QUESTION

Question

I am a student developing a tracking/telemetry board for a club project.  I am trying to use the M20048-1 RADIONOVA module with the internal antenna for a GPS fix.  We produced a first rev of the board on which everything except the GPS was able to be debugged and fixed.  In the course of testing we only obtained a fix once, even in ideal tests in large open fields.  Admittedly, the first revision did not really respect the layout guidelines as it should have (old layout at end).  Having done a bunch of research I should have the first time, I have completed a rev 2 layout, but I'd like a bit of extra confidence that the new layout will address most of the issues we had the first time.  The current rev 2 layout is shown below (GPS at bottom).  The board is two layers with a ground pour on the top and bottom layers (not shown for clarity).  I have included the dimensions I thought might be important.

For comparison, the old design (GPS upper left).  The line indicates the portion of board which was Dremeled away in an attempt to actually have it on the 'edge' of the board (the power trace was rerouted through a jumper).  Even after that fix we were unable to obtain a fix:

Answer 1

Hello,
Firstly, ensure you have a ground pour void underneath the antenna section of the M20048-1. See Page 14 of the Product Specification document (Internal Antenna Clearance).

“The M20048-1 module internal antenna requires a small clearance on the host PCB to

operate. The clearance means that no Ground or tracks of any kind can be within this area. This

must also be clear through the entire PCB stack up. The minimum area needed clear is 6mm x

4.87mm.”

Additionally, Page 17 Module Placement Guidelines: Follow these closely.

1. Better to place Module along to longer side of the PCB (currently you have placed it on the shorter)

This should be enough to get a fix

Kind regards,

Paul

Comments

I do have the void (looks like I accidentally had the keepout layer hidden in the picture).  I could move it to the left side but as I only have 2 layers, I'm not sure I would be able to maintain as continuous a ground plane on either layer with the various signals that will have to run up and down the board.  Due to integration concerns I can't easily extend the board widthwise.  I guess my question is whether a solid contiguous ground plane or a wider edge is a more important concern.

---

Hi

 

Firstly, I think you can most definitely arrange the PCB layout to have the M20048 on one longer edge and all other components along the other long edge half of the PCB leaving a long clean ground plane along the M20048 side. You should add additional stitching throughout the board to help RF with the 2-layer design.

Other than this, you may need to get the M20048 internal antenna tuned. However, this is quite complex and may be out of scope for your club project. Firstly, you need to properly put a UFL connector between pins 4 and 6 and use a VNA to record the information required to send to Antonova to provide you with the AT1 and AT2 component types/values. If you’re at a university, there will be someone there with knowledge of how to do this and the equipment required.

 

Secondly, if you don’t have the capacity to performance antenna tuning, you could look using separate RF MODULE and antenna, I have used a SR4G013 antenna on a similar sized PCB and achieved good results prior to tuning. It would require you to use a standalone GNSS RF MODULE, Antonova has a few options. However, I have found although this antenna works quite well without tuning, in a tight enclosure you will need antenna tuning to achieve high performance.

 

Kind regards,

Paul

---

I followed your advice and this is the new layout (the new positioning actually worked out way better than I thought it would).  The faint rectangles are voids.  The GPS one goes though both layers, while the upper one is an SD slot and it only on the top layer (and will have the grounded case of the holder over it too).

Answer 2

Hi,

I used the M20048 for a LoRaWAN tracking device last year and initially had similar issues, getting an initial fix was proving a challenge. Following some testing I tracked it down to antenna tuning (involves changing two passive components L1 and R8 on your design). After contacting antenova, they provided a tuning service that I took advantage of (It was free, it may still be) where I sent off a finished device (including the enclosure is important as it affects the signal) and they tuned the antenna for me, sending back a very professional presentation showing how the tests were carried out and the results and recommendations. To put it in perspective the default components are a small inductor and a 0R resistor I believe, I ended up with 10pf and 8.2pf capacitors. This made all the difference and resulted in <45s cold start first fix virtually every time. I tested the units in the Brecon Beacons last November and they performed extremely well. So I would recommend that you contact antenova and see if they can help.

Matt