search

TheThingsIndustries / generic-node-se

Generic Node Sensor Edition
https://www.genericnode.com
Other
110 stars 31 forks source link

Potential ANT performance issue #89

Closed azerimaker closed 3 years ago

azerimaker commented 3 years ago

Summary:

In their first layout review Fractus has mentioned to place the antenna on the back side of the mainboard. As can be seen from the review, we followed all their design recommendations, but one. To simplify the manufacturing and bring more prominence for the component I've placed it on the front side of the board. Now that the component is slightly closer to battery contacts, which might result in slightly degraded RF performance and Fractus is a bit concerned about this. ant1

Next step:

Once we have the first prototypes, I'll do in-depth testing to compare the performance penalty. If the performance is severely degraded, I'll place the antenna on the back side (as shown below). ant3 ant-2

elsalahy commented 3 years ago

@azerimaker The new node (v0.2) joins the LoRaWAN network via OTAA and it seems functional which is great news. but the RSSI doesn’t look good (-91), so we need to check a couple more nodes to see if this persists in the other nodes. image

azerimaker commented 3 years ago

One definite difference is the missing enclosure, because the matching network was selected with the enclosure taken into account. So putting the device inside the enclosure will definitely affect the RSSI value, but nonetheless device needs to go through antenna tuning process provided by our partner.

azerimaker commented 3 years ago

I did a crude antenna return loss (efficiency) and VSWR measurements with my NanoVNA. Without any matching network tuning antenna works best for 938MHz with VSWR value of ~1.73 inside the fully assembled enclosure. Without the enclosure frequency shifts to ~960MHz with VSWR value of ~1.04. The antenna clearly performs better without the enclosure, but once it goes through matching network retuning with the enclosure we can get good performance out of it.

Test with the enclosure: nano-vna-test

Test without the enclosure: nano-vna-wo-enclosure

elsalahy commented 3 years ago

This issue is part of Q2, do we have any progress on this?

azerimaker commented 3 years ago

As suspected overlapping battery terminals on the ANT clearance area does slightly reduce the efficiency, but after we've gone through the matching network retuning on the actual device it became evident that the real device performs better than the simulated results. Please see #128 for more info.