GPS: External Antenna for V.KEL VK2828U7G5LF

[0xPIT]

Hi,

I suck with RF and I'd like to ask for help.
I want to attach an external GPS antenna from my bin to a VK2828U7G5LF (

) module, so that I can mount it to another PCB that will land in some enclosure at some point.

I've desoldered the stock antenna already, so what is a good approach to attach coax for an external antenna?

Can I simply solder coax to the pad where the old antenna was, and the coax' shied to GND?

Is there special consideration required, maybe in order to match impedance?
The passive parts and LNA between antenna pad and the ublox G7020 can stay as-is, I presume?
Any RF magic or some sort of sacrifice required?

Thanks,

pit

[pigrew]

It's likely perfectly fine to do what you propose.

You need to consider biasing and active vs passive antennas. My guess is that the board you show has a 50-ohm antenna, and a LNA+filter. Since it has its own LNA, I don't think that it has a bias-tee on it, so I think that no power would be provided for an active antenna. If this is the case, you'll need to add a bias-tee for your antenna somewhere. The board as it is seems designed for a passive antenna.

The antenna is most likely 50 ohm, so my expectation is that things will already be matched.

The other thing to worry about is RF power levels. I think that the modules usually like 20-40 dB gain from the LNA. It's best to put the LNA as close to the antenna as possible (in order to improve the noise figure). So, you may want to remove the LNA from that board (but probably leave the filter). Check the ublox hardware integration manual for ideas. It describes various designs for active and for passive antennas.

For connecting a coax cable, sure, you can directly solder it onto the board as you describe.

[cdev]

Ubox should have a "hardware integration" document for the specific chip family. It will give you info on the specific antenna interactions needed for that particular chip. Some need an active antenna or need to satisfy some specific set of conditions. Its not that simple.

Also, RF at 1.5 GHz is a bit particular about the quality of the connection. Keep leads short, and try to maintain the right impedance, is the best advice I can give you.

[0xPIT]

Thanks for your answers, sounds simple enough.

Now, as I removed the shield at the antenna to extend the coax, it seems that this is actually an active antenna.
Obviously, I have no datasheet for it.

Do you have any advise on how to power the antenna?
Will some cutten limiting resistor like in the ublox hardware manual work?

[cdev]

short term, its awful nice to have a bias tee around when you need one. Then you dont have to worry about the power until everything else is working.

[cdev]

Do you have any vero board lying around, or better yet, old PCBs that have chunks of impedance controlled microstrip that you can carve up ? If so then you can try to make a ghetto bias-tee. You may have to experiment on the values because real world components vary a great deal in the parameters that matter to bias tees.

The simplest bias tee is just a DC block capacitor (to block DC from going into the receiver. Dont forget that!) and an inductor (choke) for the DC, and maybe a second bypass cap (right as the power comes in) For a broadband bias tee (which Ive always had some issues with, I have never made a good ultra boradband bias tee..) you would want to bring the power out to the side and zig zag with the Ls and Cs. There you also may want to put lowish value resistors in parallel to the inductors to reduce the Q.  But for one frequency use, screw it. For just one fairly narrow band of freuencies like GPS/Glonass/Beidou dont waste all that time, just try to get the core values right. Look in the ublox book for their recommendations for capacitors. Spend an extra few pennies getting the ones they recommend. Keep your leads/traces as short as is humanly possible.

Do you want some virtually foolproof GPS antenna ideas that work without an LNA?