Power calibration & Certification

Hi all,

In certification field, I’m wondering if an outdoor LT1 system is possible (regarding ETSI EN 302 065-2 Harmonised Standard).

They say that “fixed outdoor LT1 transmitters are not permitted. Typically, LT1 transmitters are mobile location trackings tags which are attached to people or objects, and tags are tracked using a fixed receiver infrastructure to only receive the UWB emission emitted by the tags.”

To compute distance, we need some exchange messages ==> receiver is not only receive UWB emission, it also emits.

Does it mean that outdoor LT1 application is not possible ?
Or do I misunderstand it ?

If I have to use LT2 application, I’m annoyed because of the ultra low duty cycle (1.5% of the Ton) !

Best Regards,


Hi Xavier,

How about using other frequency bands for data transfer and synchronisation? Your anchors will then silently listen for UWB packages while performing data transfer and synchronisation e.g. using the 802.11x U-NII bands (https://en.wikipedia.org/wiki/U-NII). If the distance is short and your bandwidth requirements are low, you might even be able to get the job done using the 2.4 GHz radio of nRF52 (lives inside the new DWM1001).

Best Regards,
Göran Nordahl

To be more specific:

For our TDOA system to work with wireless synchronization we need our anchors to transmit to each other. In most regions this is not allowed outdoors. So the idea is to use DW1000 to transmit a UWB signal at a center frequency of say 5491MHz with a bandwidth of say 250-300MHz for wireless sync.
This band would fall in the 802.11x U-NII bands. The operation bands for WiFi 5G are shown below:

Band 1: 5150 ~ 5250MHz
Band 2A: 5250 ~ 5350MHz (DFS)
Band 2C: 5470 ~ 5725MHz (DFS)
Band 3: 5725 ~ 5850MHz

If one wants to transmit a signal with a 250~300MHz bandwidth, it’s better to use the Band 2C with a center frequency of say 5675MHz. (Note Band 3 may also be an option with the use of an external bandpass filter)
Since for Band 2C, the operation frequency range is almost 400MHz and the power limit is <250mW, the DW1000 can handle the power level and bandwidth! Only challenge is that one would require Dynamic Frequency Selection (DFS) if operating in this band.

So for the DW1000, for the Frequency synthesiser you need to set 2B.07[3:0] to the multiple of 499.2MHz that you want:
For 4.992 GHz, set 2B.07 = 0xA, for 5.4912 GHz, set 2B.07=0xB or for 5.9904 GHz, set 2B.07=0xC

To get other intermediate frequencies, it would be necessary to use a crystal with a frequency close to 38.4MHz but changed by a small percentage. E.g. using a 37MHz crystal and 2B.07 value of 0xC would give a centre frequency of 5.772GHz.

Bit[4] of register “B.07 is used to set the “PLL High Band Mode”, so you can adjust this setting and see what is more stable and flat on a spectrum analyser. Historically this was only needed for 6.5GHz, because the next lower channel was 4.5GHz.
There are other tuning bits that may need to be set to ensure the TX spectra is somewhat tuned. The easiest way to do this in DecaRanging is to configure the device with the defaults settings for channel 3 or channel 5 before applying the above synthesiser changes. Pick which initial configuration that gives the best response.

The power level is set via register 0x1E, to max this out set to 0x1F1F1F1F, but this may cause compression artefacts. If the spectrum doesn’t look good, back it off to 0x3E3E3E3E or lower. This can allow the transmit level to get to a maximum of 0dBm but realistically only about -3dBm. For higher transmit power levels, an external power amplifier is required.

A problem that the user might run into is that many of the UNII band regulations require a much smaller bandwidth than the default of 500MHz. The bandwidth is set by programming the PG_DELAY register which controls the pulse width. This is register 2A.0B. The default pulse width in this register is 0xC5. The larger this number is, the greater the pulse width in time. A wider pulse has a narrower bandwidth, so a value of e.g. D5 will be narrower. If however, the number is set too high, the pulse shape becomes distorted. The minimum bandwidth that can be programmed with this method is approximately 250MHz. To get smaller bandwidths, and external filter would be required. Using a smaller bandwidth will affect the time of arrival accuracy a little bit, but not very much.

So with all of the above information it could be possible to either use Band 3 with a BP filter or Band 2C with some kind of DFS implementation.

Best Regards,
Göran Nordahl