Project

General

Profile

500m link?

Added by tve over 5 years ago

I’d like to set up a 500m link between two JeeNodes, outdoors with direct line of sight. Has anyone done something like this? Recommendations? I suppose I should use the 432Mhz radios over the 915 ones? And switch to the lowest data rate? And perhaps nice quarter wave whips with ground plane? Or will I still be far off bridging 500m?


Replies (16)

RE: 500m link? - Added by lightbulb over 5 years ago

tve,

I have never done over 150m with two jeenodes L.O.S (because I never needed to), but if you have issues with 500m and if you don’t mind adding more jeenodes into mix, you can relay using the relay sketches.

I did need to get data from jeenodes over 600miles apart, with little infrastructure except at LAN connection at other end, so ended up using a small great little UART->RJ45 device with inbuilt protocol stack (less than 10\$ from HK). Puts your JN UART right on the network (with whatever security you wrap around it), I then used ‘socat’ to bring the i/o out on my local system to treat it as a serial device ‘/dev/remoteJN?’. Has been working flawless, even reporting directly back into HouseMon for over 3 months.

A friend of mine does 500m+ all the time, but uses xrf (and not on battery).

—lightbulb

500m link? - Added by martynj over 5 years ago

tve, you might find this long range working from the old forum useful.
The key is to keep the signal to noise ratio high enough in the bandwidth of interest i.e. lower the data rate so that you can lower the receiver bandwidth. Some older detail here and some ‘magic numbers’ from here onwards. 
Using a multi wavelength vertical above a good ground plane will help a lot - it is directional in the sense that sending energy off into the sky is reduced by shaping it more towards the horizontal.

RE: 500m link? - Added by dabbishaw over 5 years ago

lightbulb Do you have any more details on the UART->RJ45 adapter, one of those sounds very handy and a search on ebay failed me.

lightbulb wrote:
> tve,
>
> I have never done over 150m with two jeenodes L.O.S (because I never needed to), but if you have issues with 500m and if you don’t mind adding more jeenodes into mix, you can relay using the relay sketches.
>
> I did need to get data from jeenodes over 600miles apart, with little infrastructure except at LAN connection at other end, so ended up using a small great little UART->RJ45 device with inbuilt protocol stack (less than 10\$ from HK). Puts your JN UART right on the network (with whatever security you wrap around it), I then used ‘socat’ to bring the i/o out on my local system to treat it as a serial device ‘/dev/remoteJN?’. Has been working flawless, even reporting directly back into HouseMon for over 3 months.
>
> A friend of mine does 500m+ all the time, but uses xrf (and not on battery).
>
> —lightbulb

RE: 500m link? - Added by dabbishaw over 5 years ago

tve,

I tested a jeelink in the house plugged into my desktop and a simple jeenode running a counter every second with a battery attached. I then simply walked away from the house until the packets stopped. I got to just over 150 meters and roughly at this point the line of sight to the house disappeared. I suspect if you use the low bandwidth settings and you have true clear line of sight you should get close to the 500 meters without anything fancy.

I used this setting to drop the bandwidth.

rf12_control(0xC647); // 5 kbps bandwidth

Why don’t you do some tests and see how you get on.

I should also add that I built a jeenode with a whip antenna and for me it made no difference for broadcasting from trouble spots like under the hot tub.

tve wrote:
> I’d like to set up a 500m link between two JeeNodes, outdoors with direct line of sight. Has anyone done something like this? Recommendations? I suppose I should use the 432Mhz radios over the 915 ones? And switch to the lowest data rate? And perhaps nice quarter wave whips with ground plane? Or will I still be far off bridging 500m?

RE: 500m link? - Added by tve over 5 years ago

dabbishaw, which freq band were you using for your 150m test?

lightbulb, thanks for the relay suggestion, but I’m trying to bridge a 100m deep canyon between my house and my mailbox, so relay nodes are not practical. I might as well string a cat-5 across the canyon ;-).

martynj, thanks for the links, I’ll go on reading..

