Electronics > RF, Microwave, Ham Radio

Antenna selection on generic GNSS modules (Ublox etc.)

(1/3) > >>

I have a number of GPS etc. modules which I'm using for experimental purposes, with (genuine) Ublox neo-xx cans onboard. In general I select these to have provision for an external antenna, very often they support both an external and an onboard antenna.

I've seen various mutterings that a capacitor should be removed so that there is only a single RF source. Can anybody provide definitive or at least informed information on this?

In more detail, the attached photo shows part of a GPS module with SMA and IPEX connectors paralleled at the bottom and left, and the centre connection of the onboard (ceramic patch) antenna towards the top right.

The external antenna is connected to C8 and L1 in parallel, with C8 connected to U3 which is an unmarked six-pin device.

The internal antenna is connected to C4 and L2 in series, hence to U3.

The fragmentary instructions I've seen suggest that C8 should be removed if the external antenna is being used... which is of course rubbish since that's the in-path device. However the seller who advised that must have got the notion from somewhere.

If U3 is a passive combiner, e.g. from Mini-Circuits, then is is reasonable that an unused RF source should be isolated since an out-of-phase signal would mess things up.

If U3 is an active switch, e.g. from Peregrine as mentioned in https://www.u-blox.com/sites/default/files/products/documents/GNSS-Antennas_AppNote_%28UBX-15030289%29.pdf , then it's not reasonable to expect to have to remove components.

I lean towards U3 being active, since the component layout on the module is approximately as described in the Ublox appnote- even if the open-circuit detection has been omitted or simplified. However I'd be interested in the community's opinion or experience.


Hard to tell from the picture but my guess is:
1: SMA ----> C8A ---> L1A -----> 2nd module pin is the RF microstrip connection from the UFL & SMA into the module pin.

2: L1 <-----<---R2----< 5th module pin is a series circuit which probably enables the module to feed DC bias voltage as a bias Tee  along to the SMA and UFL antennae for their LNAs if they need DC bias on the feed for such.

Looks like: module #2 RF RX <--- C8A <-- C8B <-- microstrip RF 50Z trace <--- U3-6.
Therefore maybe we have:
U3-3 <--L2A <--- L2B <--- C4B <-- C4A <--- patch ant.
Therefore U3 should perhaps be a LNA and the patch antenna signal comes through the PCB via into C4A and U4 is not relevant to antenna switching.  If the patch ant has a built in LNA then I do not know what U3 would be useful for.

So if the UFL and the SMA antennae are not connected then all is well.  But to connect either optional antenna, C8 should be removed if U3-6 is an RF LNA output feeding through C8.

If u3 is a switch, where would its control pin be? I see perhaps:
U3-1: ground plane connection?
U3-2: ground plane connection?
U3-3 RF trace from L2 --- C4 --- via to patch antenna?
U3-4 C5A in parallel with a thick trace.  I'm guessing this is a decoupled VCC since C5B is maybe grounded and the trace seems to go to other possibly decoupling capacitors elsewhere.
U3-5: No connect?    Or maybe tied to U3-4 under the chip to tie it high?  Maybe this is the LNA enable pin.
U3-6 RF trace.

Looks possible to be LNA, not possible to be a switch unless there is a control signal

I don't know why L2 would be in series with C4 though that is a little odd forming a series resonator if that's their connection.

And the connection from L2 to U3 is not so much like a 50 ohm microstrip as is on other traces.

The five-pin U4 is a voltage regulator, marked 4A2D.

I suppose that one thing I could usefully try is looking at the six-pin chip (U3) and finding whether it has CMOS-style power on it, rather than being a passive splitter/combiner.


Is pin U3-5 high?
Does U3 look like it could be in the following package, for example?

q.v. that data sheet.  Looks like the physical connections I'm inferring match the pinout if the enable pin is high.

It couldn't be combining / splitting anything if it has only one RF input and one RF output.
Though the lack of consistent transmission line use is a bit confusing it seems clearly to be a 1-in 1-out device
with a Vcc, Ground, and a couple other pins.

I'm fairly convinced it's a LNA similar to the guesstimated one I mentioned.

The RF layout of the board isn't great though with the large stubs everywhere depending on what antenna is being used but I guess it works.

--- Quote from: MarkMLl on June 13, 2021, 05:06:23 pm ---The five-pin U4 is a voltage regulator, marked 4A2D.

I suppose that one thing I could usefully try is looking at the six-pin chip (U3) and finding whether it has CMOS-style power on it, rather than being a passive splitter/combiner.


--- End quote ---


[0] Message Index

[#] Next page

There was an error while thanking
Go to full version