Remember in the UK there is no such thing as 'unlicensed' spectrum, it's all license exempt - which means that specific Acts of Parliament (Statutory Instruments) have been passed to allow use of specific bands without a specific license and as long as you stick to the conditions in the SI you don't need a specific license.
Comms Act has obligations for anyone running a communications network, just more if a Public communications network.
Thomas,
Thank you very much for the reply.
Indeed the channel configuration was the primary issue. I used "mac set ch status X off" for all channels except 48,49,50,51 which are configured in the MultiTech Conduit's global_conf.json file.
As for the RX2 frequency, I have not been able to test it, but the LoRaWAN Spec V1.0 gives 923.3MHz at DR8 as the default RX2 window for USA (see paragraph 7.2.7).
Still working on getting data to show up in the API, but your pointers were of great help. Thank you.
-Chris
Melbourne`s TTN community discussion
If anyone wants to join our community, please join here at the forum first, so we can start the discussion about the planned tasks and ongoing activities.
Thanks 
Looks nice and easy this way.
1. I miss the "mac get status" call. I use that allways before a transmit to check if the devices is ready.
2. From the example, ttu.sendString() returns true or false , not the results you really need, for instance "Rejoin needed"
You could put the return message in a status field, or handle the session control automagicaly and build that into the class
3. also "mac get deveui" could be very usefull, since you need it for OTAA
Hey @Yesway, if you could email me (community@thethingsnetwork.org) the picture, I could update the Lincoln page.
anyone familiar with this instrument ? ' Houston , do we have a problem ? '
(or is this normal behavior ?)
And note that things might change soon: New design community page.
I just ran into a nice explanation of 433 MHz and 868 MHz Antenna Design Examples. It sounds good, but I don't know if all details are correct. Also, I don't know if the given coil diameter is for best reception, or is just a number that, combined with the number of turns, gets one the 1/4 wave length (8.6 cm for 868 MHz). (In the latter case, I guess I'd just mark that length before starting to wind the wire.)
For most applications, where space constraints exist, a simple Helical Antenna can be used. The Helical Antenna is a simple construction made from wound solid copper wire.
Some technical data for common frequencies:
433 MHZ:
- Turns = 17, Coil Diam = 5.5mm, Wire Diam = 0.5mm, Coil length = 22mm
- The groundplane stub between last turn and solder stub = 7mm
868 MHz:
- Turns = 9, Coil Diam = 5.5mm, Wire Diam = 0.8mm, Coil length = 13mm
- The groundplane stub between last turn and solder stub = 4.5 mm
Here is how to make such a helical antenna.
The material of the coil can be either solid copper wire or silver plated copper wire. The solid copper wire when pulled strongly gets a nice tension.
To pull the copper wire you start using about two meter of 1mm solid copper wire which is fixed to a sturdy anchor point. Take a piece of wood and wind the other end to the wood. Now start pulling with both hands on the wood side very slowly, until you notice the tension in the copper wire is getting loose - now pull it 30-40% in length with increased speed. This results in a very strong copper wire with a much higher tension, stiffness and perfect straightness with a diameter around 0.5 - 0.8mm.
Now take a rod like pencil or a drill with 5mm diameter and wind the wire around the rod as many time as mentioned above in the drawing according to the desired frequency.
Take care to create the groundplane stub on the lower end. It requires you to bend the wire in the axis away from the coil until it makes a small knee. Now add 3mm of wire by bending it 90 degree to the groundplane stub and the antenna is finished.
Thanks !
This looks like a very cost-effective solution !
Did you just plug a single antenna out of the iC880A or did you use a powered RRH to increase signal power ?
I see that the iC880A advertise a 15km LoS range with a +20dBm output. Did you check the actual range of the cell in the parisian urban environment ?
Is the gateway located inside and/or outside ?
Sorry for all the questions but I really would like to know more about your experience.
By the way I have submitted the idea of a parisian LoRa network to the budget participatif and it is somehow getting some traction.
Would you be interested in participating in this project and bring in your experience ?
If you are, could we exchange through paris.lora.network@gmail.com to have a more direct conversation ?
Cheers
This is my node running TXfastViaTTN.ino on a Arduino Pro mini and a Microchip RN2483, powered by a 9V battery.
Thanks for sharing!
Hey Almarion,
I am interested. 30+ years experience in RF (military, AM/FM radio stations, HAM). I have a large building contractor who is interested in helping/sponsoring us.
Nice Marco,
I ordered the holiday cards !
How does the range of such SX1276-based gateway compare to "true" gateways?
(Lacking any TTN in my area, I'm trying to figure out if the bad range I get is caused by my node or my single channel gateway. I tried many spreading factors, all on channel 0, and both gateway and node are using a HopeRF RFM95W with a solid 8.6 cm wire soldered on the board as an antenna, with no explicit ground plane—which might be bad? All fine when in close range, but the maximum range is some 200 to 250 meters... But maybe that's all to be expected?)
Just got a single chan gateway up and running (801F02FFFF7E56D1) with one node (02E00201) sending "Hello, world!" for a few test packets then turned off.
Node is a mega2650 w/ a Modtronix inair9b and gateway is a RPi model B rev 2 with another inair9b and the single_chan_pkt_fwd from github. Node will soon have a DHT22 temp/% rel humidity sensor for real data. Eventually hope to collect water quality readings from the nearby major river.
Lessons learned - The damnable RPi wiringPi pins (Actually love the RPi 
took a while to sort out but ended up using:
// SX1272 - Raspberry connections
int ssPin = 10; // pin24 , SPI_CE0_N is GPIO08 and '10' in gpio readall wPi column
int dio0 = 7; // pin7, GPIO07, 7 in wPi column, also labled GPIO04 in other pinouts
int RST = 0; // pin11, GPIO00, 0 in wPi column, also labled GPIO17 in other pinouts
and had to change 'eth0' to 'wlan0' in main.cpp and also tcpdump used to verify traffic off to thethingsnetwork. Anyway, getting the proper gpio pins worked out (using the gpio readall and testing with a scope - or an led/limit resistor would work also) was the breakthru in my case.
Hmmm, I did not change that, and uploading to TTN just works fine on WLAN. But I did have an ethernet cable connected at some point in time; maybe that matters.
Hi Arjen,
This depends very much on where the gateway (-antenna) and node are located. Although line of sIght (LoS) is not required for LoRa it does help a lot. My 1ch gateway also using 8,6 cm antenna is placed inside in the attic (about 6 meters above ground level) in a "city area", also range isn't that good but I do get up to 2 km when also the node is high (like on a bridge). I now also have an iMST iC800A gateway at the same location with a better antenna. This performs "somewhat" better but not dramatically. I did not get a greater maximum range, but do get little further in some area's and reception is more reliable in close range (I have put my nodes on 1 channel for comparing this). This limited range is not a limitation of the iMST board, I know it can perform much better with a proper antenna on a good location,
I also see there is some "directional preference" for some area's (this probably has to do with buildings nearby). My RFM95 based node with wire antenna gets up to 17 km range when talking to a Kerlink gateway (in "almost" LoS conditions).
I plan to get an antenna for the gateway up outside in the next weeks and hope to get some better range.
Yesway you are right, all equipment operating on less then 50VDC or 75VAC are exempt from testing according EN60950 (safety). However the R&TTE directive clearly states that as soon as there is wireless communication is the product, the product should be tested to EN60950 too, even if voltages are less then 50VDC/75 VAC (sorry to disapoint you, but a warned man is worth twice) 
