DW1000 Chipset vs DM3000

Hi ,

We are planning to use UWB tracking for Human Proximity detection

Online we searched for available UWB parts, and we came across 02 parts DW1000 & DW3000.

Most of the technical details for DW1000 are available on the website, but there is no detail available for DW3000 part.

Please help me understand the difference between DW1000 and DW3000 from Cost, RF Power, Accuracy , Technical point of view.


Hi Chetan

The DW1000 is our first, mature UWB chip. We also have the DWM1000(C), DWM1001 and DWM1004 modules and the DWS1000 Arduino compatible shield based on it.

The DW3000 is a new UWB chip range that is still being released, which is why there is limited availability and documentation. We also plan to release DWM3xxx and DWS3xxx shields based on it. Mass production is planned for late June - early July.

The main differences between the DW3000 and DW1000:

  • The DW3000 is designed for lower power operation
  • The DW3000 supports higher SPI speeds
  • The DW3000 should result in a lower total BOM cost and count (e.g requiring less caps)
  • The DW3000 supports UWB channel 9 operation (important for TELEC/Japan market)
  • The DW3000 has embedded AES-256 encryption support for secure timestamping, 4z BPRF compliant
  • The DW3000 has a slightly lower TX power than the DW1000
  • We plan to release versions of the DW3000 that support PDoA in a single chip
  • The DW3000 will be available in more compact (CSP, QFN) packages.

See also this post by @Kenneth_Dwyer_DW: Decawave in Japan

Note that the DW1000 and DW3000 are not register compatible.

You can contact sales if you are interested to learn more and get samples of the DW3000 at sales@decawave.com


Thanks Seppe for sharing this information.
it is really helpful.

Now, if we use DW1000 chip set in our design and implement Peer-to-peer positioning.
Please let me know, How many (maximum) Tags we can correctly locate ?
We are looking for 1000-1500 Tags to be located (consider a case of office floor or Production floor etc.)


Hi Chetan

This depends on a lot of factors like the frame configuration, datarate, distance between tags, how many tags are in the same “radio space”, the polling rate, … In short: this depends on the firmware.

We provide PANS for our DWM1001 module, the limitations of which can be found in the System overview document, mainly section 5, “Scalability”.

Your own software solution can have work differently, but based on the information in section 5.3 1000-15000 tags would work in PANS if a location rate of ~0.1Hz is good enough for your use-case.



will the DW3000 IC be available at common distributors like digikey, mouser etc. after the mass production has started ?

best regards


1500 tags within range of each other represents possible range pairs that number over 1 million. Literally, 1500 tags create 1,124,250 possible tag to tag range pairs.

You can’t possibly perform 1 million range operations in any reasonable amount of time, even if you do all sorts of tricks to coalesce ranging operations together.

Even 1000 tags creates 499,500 possible pair sets. That’s still way too much to actually perform ranging.

With so much activity, and no central coordination, battery life will be dreadful. It may take an 18650 cell, or two AA cells, for each tag and you may have to charge it every day.

There are various bits of cleverness you can use to try and make this problem somewhat more tractable. You can use BLE to coordinate to save power (but that saturates BLE bandwidth), you can use RSSI to down select ranging pairs (which may mean you miss pairs you should have ranged, false negatives), you can go really slowly (which means you miss short term contacts), and so forth. But even with all those tricks, there’s no way you handle 1500 tags in one area using only tag to tag methods.

The no infrastructure contact tracing application seems simple only when considering it at a trivial scale. Once you look at the practical and realistic implications, it falls down. The times you need it most, large numbers of people, is when it fails. It will work for trivially small groups, like 10 or 20.

If you have to cover 1500 people in one area, there is only one way to do it, using infrastructure, an array of anchors, which locates all the tags. Such a system can have very high capacity (say 17,000 tags at 5 second interval), and very long battery life (tags only need to beacon, not hear other tags, so they can last months if not years). These are not theoretical numbers, there are systems that do this today.

Mike Ciholas, President, Ciholas, Inc
3700 Bell Road, Newburgh, IN 47630 USA
+1 812 962 9408

Yes, I believe that is the plan, but I can’t confirm anything officially.

You can see our list of distributors here: Our Distributors - Decawave

Note that there isn’t just one DW3000 IC, as said in my previous message we plan to release multiple versions with different capabilities and in different formfactors. Not all of them will be released at the same time.

“Within range” being a key element here. It also depends on what is meant with “Ranging”. Ranging in between 15000 nodes will indeed be impractical, but the PANS example will only range between the tags and 3-4 fixed “Anchors”. Instead of 1,124,250 pairs, this means (simplified) only 1500*[3-4] = 45000 - 6000 ranging pairs in the same radio space.

It is also possible to lower TX power, thus reducing the size of the radio space and thus the amount of nodes in the same space. This combined with a smart TDMA / CSMA scheme can make it possible to implement a “secure bubble” use-case with 1000s of nodes.

As I said before, the answer will depend on a lot of factors. 15000 nodes in a small area will probably not work, but if devices are spread out over a large area and you only want to range between devices close to each other it should be possible to get that working.

Hi Seppe, Mike,

Thanks for sharing all these information, It is really helpful.

We are planning to use DW1000 Chipset, and wants to check availability of 100K chips
Please suggest from where I can get quotation and chips for our device production.


Hello Seppe.
Considering developing projects, show we wait for dw3000 and it’s versions dw3XXX instead of dw1XXX ?


Hi Ramiro

This depends on your application and how urgently you need it.

As stated before, the DW3000 will have new exciting features that will ensure compatibility with upcoming UWB standards, allows lower power consumption and optional single chip PDoA.

If your application needs or could strongly benefit from these features, waiting could be worth it and you can contact our sales team to see if we can help you.

However, it will take some time before DW3xxx parts will be stocked by distributors and we have a lot more demo applications available for the DW1000, like the PANS RTLS system for the DWM1001.
These types of applications are often very beneficial to get started with UWB quickly. It will take a while before we have a mature set of demo applications available for the DW3000. Same goes for modules with integrated MCUs.

For smaller, simple projects the DW1000 might offer enough features and the advantage of modules being readily available.

The DW1000 can also have a bigger range, which can be important for certain applications.

1 Like

Very clear.

Thanks @seppe