Sierra Wireless XR80 and EM9291: A Quality Inspector's Field Checklist

I'm a quality/compliance manager at a wireless hardware integration company. I review every cellular router and module before it reaches customers—roughly 200 units a year. We've done maybe 200 unique acceptance runs this year. Maybe 180, I'd have to check the system.

I've rejected about 14% of first deliveries in 2025. Most of those failures were avoidable.

This checklist is for anyone specifying, integrating, or accepting Sierra Wireless gear for IoT, fleet, or mission-critical devices. It's not about datasheet marketing. It's what to check when the box arrives.

Before you start: know which Sierra Wireless part you're really using

The first failure I see is confusion between modules and routers. The Sierra Wireless - EM9291 is a module inside a larger device. The Sierra Wireless XR80 is a standalone industrial router. If you're testing the XR80, you're testing a finished product. If you're testing the EM9291, you're testing a component that only works when the host board, antenna, and firmware are right.

Searching sierra-wireless by itself is a waste of time; include the SKU. Searching sierra wireless xr80 or sierra wireless - em9291 without your regional variant gives you pretty brochures. Not enough. Start with the exact part number on the label and the firmware version. The label is your first checkpoint.

Step 1: Verify the part number, variant, and firmware against the order

This sounds too basic, but it's where I've wasted the most money.

In my first year, I made the classic variant error: I assumed EM9291 meant the same thing across every purchase order. It doesn't. Different regional variants support different band plans. The same part name can ship with different carrier configurations.

Checklist:

• Compare the actual label to the PO line item. Not the box, the label.
• Record the IMEI or serial number and the firmware build.
• Verify the region code: part numbers ending in different letters often mean different carrier or band profiles.

That 90-second check caught a $22,000 mistake last year. We had ordered XR80 units for a public safety project and received the wrong regional SKU. The vendor claimed it was within industry standard. We rejected the batch, and they redid it at their cost.

This also applies to blood pressure monitors or any connected medical device using a cellular module. The module's region code determines whether the device has legal approval for that market. Don't assume the manufacturer handled it.

Step 2: Check certification marks and carrier approvals before powering up

Everything I'd read about certifications said that once the module is approved, you're done. In practice, system-level certification still fails. The module can be approved, but the host device may not be.

For US deployments, check for the FCC ID on the label. For the EM9291, that's usually on the module label or the host device label. For the XR80, it's on the router chassis. You also want to see which carrier approvals the exact SKU carries. PTCRB is the baseline for many carriers, but carrier acceptance lists go beyond it.

Ask the vendor for the specific carrier acceptance list. Don't accept a generic yes, it's certified. A carrier can approve a module for one band plan but not the band plan your project uses.

Oh, and that step matters even if the module is already certified. A blood pressure monitor does not automatically inherit the module's certification. The whole device goes through additional testing.

Step 3: Run a network tester pass, not just a signal bar check

This is the step most checklists skip. Everyone checks that the LED is green. Hardly anyone runs a proper network tester session.

Why does this matter? Because connected and stable are different things. I've seen devices show full bars and then drop packets during upload.

Here's the network tester routine we use:

1. Insert the correct SIM for the carrier and region.
2. Put the device in a known location with a good test antenna.
3. Run a throughput test for at least 10 minutes, not 30 seconds.
4. Force a cell handover if you can. Observe whether the connection drops.
5. Log signal quality metrics: RSRP, RSRQ, SINR, not just bars.

A basic network tester will save you a $500 field visit. It'll also catch the intermittent issue that appears after a device has been running for an hour. We had a fleet of XR80 units that failed only under load. The network tester caught it on the bench, and we never saw it with a phone app.

For a blood pressure monitor with an embedded EM9291 module, add one more test: make sure the cellular transmissions don't interfere with the measurement electronics. That's a network tester discipline—run a transmission and a reading at the same time. If the test shows noise, the board layout is wrong.

Step 4: Load-test power and thermal behavior

This is the step that feels like overkill until it ruins your project.

A field failure in March 2023 changed how I think about connectors. One router lost connectivity after a heat cycle. The module was fine, the connector wasn't. Normal temperature range looked okay on paper, but the thermal expansion on that specific connector was too much for the solder joints.

Do this:

• Run the device at maximum transmit power for 30 minutes.
• Measure the case temperature at the module and at the power supply.
• Cycle the device through its worst-case temperature range, not storage range.
• Check that the power supply can handle the peak current during LTE or 5G transmission bursts.

Like most beginners, I approved a batch based on desktop testing. Learned that lesson the hard way when a 1,000-unit deployment had intermittent uplink issues. The lab was air-conditioned; the field cabinet was not. The difference was 9 degrees Celsius.

If your product is a blood pressure monitor, this step is even more important. The modem transmits near sensitive analog circuits. Excess heat can shift sensor calibration. Let it run, then measure the reading against a reference device.

Step 5: Inspect connectors, antennas, and mechanical integrity

When someone asks for a Cypress vs. something comparison, my first question isn't about the chip. It's about how the thing will be physically handled. A great module dies instantly when the antenna connector is loose.

Checklist:

• Torque the antenna connectors if the spec requires it. Loose connectors are the #1 cause of intermittent RF issues in my experience.
• Pull test the cables gently. The connector should not move.
• Inspect the shielding: are there gaps around the module shield can?
• Verify the enclosure doesn't stress the PCB when screws are tightened.

That last one is one of those nobody-checks-until-it-fails items. We once rejected a batch of 8,000 units because the screws pinched the antenna cable. The team who assembled them thought it was fine. It worked in final test. It failed in the field after vibration.

There's something satisfying about a perfectly executed mechanical inspection. After all the test data and certification documents, seeing the physical hardware match the drawings—that's the payoff.

Common mistakes to avoid

Here are the errors I see repeatedly when teams set up their own acceptance testing:

• Testing with a development board instead of the real enclosure. Antenna placement changes everything.
• Using a SIM that's already in another device. Some carriers lock the SIM to the IMEI.
• Skipping the firmware update because the unit works out of the box. Works, but not securely.
• Trusting the vendor's test report without verifying the test date. A report from two firmware revisions ago is meaningless.
• Only checking the module and not the complete system. The Sierra Wireless XR80 can pass every test while the device connected to it fails.

The question isn't whether you can afford to run these checks. It's whether you can afford not to. The cheapest test is the one you run before the batch goes to the field.

That's it. Five steps, plus the mistakes that get your budget cut.