoretically involved (I even tried ultra low loss coaxial cable, which is about 2cm in diameter!)

SWR. I touched on this in another post and although at these power levels there is very little chance of damaging the transmitter, any mismatch will again attenuate the signal.

Circuit board connection point for antenna. You cannot just solder on to the existing loop if you want an efficient system. The existing loop must be isolated (ie the track cut on the circuit board) and the coax connected directly to the output of the transmitter. Easy on a big old amateur radio set... infinitely fiddly on an all in one ic!

Putting a parabolic reflector around the transmitter would give some small gain (unlikely to reach 3db though). Unfortunately, the last thing that you want near the temperature sensor is a mirror focused on it!

Other options that I've considered are a small radio repeater that could retransmit with increased (and therefore illegal) power or actually removing and then hardwiring the unit into its own container and relocating the transmitter itself to a higher, closer position.

I would love to find some way of extending range as I live halfway up a mountain and want to situate an anemometer around six hundred feet above on the summit... it only gusts to 100mph down here in the "valley"

_________________

This looks interesting
http://www.maxim-ic.com/datasheet/index.mvp/id/6050

_________________
Roy
CWOP: DW6272
WUnderground
RadstockWeather on Twitter
Met Office Wow Site

Actually you can get significantly more gain with a parabolic reflector than 3 dB, and unless you are aiming at (or near) the sun, it won't impact temperature readings. If you can find an old DBS TV dish, mount the transmitter unit where the LNB is, and then aim at your house, then, depending on the frequency of your unit, and the size of the dish, expect 20 dB or more gain, or about 10X the range. This is the simplest and best way for you to solve your problem.


The law specifies effective radiated power, so adding antennae, parabolic reflectors, etc are all just as likely to make you over the limit as is a small repeater. Again, focusing the energy on the target (your house) will significantly reduce the likelyhood of you interfering with another service, which is the actual intent of the law. Use discretion - if 100 mW will solve the issue, don't use a KW! In the worst case, someone will tell you to remove it.

Having said that, an old DBS dish is the way I'd go first.

Hi,

Apologies for a late response but I've only just registered with this excellent forum.

IMHO, timfy is correct that connecting and matching an external cable/antenna to an existing PCB (with integral antenna) is far from easy. But I don't believe that attenuation in a (short) cable is a significant factor. However, an unmatched (e.g. unterminated) short cable can actually make an excellent notch filter, because a reflected, unattenuated signal can null out the source at certain frequencies. So if you have a "problem" cable, try changing its length by a quarter-wavelength (noting that the speed of the wave in the cable may be about 60% C, so try cutting off a 15% lambda length).

Also, Charlie is correct that making almost any change to a transmitter (including just adding a reflector) can potentially make it "illegal". However, I don't believe that a parabolic reflector "dish" will give a useful improvement anyway. It depends on the size of the dish and the ISM frequency, which might be 434, 868 or 915? MHz for some UK, European and USA systems respectively. A dish might be useful at the license-exempt microwave frequencies of 2.4GHz (Bluetooth/WiFi) and 5GHz.

A two foot diameter dish (which is quite large by modern DSAT standards) has a diameter similar to the wavelength at 434MHz so the theoretical beam width might be about 70 degrees, with just a few dBs gain. At this frequency the "focus" is probably irrelevant, the dish will effectively act as a flat plate reflector which needs to be a quarter-wavelength behind the antenna and might give about 3dB gain (by "folding" the rear lobe onto the front lobe of the radiation pattern).

Personally, I would try a TV-type yagi antenna for the "main link", a small loop antenna perhaps a few feet from the existing transmitter/receiver and a low gain amplifier (perhaps as much to sort out impedance matching issues as to add gain). This might be either a "masthead" type (for low noise) or a "distribution/splitter" amplifier (for ease of power supply).

At 434MHz a normal UHF TV yagi (group A in the UK) should work fine, a "contract" type giving perhaps 10dB or a "fringe" type with balun, mesh reflector and numerous fancy directors etc., perhaps 20dB. 868MHz and above might be more tricky because yagis (and most dipole antennas) have an almost "cliff edge" response above their design frequency.

Interfacing a simple loop antenna to an existing station can probably only be done by trial and error. "Low frequency" theory indicates that the loop should receive "edge-on" but considering it as a folded dipole, it would receive "face on". That's probably why they are generally considered as omnidirectional.

You might first try such a "repeater" arrangement at the receiving end (which would be "legal" and easy to power, but least likely to work well) then at the transmitting end and finally at both ends if all else fails.

Cheers, Alan.

Hi guys.

am not really sure why some of you folk are sceptical about the gain offered by reflectors.

If you had looked at the article quoted you would have seen that someone has already done the calculations for the job.

The 'cheap and cheerful' paper/foil reflector has been well documented for wireless router work with outstanding signal path improvements 'quoted' !

The frequency is irrelevant as the device is made to order.

Am not sure what the output power of the TX @ 860 MHz is but my guess would be < 100 microwatts, so 3 dB gain would be well under the UK SRD limit of 10 mW

Interestingly, i have not been able to detect any RF @ 860 MHz using a colinear (2/70) @ 10 m agl < 10 m from the transmitter

_________________
Phil - G0DOR

What people were saying is that to be effective, a parabolic reflector needs to be several wavelengths across. There is a big difference between the 2.4GHz used for wireless networks and the 434/868 MHz used by these weather stations. Are you sure the calculations were done for these lower frequencies? I agree, a small reflector will direct energy back in the right direction - as will a linear reflector.

_________________
Gina

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

Hi,

The designer appears to claim only 3dB improvement whereas Charlie talked of (and timfy may need) +20dB. I don't know how the 3dB figure was obtained but it actually seems rather low, perhaps because the designer appears to have used a very unfortunate spacing of a half-wavelength to the reflector. If I recall my antenna theory correctly the reflector should be spaced at a quarter wavelength (because phase inversion occurs on reflection).


<10m might be too close to use such a distributed element antenna, or perhaps the transmitter is working at 434MHz which the majority of FO devices sold in the UK appear to use?

Cheers, Alan.

Yes sorry, I keep forgetting that there are other frequencies used by these units. My unit here in Canada is 915 MHz. Our old DBS dishes come in a variety of sizes too, with older being larger. So some real examples:
24" dish, 915 MHz = 19.25 dB
18" dish, 432 MHz = 10.23 dB
These are of course calculated, not measured. YMMV

So, if you want to significantly extend the range, either of these will do nicely. Aim carefully, and hopefully there is nobody that you might interfere with in the line beyond your receiver. Shooting down a hill to a lone house sounds ideal and shouldn't bother anybody.

The tinfoil on cardboard parabola slice will provide some gain, but it's a bit tricky to calculate, and then there's construction acuracy... the focal point needs to be just right... not for the feignt of heart. Hence the (2.4 GHz) reports of "works great" and "does nothing" and "actually made things worse". All are likely true.

just an fyi- I opened my new 3080 base unit, and the antenna is an insulated wire in an "L" shape, 225mm long.
I've straightened the wire and led it out through one of the hanging holes at the back; reception seems improved.