RE: 500m link? - Added by martynj over 5 years ago

tve, I can see the usefulness of saving a wasted trip to the mailbox ;-)
With such a clear line of sight and reduced data rate, it should be achievable.  There may well be a problem when it rains though.

RE: 500m link? - Added by lightbulb over 5 years ago

dabbishaw - http://www.hlktech.net/product\_detail.php?ProId=39

tve, I’ll lend you my dog, he fetches the mail (even in the rain) - you may even get a hunk of the posties’ leg to boot ;)

RE: 500m link? - Added by wiso over 5 years ago

I was going to post the results from a UAV flight but I see that martynj already beat me to it!

RE: 500m link? - Added by dabbishaw over 5 years ago

tve I’m using 868mhz.

500 meters to the mailbox, oh boy I need to move!

tve wrote:
> `dabbishaw, which freq band were you using for your 150m test?

`lightbulb, thanks for the relay suggestion, but I’m trying to bridge a 100m deep canyon between my house and my mailbox, so relay nodes are not practical. I might as well string a cat-5 across the canyon ;-).
>
> martynj, thanks for the links, I’ll go on reading..

RE: 500m link? - Added by padvinder95 over 5 years ago

I think you should easily get to 500m if you lower the baud rate and make the antenna a tad more directional than the standard wire. Maybe a small dish reflector (http://en.wikipedia.org/wiki/WokFi) or a waveguide antenna (http://en.wikipedia.org/wiki/Cantenna) (but don’t use Pringles, rather a food tin) would be best?

RE: 500m link? - Added by tve over 5 years ago

Well, I did some preliminary tests today. One node with a small dipole and at the other end (500m away) two nodes with a monopole with 4 angled-down radials (as depicted in http://jeelabs.net/boards/6/topics/2152?r=2158\#message-2158) and a commercial 10 dBi yagi. I sent 4-byte messages back and forth at 19kbps. Well… the yagi resulted in relatively frequent successful exchanges but the monopole only occasionally completed a ping-pong.
Looks like with some additional antenna tuning and perhaps a further speed reduction it should be possible to make a link work as long as I’m not sending long messages and retransmit a lot.

RE: 500m link? - Added by martynj over 5 years ago

tve some encouraging results. Remember that the benefit from speed reduction is linked to allowing a reduction in receiver bandwidth - their interaction is complicated, it is easiest to stick with the ‘magic numbers’ referenced above. Did the implied aerial polarisation match in the two test cases?

RE: 500m link? - Added by tve over 5 years ago

I’m using the magic numbers, although below 19kbps there is no further reduction in bandwidth, if I remember correctly. I’m using:
#define RF12RATE 0xC611 // 19.1kbps
#define RF12BW 0x94C1 // VDI:fast,–97dBm,67khz
#define RF12DEVHI 0x9820 // 45khz, full tx power

I did ensure that the polarization matches. More tests will be necessary when I have a bit more time to play.

RE: 500m link? - Added by martynj over 5 years ago

tve, it is possible to squeeze more signal/noise ratio by dropping the baudrate/deviation/receiver bandwidth further, but it usually requires trimming the crystals to match better. As the rx bandwidth reduces, it is more demanding on the transmit side falling into the slot for the AFC to pick up anything to align to.

RE: 500m link? - Added by jwp over 5 years ago

You could also pick up an RFM12BP (P for power amplifier) to replace the RFM12B. It would easily bridge that 500m. It’s not exactly pin and size compatible, but with some moderate soldering skills it’s fairly easy (all RFM12B pads are available through-hole just left of the RFM12B). You’ll also have to sacrifice two ports because you need 2 digital pins to control the power amp and a small modification to the RFM12 driver is required to control the amp on RX/TX.

RE: 500m link? - Added by martynj over 5 years ago

Hmm - yes and no. The RFM12BP will certainly boost the received S/N ratio by increasing the transmitter field strength, but if you want reliable ACK’s, then both ends of the link need to use the higher powered module. Making the signal directional by a suitable improvement to the simple omni-directional quarter wavelength stub antenna can achieve a similar link performance without the power consumption penalty of higher powered transmitters.

As usual - many ways to skin the same cat.

    (1-16/16)