DWM1001 Power consumption

Decawave has an utility called power calculator developed in java. It is just a front end to an excel file . This front end is not working properly because java is giving an errror message : java.io.exception: server returned HTTP : response code 403 for url:// www.decawave.com//TWRCalculator/TWR_power_calculator.xlsx . So , I downloaded this excel file to calculate the battery duration . By default, factory configuration of Anchors comes in mode 2, ie, 6.8Mbps, 12 bytes of payload. I´m using a rechargeable battery RC123 of 3.7V and 650 mAh. I put this data in the worksheet and results in 122 days of battery. Also I´m using a development Kit DEVK-1001.

*My tests shows only one day of life. I would like to know what I need to do to improve battery duration, if I need to select another mode or if there is something else that I should do

Hi,

Could I just confirm with you what device you’re powering with the RC123? Is it the DWM1001 (module on its own) or the DWM1001-DEV? (There are 12 DMW1001-DEVs in a MDEK1001)?

Either way, the power calculator is based on the power required for the DW1000 IC , on its own. It does not calculate and include the power requirement /usage of all other components on the board like MCU, accelerometer, DC-DC Converter and other components.

To minimise current usage (of the DWM1001-Dev) one could take some measures by desoldering some of the jumpers as described in section 8 of the DWM1001-DEV datasheet.

For example it describes that solder Bridge J4 can be desoldered and a resistor placed across connector J2 to allow measurement of the modules current consumption.

It describes also how the disable the current drawn by the Jlink and Leds.

For other hints on reducing power I suggest you have a look at APS001 APPLICATION NOTE DW1000 POWER CONSUMPTION and the DW1000 IC user manual.

Regards,
Leo

Hello Leo. Yes I´m using DWM1001-DEV.

I found a comment in this forum from someone that has the same problem I experienced. His battery duration was 15 hours, the same as mine. The answer was that anchors are not optimized for power consumption . So in fact we need an external power supply. Do you agree ?

Is there a possibility to put the anchor in sleep mode in specific periods ?

I need a solution for outdoor environment. Have you seen any solution which uses solar panel ?

So even disabling Jlink and leds (only one. The others I can disable via minicom command), I still not be able to reach the expected battery duration using the excel spreadsheet , correct?

Thanks for your help!

Generally, yes, anchors are not optimized for power consumption as they expect to be listening mostly all the time and expect to have unlimited power.

It takes very careful protocol design to build such a system where both sides, tags and anchors, can be off significant amounts of time. Such a system suffers from the potential they don’t hear each other. Then, even if they do, you have to set up a means to schedule both sides so they awake and hear each other in the future. With clock drift and other issues, this becomes a very non trivial project which takes very careful protocol design.

The real issue here is that receive power on the DW1000 is so high, about 400 mW. So the major requirement is to be in receive as little as possible. Ciholas is working on such systems now, wireless anchors which operate for long periods on batteries. They are best used in places with high install costs, or temporary installs, and with systems that have fairly low capacity requirements. If you end up using full air time capacity, then anchors are on most of the time and the battery is gone.

Another issue with battery anchors is that eventually somebody has to charge or swap the batteries. Anchors are usually mounted up high for best performance, so this isn’t trivial in many cases.

That is possible. You’d need a sufficiently large battery to work overnight and to bridge cloudy days. In cold climates, snow can cover the array for a long period of time.

Some numbers: estimate anchor uses about 15 WH per day (10 WH for the DW1000, rest of system about 5 WH). Battery should be about 6 times that number to bridge a week, so 100 WH or so. Solar array should recharge battery in 6 solid hours, so about a 15 to 20 watt panel. That is about 30 to 40 cm square panel, so not small, but not huge. So, seems doable. Does add some extra install steps and requires the anchor sites have good sun visibility.

Wireless anchors are on everyone’s hot list, but when the battery servicing burdens and capacity limitations become evident, it soon becomes apparent that wired anchors are preferred in most permanent and professional installations.

Mike Ciholas, President, Ciholas, Inc
3700 Bell Road, Newburgh, IN 47630 USA
mikec@ciholas.com
+1 812 962 9408

Mike. Thanks for the answer . I know that there is a new module DWM1004C in a MDEK-1001 KIT. The tech description shows that it has less power consumption.

My question is : using this new chip can improve a anchor power consumption ?

Also I see a warning message in documentation that this product is not certified yet. Does anybody knows when this will happen ?

The power consumption of an anchor is dominated by the DW1000 UWB chip being in receive most of the time. Given the change between DWM1001 and DWM1004 is mostly the processor (STM32L041 replaces nRF52832), that’s not going to change the power profile of an anchor using either of these modules.

The fundamental power profile of an anchor is the receive duty cycle of the DW1000 chip. If you don’t have some sort of complex protocol to keep the anchor in sleep, your power is basically the DW1000 chip being in receive listening for tags, or about 400 mW.

Datasheet says “planned for October 2019”.

Certifications always seem to take longer than they should. It isn’t so much the technical requirements, it is all the non technical things that go with these approvals and UWB is more complicated than most RF systems in that regard.

Mike Ciholas, President, Ciholas, Inc
3700 Bell Road, Newburgh, IN 47630 USA
mikec@ciholas.com
+1 812 962 9408

To clarify:

DWM1004C is not part of the MDEK1001 kit or system. It is designed as a Tag for a TDoA system.
The software available for the DWM1004C is TDoA blink code.
DWM1001 uses PANS software stack, a TWR based network, not TDoA.

Hi,

Any news on the wireless anchors which operate for long periods on batteries? Is the protocol development still a bad idea for beginners?

Thanks,

Guillem Balcells.