2.4GHz Wi-Fi – Effect of Settings on Second Harmonic at 4.8GHz
When using a 2.4GHz radio inside a product, one of the most common radiated emissions problems observed is the second harmonic at 4.8GHz approaching or exceeding the limit line.
This can be caused by a nearby parasitic antenna structure (e.g. metalwork, PCB cutouts, wiring looms) or by using high output powers in the Wi-Fi radio.
In this case, the product was a handheld battery powered device with a uBlox Wi-Fi module.
The customer wanted to try different settings to judge the effect on the 4.8GHz harmonic to remedy a test failure at another lab.
Radiated emissions being measured to EN 55032 Class A in our Fully Anechoic Chamber.
Takeaway
The important takeaway from this exercise (and something that we will be doing in future) is to measure the 2nd harmonic radiated emissions of Wi-Fi
— at both channel 1 and channel 11 (6dB difference in this case)
— and at the lowest modulation / data rate (7dB difference in this case)
Cumulatively, this could make up to *13dB* of difference based on these figures.
Experiment 1) Add a 2.4GHz Notch Filter
The first thing we would do when measuring an RF emission second harmonic is to evaluate how much is being caused by overload of the measurement system (preamp, spectrum analyser). We add a -30dB 2.4GHz notch filter in series with the antenna inside the chamber before preamp or any other active measurement equipment.
This removes the RF carrier from the input of the measurement system, preventing overload. The overall insertion loss at 4.8GHz is in the order of 1.5dB
Wi-Fi Channel |
Modulation | 2.4GHz Notch Filter | Output Power | Polarisation | 2nd Harmonic Margin (to Class A limits) (dB) |
1 |
1Mbps/20MHz | No | 6dBm | V |
-3.2 |
1 | 1Mbps/20MHz | Yes | 6dBm | V |
-7.52 |
Accounting for the 1.5dB reduction in the filter, we reduce the 2nd harmonic by 3dB. So what we are measuring is mostly real and not an artefact of the measurement system.
Experiment 2) Increase Output Power
This was to see at what power level the 2nd harmonic would meet the limit.
The non linear characteristics of the curve certainly point towards saturation of the power amplifier part of the Wi-Fi module RF output.
One might be tempted to pick an output power of 14dBm as having sufficient margin. However, given likely variations in operating temperature, supply voltage, tolerance across units, etc. an operating point of 12 or 13dBm would be more prudent.
Wi-Fi Channel | Modulation | 2.4GHz Notch Filter | Output Power | Polarisation | 2nd Harmonic Margin (to Class A limits) (dB) |
1 | 1Mbps/20MHz | No | 6dBm | V | -3.2 |
1 | 1Mbps/20MHz | Yes | 6dBm | V | -7.52 |
1 | 1Mbps/20MHz | Yes | 11dBm | V | -7.18 |
1 | 1Mbps/20MHz | Yes | 13dBm | V | -9 |
1 | 1Mbps/20MHz | Yes | 14dBm | V | -3.1 |
1 | 1Mbps/20MHz | Yes | 15dBm | V | +3.44 |
1 | 1Mbps/20MHz | Yes | 16dBm | V | +6.18 |
This graph has an interesting characteristic. Above 13dBm we suspect that the amplifier in the Wi-Fi device is starting to saturate causing additional second harmonic products.
Further testing took place at 14dBm to see if it increased or decreased the measured level as this was located centrally on the steep part of the curve.
Experiment 3) Changing the signal bandwidth and modulation scheme.
Lower data rate / bandwidth signals appear to have the highest occurrences of 2nd harmonic emissions. Very little difference moving from 20MHz to 40MHz bandwidth.
Wi-Fi Channel | Modulation | 2.4GHz Notch Filter | Output Power | Polarisation | 2nd Harmonic Margin (to Class A limits) (dB) |
1 | 1Mbps/20MHz | Yes | 14dBm | V | -3.1 |
1 | 6Mbps/20MHz | Yes | 14dBm | V | -6.51 |
1 | 12Mbps/20MHz | Yes | 14dBm | V | -10.15 |
1 | 54Mbps/20MHz | Yes | 14dBm | V | -10.7 |
1 | 54Mbps/40MHz | Yes | 14dBm | V | -10.1 |
Experiment 4) Change Wi-Fi Channel / Frequency
The most interesting result was changing the Wi-Fi channel frequency. Increasing it from channel 1 to channel 11 caused the second harmonic emissions to drop by 6dB.
Wi-Fi Channel | Modulation | 2.4GHz Notch Filter | Output Power | Polarisation | 2nd Harmonic Margin (to Class A limits) (dB) |
1 | 1Mbps/20MHz | Yes | 14dBm | V | -3.1 |
3 | 1Mbps/20MHz | Yes | 14dBm | V | -5.5 |
6 | 1Mbps/20MHz | Yes | 14dBm | V | -7.56 |
9 | 1Mbps/20MHz | Yes | 14dBm | V | -9.37 |
11 | 1Mbps/20MHz | Yes | 14dBm | V | -9.5 |