Hello

Summary:

Using the T110 over total ranges of 20 m and 30 m, predominantly outdoors, gave me poor (unreliable) or no connectivity to a H100 hub which I tried in a couple of positions and found no useful improvement.

Measurement method: see if my phone/tablet showed the message using the Tapo v3 app  (tablet LAN connected using WiFi, or 4G phone for more distant tests at the shed and garage). An assistant was monitoring the hub for audio sounds ("open" and "closed" MP3s) and confirming whether these alerts were heard or not heard at each time.  Finally, checking the H100 logs to see what had happened (and what had not).

Detail:

The H100 hub was tried  in a downstairs room and then an upstairs room, 20 m distant from a garden gate location, all in a rural setting.  The radio path was line of sight apart from one external house wall: no foliage in the way, no ground in the way. The T110 was placed approx 1.2 m above ground on a dry brick wall and the hub was plugged directly into a mains socket such that the first attempt (through that one wall) was at about the same height and the second attempt (again, through the same wall) was about 2 m higher (an upstairs room).

Initial setup of the T110 and H100 hub was painless and rapid. Setting up a couple of rules (play this sound when the contact sensor opens, but play this other sound when it closes) was easy.  Several tests in the house showed very rapid detection and messaging, with delays always shorter than 2 s and probably shorter than 1s most of the time.  It seemed promising and it worked through a couple of indoor walls but I didn't pay attention to these tests due to the goal being an external location. For completeness, the house external wall structure consists of a single skin of brick, a cavity filled with insulating foam, and then a breeze block internal wall, and then plastered. I expect that the plaster render was applied to the blockwork (original 1950s construction). It is not likely to be (e.g. foil backed) plasterboard.  In any case I'd still expect some knife-edge diffraction round the windows in that wall to illuminate the H100 in the room.

I then moved the T110 to the aforementioned outdoor position, on a wall next to the intended garden gate installation location. The "open" command was immediately seen but "closed" didn't turn up. Numerous (20+) tests at this position  showed that the "opens" were almost always detected whereas the close" signals were intermittent.  Standing in the way of the radio path caused the opens to drop out. Not the sort of link margin I had expected.

I tried a different location - a garage the other side of the house - which was also 20 m away but the path on this occasion would have to encompass one outside wall and two inside walls. Almost no joy at all for open or closed: I think I got one or two opens and one or two different closeds out of a good dozen test pairs but I can't remember the exact detail now.

I tested again inside the house (all OK) so I moved the hub to an upstairs room; then I went back to the gate and re-tested, and the garage and retested, and found the same lacking performance.  I tried our shed which is  30 m away but through a different aspect of walls and also failed to detect any "opens" or "closeds".

The chat session I had with support prior to purchasing told me that the expected range should be up to 500 feet according to internal (and as yet unpublished) TP-Link documentation.  The tests I carried out showed my T110 and H100 unable to connect at distances significantly shorter than 500 feet (20 m is approx 70 feet).  I would have expected the link margin for a narrowband (FSK?) signal to be many tens of dB in hand over the short ranges I was trying: the building insertion loss is unlikely to be higher than 10 dB. I provided feedback via the app and I then immediately returned the items for a refund.  I can't say whether the T110 was v1 or a later hardware version nor can I say what the f/w version was, because I returned the whole lot the same day (a week ago now).

Please can TP-Link publish the documents, which support referred to, so that we can see the sorts of performance we might expect over longer distances?