Communication Range of DWM1000

Wireless Connectivity Ultra-Wideband
, , ,
1

Hi All

Does anyone test the communication range of DWM1000
In my experiance
Indoor Line Of Side (LOS) range is 7m with default configurations.
Configuration are 6.8Mbps, 16 MHz PRF, 128 preabmle length and 34 data length (this is not default value)

When i change the configurations to 110kbps, 64 MHz PRF, 4096 preamble length and 34 data length, the communication range is same,
7m indoor LOS

Does somebady have suggestion?

Regards

Metin…

2

Hi Metin,

Poor range can have various reasons. It could be configuration, the schematic and/or not following the datasheet eg chapter 5.1 Application Board Layout Guidelines and/or not taking into account multi-path signalling caused by reflections, see chapter 3.4 LOS operation with multipath in application note APS006_part 1 available on decawace.com.
Also you should increase the transmit power by approximately 3db when using modules. This because DecaRanging software is targeted at DecaWave’s EVB1000 evaluation board which has a different RF path and antennas compared to the DWM1000.
Also worth checking is APU001 CONFIGURING THE DW1000 FOR DATASHEET USE CASES.

Hopefully of some help
Leo

3

Hi Metin,

One thing that I found changed the range of the device significantly where I live (especially indoors) was the channel used. Try channel 5, it gave us the best range: 30+m indoors. The evk kit i started with used channel 2 and with that I got ranges similar to what you are getting.

Best of luck,
Niklas

4

I also suggest using the default channel 5 with the modifications specified in the user manual at the end of chapter 2 (I believe). I reached 33m LOS in an outside environment.

5

The settings we ended up using in Hayward with the LPS:

Channel 5
Prf DWT_PRF_16M,
datarate DWT_BR_110K
preamble 3
preamble length DWT_PLEN_1024
pacsize DWT_PAC32
sfd 1
sfd timeout (1025 + 64 - 32)

The output power settings we used were:

Low power: (used in hayward to reduce reflections due to large metal structures around the robots)
0x0E082848, //16M prf power
0x25456585 //64M prf power

High power:
0x0E08281F, //16M prf power
0x2545651F //64M prf power

Note that the high power settings may not be legal in all countries / areas.

We got about 330m LOS with this setup outdoors.

Hope that helps someone.
Niklas

6

Hi all,

first of all, I wanted to ask you all to describe your experiences/surveys more specifically. For example, when you say “my range is 10m LOS”, please write whether it is the last successful ranging you got or the 10% packet error rate area. Or at least describe it verbally: good or poor performance.

Second thing, it is really interesting what Niklas wrote about his 300m outdoor range. Last week I performed range test, and on 110k baud rate and 2048 PLEN, channel 3 the last succesful measurement I got was in the range of approx. 330m as well. In my case however, I was expecting much more because I use LNA (+15dB) in my application. The rx_power read on the last few measurements was around -107dBm, so it is in the limits (datasheet states -107dBm on those settings). Can you tell us more about your test? What were the antennas you used?

Regards,
Bartosz

7

Bartosz,

330m was the last successful 2way ranging we achieved when walking away from each other on a inner city street in Stockholm. The test was performed with the DWM1000 module hence only whatever antenna is already on there. The DWM1000 was mounted on the LPS (not the mini) at the time of this test.

Cheers,
Niklas

8

Then 330m is decent result I think. With that distance, you probably already disturbed the Fresnel zone so if you would place the transceivers higher, it would be even more.

Thanks for help :slight_smile:

9

Can anybody send me the zip file “TREK1000_1.0_Enabling_UART_Example.zip” from DecaWave??

10

Hey all,

one more question to Niklas: did you record the RX_power from those 330m measurements.? It is really important to me, because I can barely get 330m on the board with LNA and antennas from TREK devboard which are surely better than those from DWM1000. If you have maximum allowed by DW1000 tx_power which is -14.3 dBm, and free space suppression around 98dB (330m on 6GHz) it means that you properly received frames on -112dBm, which for me is simply not possible. Can you comment on that?

Regards,
Bartosz

11

Hi Group,

I test the communication range of DWM1000s after configure them like Niklas configuration

Channel 5 (Question: when i configure receiver channel to 5, and transmitter channel to 1 or 4, they can communicate over different channels, is it possible or am i doing incorrect configurations?)
Prf DWT_PRF_16M,
datarate DWT_BR_110K
preamble 3
preamble length DWT_PLEN_1024
pacsize DWT_PAC32
sfd 1 (I don’t understant that, how can i configure sfd 1)
sfd timeout (1025 + 64 + 32) (when sfd timeout was (1025 + 64 - 32) like Niklas said, there was no communication, so i change it to (1025 + 64 + 32)

High power:
0x0E08281F, //16M prf power (smart tx is enabled)

My last successful range was 50m. My test board is above.
Do you have suggestions for increasing the range?
How can i test that the configurations are true?
Must i increase the tx power or disable smart tx power?
I think i make some mistakes!

Regards

Metin

12

Hi all,

  1. Smart TX power settings are useless in your case, because just by PLEN = 1024 your transmission time is greater than 1ms. Read the datasheet. If you dont need to stay within regulations, you can increase power settings :slight_smile:
  2. You can improve the module placement (read DWM1000 datasheet, Chapter5.
  3. Read the Rx power on your max range and check what is corellated with receiver sensitivity.
  4. Do you use Decawave API or you wtire your own software ?

Regards,

Bartosz Kawłatow

13

Hi Bartosz,

I use my own software.
I understand from your suggestion that i must use smart tx power because of preamble length.
Can you share your configurations and last successful range, and placement of DWM1000 (if you use it)

Regards

Metin…

14

Hi Leo,

I make my tests with PCB which is made for another project for RF communication with another RF module,
i connect the DWM1000s on it with cables, i share the photo below
I can share the schematic if it is necessary, no problem, is it necessary?

Regards

Metin

15

Hi,

Now i test the range again with 0x1F1F1F1F power setting and disabled smart tx,
Communication range is
35m when data rate is 6800kbps
42m when data rate is 110kbps
There is something strange?
Do you have any idea?

Configurations:
Channel 5 (Question: when i configure receiver channel to 5, and transmitter channel to 1 or 4, they can communicate over different channels, is it possible or am i doing incorrect configurations?)
Prf DWT_PRF_16M,
preamble 3
preamble length DWT_PLEN_1024
pacsize DWT_PAC32
sfd 1
sfd timeout (1025 + 64 + 32)

Regards

Metin

16

Hi Niklas,

I assume the TX_POWER settings that you used here were with smart power on. The power setting 0x0E08281F for high power is too high for manual power. It would be great if you could explain , how did you manage to use this settings, thank you.

17

Hi Ram,

I may be wrong but as I understand it these settings whilst perhaps not legal in your area should still work in manual mode (without smart power). I’ve not tested them very thoroughly but they were the ones that worked the best for me at the time.

This is how I would configure them using the decawave api (using a different example for channel 2):

dwt_txconfig_t configTX;
configTX.PGdly = 0xc2;
configTX.power = 0x15355575;
dwt_configuretxrf(&configTX);

Note that this is not using the EVK but the LPS boards from Loligo (loligo.se).
That is essentially an Arduino + DWM1000.

Cheers,
Niklas