Improving RF signal strength emonTH emonTX

Further improvements…

I made 2 ground planes earlier from 165mm diameter pieces of cardboard, and stuck aluminium foil to the uppermost surface (see image below).

I sat the emonTX on one (instead of the sheet of foil I tried yesterday), and also sat the Raspberry Pi base station on the other (no earth connection) and immediately saw yet further signal strength improvements, despite both units remaining in exactly same locations.
Measuring over a 2hr period, I now have;

Mean Min Max Diff Dev npoints
-66.1 -72.0 -54.0 18.0 3.0 720

Which seems a big difference from the -81.8db mean average, which I had 2 days earlier.

Any particular way of attaching it to the PCB?

No, just place the Pi on top of the disk so the antenna is centered in the circle - there are no electrical connections to the foil.

NOTE; I made it 165mm diameter because I understand that it’s radius should equal the antenna length. Which in my case it was 82mm (868MHz).
I assume that it would need to be 330mm diameter for a 433MHz system which has a 165mm antenna.

I wonder if sitting it on a metal shelf would achieve the same outcome?
I think I’ll give that a try when I get home tonight!

It should do.

@Bill.Thomson contacted me earlier today, and sent me the below video that he took several weeks ago regarding the ground plane effect.

To set the scene, Bill’s emonTx is located about 75 feet from where the emonBase in the video clip is, and there are three walls between the two. Watch what happens to the RFM12 LED when he removes & replaces the aluminium foil from the emonTX…


I was thinking more of the connection to the PCB for the antenna. Straight down or a slight bend on the bottom of the wire and if so, does the bent part get included in the 82mm length?

I also wonder about putting a hole in the ground plane and putting the antenna through the hole and sitting it on the case.

The wire is completely straight & vertical. Distance from PCB to tip of antenna = 82mm (for 868MHz)

No, the complete antenna must be above the ground plane.


1 Like

Ah, and for 433MHz?

Ahh, you mean for the emonTX.

My emonTX is in a bespoke case, which enables me to have the antenna vertical, however with a emonTX in a openenergymonitor case, it’s problematic to use a ground plane, unless you drill a small hole in the case, pass the antenna through, and lay the emonTX flat.

I think that I gave details for the 433MHz antenna in a post above.

Take a look at the Sparkfun website.

You can add a length of 50 Ω co-ax and terminate that at the ground plane, then the antenna length is ¼ λ from the ground plane.

Oooh, I never even considered a dipole Robert…


Ah yes, Sorry missed that for 433MHz. Yes I have a TX but I was thinking more of the base station RFMPi board. This is currently mounted on a Orange Pi Zero so a fixed antenna would not be an issue. I don’t think I need to go to a ground plane as well.

The ground plane will help at both transmitting and receiving ends of the link.

Hi Robert, yes I get that. I just do not need that much improvement. The solid antenna on the receiving end should be just what I need. Also it is currently mounted on a wall which makes a ground plane a little tricky!

Adding a second wire of the same length to a gnd track near the existing antenna wire would work really well as a dipole. I had massive gains when I played with this on a JeeLink a couple of years back, a few days ago I went to add a link to a previous thread here as I had posted a picture back then but the pic’s no longer attached (“weak signal?” on old forum), but I’ve just found a local copy.

It’s just simple single solid core bread-boarding link wire, I have since rolled this method out on most of my emonTH’s, and rfm2pi’s with great results. plus the wires are thin enough to be discrete, especially when the wires can also be color coded to their surroundings too.

I’ve also discovered that the antenna wire and ground plane wire are not that fussy about positioning, I have one rfm2pi on a Pi in a small case and both wires are coiled inside, one above and one below the board and even that gives better reception than the original multi-core antenna wire without a ground plane.

I’d read about ‘coiling’ antennas elsewhere, but dismissed it as I couldn’t reconcile the supporting theory…
…but it’s many years since I studied RF, and I’ve forgot more than I learnt!


Oh no, I wasn’t suggesting I had done anything technically advanced, quite the opposite, I was just saying l’ve “tucked the wires away” rather than having the antenna and ground wires fully extended in opposing directions, and still getting good results.


The only “gotcha” here is a 5/8th wave antenna needs a matching section at the feedpoint. e.g. a tapped coil.

Found some video clips I’d forgotten I made.
The ground plane doesn’t necessarily need to be metallic, just conductive.
As can be seen in the first video clip below, an anti-static bag works quite well.

In this clip, we see that it works even if the ground plane is simply adjacent to the Pi vice underneath it.