Short indoor range with LoPy



  • Hi everyone!

    I am experimenting with several LoPys for (currently) indoor use. The current setup is one node acting as a LoRaMAC-Gateway and one as a LoRaMAC-node. They are both powered via USB.

    Hardware:
    LoPy (edit: corrected link)
    PyTrack
    Antenna (Connected to the slot nearest to the LoPys reset button)

    Parameters that have been tested are:

    sf = 7,12
    frequency = 864000000, 868100000
    coding_rate = LoRa.CODING_4_8, LoRa.CODING_4_5
    bandwidth = LoRa.BW_125KHZ 
    

    And tried with the parameter combinations mentioned in this forum post.

    No matter what I try the range seem to be less than 30 meters (Indoors in an open office environment, slight obstruction from line of sight). The range was slightly better with the LoPy-node in full LoRaWAN mode (with OTAA) with a multitech conduit acting as gateway, but still less than 50 meters.

    Is indoor range always this bad? Can this be fixed?

    Thanks in advance,
    J

    Edit: Solved! See description below.



  • Solved! Thanks again everyone.

    The problem was indeed that the antenna was mounted according to the specification for the LoPy1r, so it was actually on the LoRa 433-510 MHz antenna slot. After a quick range test I now receive packets at at least 400 meters distance.

    Many thanks, I thought I had gone crazy there for a while!



  • @crumble
    The hardware is indeed different (LoPy = SX1272, LoPy4 = SX1276) but the region settings that appeared in firmware 1.16.0.1 should work on both boards.



  • @jonasde said in Short indoor range with LoPy:

    Are both the LoPy4 modules defective, or am I missing something that I should have thought of? Is there anything that must be configured for the LoPy4 to be as effective as the LoPy1.0r seems to be by default?

    I can't find it right now, but pycom mentioned in this forum, that they use different LoRa hardware modules in LoPy and LoPy 4. Maybe it is related to this. The old modules are configure by hardware to your region. The new ones are configured by software. Have you set them all to the same region, so that they sending with the same strength?



  • @jonasde do you have access to the stats on the network side? Do you get similar results?

    Which firmware version are you using? I’m wondering if by any chance there couldn’t be an issue with the antenna port selection (433/868)? You did take into account the fact that the 868 antenna connector moved between LoPy 1 and 4?



  • @jmarcelino The U.FL plugs at the pigtails are not made for frequent use. After a few plugs, typicall the middle contact wears out. So better plug them in once and fix them, such that they do not move a lot when you use the device.



  • Thanks everyone for the fast response and good advice!

    On @robert-hh's suggestion I measured the RSSI and SNR using lora.stats().
    This revealed that both of the LoPy4 modules are receiving extremely low signal strengths. I have a LoPy1.0r module which I have tested with as well and it received orders of magnitude larger signal strength in the same positions and using the same gateways.
    To verify that the problem is isolated to the LoPy4 modules, I have tried the following:

    • Swap out the antenna, PyTrack, USB-cable, and case used with the LoPy1.0r with the corresponding ones used with both LoPy4's.
    • Updating all firmware to most recent versions, on all LoPys and expansion boards.

    Are both the LoPy4 modules defective, or am I missing something that I should have thought of? Is there anything that must be configured for the LoPy4 to be as effective as the LoPy1.0r seems to be by default?



  • @jcaron
    Before joining Pycom I was deploying many nodes in industrial-ish (hospital facilities) and we've had very good success with these antennas:

    https://www.rfsolutions.co.uk/antennas-c8/whip-antenna-90deg-hinged-868-mhz-sma-m-p150

    (they get a nice price break at 100+ units)

    Don't skimp on the pigtail cable either, we also found big variations. We ended up using these:

    http://www.mouser.co.uk/ProductDetail/Taoglas/CAB011/?qs=sGAEpiMZZMufBZYvsU/be/f4OCniACggye77MWZ4iIY=

    As bmarkus was saying we also covered entire hospitals (3Km+ radius including many buildings) with only one gateway on a tower. It wasn't even Kerlink, just a cheap RisingHF. Later we added a Laird gateway which worked mostly as a backup.

    It does pay to have a good antenna at the gateway. The professional choice seems to be these: http://procom.dk/products/531-cxl-900-3lw-nb868-mhz



  • @bmarkus Interesting. Just for reference, that’s in the EU region, right? Other regions have different max power so the results may be quite different.

    Also, even though the network side of things is not under our purview here, can you share details of the antennas you’ve had good/bad experiences with? I’m pretty sure others will appreciate the information!



  • @jcaron said in Short indoor range with LoPy:

    @bmarkus T
    Do you have a different (better) experience?

    I have no experience with LoPy in this environment, but e.g. we have a Kerlink IoT Station installed on the top of a 4-level office building part of a mall, 2 levels are higher then usual. No problem to receive trackers from garage in the basement with SF7 and practically from anywhere in the building.

    Customers reported they can cover large industrial areas, hypermarkets using single Kerlink ifemtocell indoor gateway, and so on. It expects a proper antenna. We have found expensive outdoor ISM antennas offering 3-4 dB gain but using them distance was very low and measuring them learned they were useless crap, not antenna. Replacing them with a professional antenna problem was gone.



  • @crumble lora.stats() tells you the rssi and snr of the last packed received.



  • @jcaron said in Short indoor range with LoPy:

    Would be curious to hear experiences of other people in the same situation.

    I had no problem with such a distance between two buildings with the antenna of the kit inside the buildings. You may have a lot of noise on this frequency or bad luck with reflections.

    An API to get signal strength and quality would be nice.



  • @bmarkus The gateways are provided by the network operator, we only operate the nodes. The one they provided me for testing is from Sagemcom. Not sure about those installed on site, not even sure they're all the same.

    Do you have a different (better) experience?



  • @jcaron said in Short indoor range with LoPy:

    My (very limited) experience indoors is that range is indeed not great. Whereas outdoors frames are received kilometres away, indoors we're observing about the same kind of range you see, though it varies a lot depending on the exact relative position of the gateway and node.

    We're currently at 4 gateways to cover an 8-floor 60 m x 170 m building, and there are still areas with very high packet loss.

    Which type of GW are you using?



  • My (very limited) experience indoors is that range is indeed not great. Whereas outdoors frames are received kilometres away, indoors we're observing about the same kind of range you see, though it varies a lot depending on the exact relative position of the gateway and node.

    We're currently at 4 gateways to cover an 8-floor 60 m x 170 m building, and there are still areas with very high packet loss.

    Would be curious to hear experiences of other people in the same situation.



  • @jonasde Which antennas do you use, and how are they assembled together with the board?



Pycom on Twitter