we run long-term HIL tests with typically 3–8 nRF9151 DKs:
- DKs spaced only about 1–3 cm apart
- short USB 3.x cables to powered USB 3.2 hubs directly below them
- sometimes two such hubs next to each other
- hubs powered by 20 W / 5 V switching supplies
- Intel Core i7 laptop about 30 cm below
- some DKs have an external sensor connected via ~3 cm I2C/VDD/GND wires, read every 2 minutes
The DKs use their onboard LTE antennas.
We see quite a few LTE reconnect failures after PSM and also general LTE connection/registration issues.
Our main question is how Nordic would classify such a setup:
Could this actually be a useful IoT field stress-test simulator, exposing behavior that may also occur in difficult real-world RF/EMC environments?
Or is the setup so unfavorable — due to closely spaced cellular transmitters, USB 3.x noise, hubs, power supplies, etc. — that LTE problems are essentially expected and the results should not be considered representative of problematic real-world deployments?
We are not looking for calibrated RF measurements, only for a useful long-term HIL/soak-test environment.
Would you consider this setup acceptable, or would you recommend significantly more separation between the DKs and surrounding electronics?
What would the simplest cost and space effective clean 4–10 DK HIL setup look like in practice?
For example: is a powered USB hub sufficient, or is a lab PSU preferable — and if so, one PSU for all DKs or one per DK? Would ~2 m USB cables and ~1 m spacing between DKs/antennas be a reasonable setup? Anything else?
Would the nRF9151 SMA DK be the recommended solution for such a multi-device HIL setup?
For example, keeping 4–10 SMA DKs near the USB infrastructure, but placing their external LTE antennas ~0.5–1 m apart and away from hubs/PC/power supplies?
