GNSS Antenna Placement in Small Trackers

Concept render

Vendor guidance for patch GNSS antennas is a ground plane of about 50 x 50 to 70 x 70 mm, 4 to 10 mm of clearance, and a 5 mm keep-out for chip antennas. No vendor we reviewed publishes a battery keep-out, so log C/N0 in the enclosure, with the battery, and with the modem transmitting.

A tracker can show good satellite signal on the bench and weaker signal once the PCB is in its housing next to a battery and a cellular antenna. Rules of thumb on forums conflict with each other. This guide collects what antenna vendors publish, shows how to compare it with your own layout, and lists the measurements you need to take on your sample. It contains no Duobay price and no test result of ours.

What do the antenna vendors publish?

Each row comes from a vendor document, read on October 31, 2026. These are vendor-specific guidelines for the parts they describe. They are not universal rules.

How do you compare your board with the guidance?

Start with ground area. For a rectangular ground plane, the area in square millimeters is length times width, with both in millimeters.

A_board   = L × W                       (mm², your ground plane)
A_low     = 50 × 50 = 2,500 mm²        (lower end of u-blox range)
A_high    = 70 × 70 = 4,900 mm²        (upper end of u-blox range)
A_opt     = 70 × 70 = 4,900 mm²        (plane the Taoglas 25 mm patch is optimized for)

ratio_low = A_board / A_low

Worked example with a hypothetical board. Suppose L = 40 mm and W = 25 mm, which are invented dimensions for illustration only. Then A_board = 1,000 mm², and ratio_low = 1,000 / 2,500 = 0.4. That board has 40 percent of the area at the lower end of the guidance and is far below the 4,900 mm² plane the Taoglas 25 mm patch is optimized for. Replace L and W with your own measured values. The guidance describes a compromise for ceramic patches, so a ratio below 1 means: read your antenna datasheet and measure the board.

Then check clearances with a ruler on your assembly: 4 to 10 mm from board or housing for a patch, 5 mm keep-out for a chip antenna. The Taoglas datasheet also asks for a square, symmetric ground plane and a symmetric environment around the antenna, with the antenna centered. Write each measured distance next to the vendor number.

What about the battery and the enclosure?

This is where the public vendor documents say the least. We found no official source that quantifies a battery to antenna keep-out distance, a dB penalty from battery proximity, or a detuning effect from the plastic enclosure. Vendor figure: not published, measure it on your board. We have not measured these effects either, so this article states none.

What the vendors do say is about the surrounding environment. Taoglas recommends clearance from the housing and asks for a symmetric environment. u-blox calls out the keep-out area as having an important impact on chip antenna efficiency. A large metal-cased cell is part of the environment. Treat its position as a variable to test.

What are the common pitfalls?

These follow from the vendor statements above. They are engineering reasoning from those statements. No company incident is described.

• Following one rule of thumb for every antenna. The ledger sources give different numbers for different parts: 50 x 50 to 70 x 70 mm for patches in the u-blox note, a 70 x 70 mm optimization point in one Taoglas datasheet, and a typical 5 mm keep-out for chip antennas. Vendor figures belong to the part they describe. Use the datasheet of your antenna first.
• Assuming a bigger ground plane always helps. Taoglas says gain generally rises with ground plane size, while u-blox calls 50 x 50 to 70 x 70 mm a compromise between gain and polarization loss. Check both effects in your antenna datasheet.
• Tuning the patch on a bench plane and shipping it on a smaller board. Taoglas states that ground plane size shifts the patch center frequency. A patch tuned on a large test plane may sit at a different frequency on your small PCB, so the tuning must be checked in the real device. This is engineering reasoning from that statement.
• Putting a chip antenna on a board with a thin ground plane. u-blox says an inadequate ground plane makes a chip antenna insufficient for GNSS. A tiny board that cannot give the chip antenna its footprint and keep-out will not meet the line you hoped for.
• Placing the GNSS and LTE antennas as close as the enclosure allows. The u-blox white paper advises maximizing the antenna distance when a cellular transmitter is in the same device. Short spacing in a compact housing works against that advice.
• Testing only the bare board. Taoglas asks for clearance from the housing and says the patch must be tuned in the real device. A pass on the open PCB therefore says little about the closed product. Test in the enclosure as built.
• Trusting a battery rule you cannot find a source for. No official vendor source in our research quantifies a battery to antenna keep-out distance or a dB penalty. A number found in a forum or a repost has no basis here. Vendor figure: not published. Measure it on your board.

What should you measure on your own board?

The measurement is average C/N0 of the high-elevation satellites, logged from the module's NMEA GSV sentences or UBX-NAV-SAT messages, in each state of the product. Change one thing between runs. The pass line is set by you from the product requirement, and you should record the basis for it. The 44 dBHz line from u-blox is one reference point. Your product requirement still sets the pass line. This is an engineering summary of public vendor guidance and does not replace your antenna vendor's support.

Fix the pass line before you look at the logs. A line chosen after you see the result tends to land just under it.

What should you ask before you freeze the layout?

Take this list to your RF designer and to any supplier who proposes an antenna placement.

• Is our antenna a patch or a chip, and does our board offer the ground plane that the datasheet for that exact part expects?
• What are our ground plane length and width, and how do they compare with the 50 x 50 to 70 x 70 mm guidance and the 70 x 70 mm plane the Taoglas datasheet is optimized for?
• Does the antenna sit centered on its ground plane with a symmetric environment around it, as the Taoglas datasheet asks?
• What is the measured clearance from the antenna to the housing, the battery and any metal, and where did we write it down?
• Is the patch tuned in the real assembly, or only on the vendor evaluation plane?
• How far is the GNSS antenna from the cellular antenna in the final enclosure, and which LTE bands will the product use?
• What state does the product spend most of its time in, and have we logged C/N0 in that state?
• Where is our pass line written down, and what product requirement does it come from?
• Who repeats this test when the battery supplier, housing material or modem changes?

If you want help laying out and testing a tracker, tell us about your device. You can also read how LTE and GNSS tracker development is organized.

Sources

All pages accessed October 31, 2026. Vendor guidance is specific to the parts described, so re-read your own datasheet before you decide.

• u-blox, GNSS antennas application note: https://content.u-blox.com/sites/default/files/products/documents/GNSS-Antennas_AppNote_(UBX-15030289).pdf (accessed 2026-10-31)
• u-blox, GNSS interference white paper: https://content.u-blox.com/sites/default/files/GNSS_Interference_white_paper_FINAL.PDF (accessed 2026-10-31)
• Taoglas, GPS patch antenna integration application note: https://www.taoglas.com/wp-content/uploads/pdf/TAOGLAS-GPS-Patch-Antenna-Integration-Application-NoteAPN-12-8-002.B.pdf (accessed 2026-10-31)
• Taoglas, 25 mm SMD patch datasheet: https://cdn.taoglas.com/datasheets/ASGPDF254.A.pdf (accessed 2026-10-31)