Seems like there was a bug in the library I'm using. The default mode is MODE_LONGDATA_RANGE_LOWPOWER. There is a function to change this, used in the example sketch to select MODE_LONGDATA_RANGE_ACCURACY. However, this code looks suspiciously like the defaults are being set again before each transmission.... void DW1000RangingClass::transmitInit() { DW1000.newTransmit(); DW1000.setDefaults(); } After editing the library to change the default to MODE_LONGDATA_RANGE_ACCURACY, I was able to get signal at longer ranges, but still nowhere close to 300 meters. It was pretty reliable up to 30m, then patchy until 50m after which there was mostly no connection. The furthest distance I saw was 72m. According to the datasheet MODE_LONGDATA_RANGE_ACCURACY should be the best for range, ie: MODE_LONGDATA_RANGE_ACCURACY[] = {TRX_RATE_110KBPS, TX_PULSE_FREQ_64MHZ, TX_PREAMBLE_LEN_2048}; Does it seem reasonable that with better antennas this might make it to 300 meters? I'm not sure, perhaps it would. I'm also now wondering if there might not be more problems in the library...
iforce2d if the pcb antenna isn’t tuned right, it could totally be the problem. Also I don’t know if the radiation pattern is the omnidirectional, do you know if it is?
I think it was an very interesting video. I think something was off. No way would a company put out something so under spec. But anyways Kicad is another layout program and it is free and easy to use. Maybe not as easy as Easyeda but you keep the rights to anything you design. Also, don't know if you care but arduino ide has a dark mode (3rd party) that makes it much easier on they eyes. I like your regular guy approach to teaching / showing technology.
Hi mate, Any chance you have a git with your version of the lib? Really priceless information right here. I think this project is worth pursuing 👍 even 3-4 anchors with 30m of range surely has some use cases for your drone project.
From looking at the PCB, it looks like there is a couple of things. One is the size of the PCB itself, the datasheet states in 5.1: "It is also important to note that the ground plane on the application board affects the DWM1000 antenna radiation pattern. In Figure 8 below, d must be a minimum of 10 mm. This gives the most vertically polarized radiation pattern. As d is increased from 10 mm the degree of vertical polarization reduces." In your case d is 6.096mm, so that could be one reason you're getting poor performance. Also I'd recommend a slightly better ground plane, D9 is almost splitting your ground plane in half, the module ground has to take a longer path which can affect performance. Add top ground plane and via stitching between the planes. I'd also recommend getting the decoupling caps (100nF caps) closer to the ATMega and giving the module a few decoupling caps. If you want, flick me a youtube message and I can help drive some board improvements via Skype or Discord.
This video is quite old, and currently im doing some research on these for a work project. What isnt well documented is that you need to set the antenna delay to get decent accuracy. Otherwise as you indicated the antenna delay will contribute to the distance measurement. THere is an arduino sketch somewhere for auto calibrating the ADELAY. do that and the accuracy will increase significantly.
I wonder how much difference the power supply makes to these. You are using USB from your phone which will be stepped up from the internal battery voltage and probably quite noisy. I've found that RF stuff often sucks with USB power, e.g. GPS performance is severely reduced.
In your projet, is the stationnary one coded as an anchor and the one you read serial output on your phone coded as a tag ? Cause' i've built a prototype with two ESP32 board with a newer Decawave chip, the DW3000 chip (compatible with Apple's U1 chip). and got better results, but I'm still guessing on how I could built a scenario where the tag knows how far it is from a particular anchor, and not just "an anchor". How can I "set" the ID of an anchor and be able to read it on the tag.
I wish I was smart enough to look at the documentation for this and understand what I'm looking at... I was looking for cheap solutions to create tracking system for players in paintball/airsoft/nerft/lasertag with +-1m accuracy to implement capture zones domination style that would count players within specified area akin to video game and pushed capture towards the team with more alive players within the capture zone. Guess there aren't dumb-proof hand holding guides for everything on the internet T_T
Intresting stuff! My first thought when this video started was also makeshift super-acurate 3D Positioning, too bad the range is that bad. Imagine you could do this with NRF24L01 modules :)
Thanks for your video. In trying to bring up your PCB design in EasyEDA, can find it but it does not load, saying "document does not exist". Can possibly update the library?
Seems to work fine for me in both Chrome and Firefox, even when not logged in to EasyEDA. So not sure what I could even change that would help. If you're only looking to get the board I used with no changes, I have a bunch of bare PCBs here, could sell you. Let me know at iforce2d at gmail if interested.
Hello. I am trying to communicate two ultra wideband bu01 dw1000 chip using esp32 d1 mini without any breakout board. What i did was to etch a pcb board and only solder the necessary pins to communicate 2 uwb with each other. The pins soldered was the SPI pins (CLK, MOSI, MISO), CS pin ,IRQ pin, RSTn, VSS, and VDD3v3. I checked the soldered pins with multimeter and there is no shorts in wiring connections or whatever. But it just doesn't work. Any ideas why?
I have never used ESP32, and I don't know anything about your custom PCB, or anything about the program you're using. So there's zero chance I can help you sorry.
Aha! I know just the place to test it. Cracking idea. Folks would go nuts for that, renders and wild claims required. What about a precision flight path into the pad as well a line to fly down? All the best from #VertiportAfrica
I remember reading something about different tx powerlevels in datasheet. Would be useful if you output the registers and check. Also it might be silly but are you sure your modules dont transmit at same time? You usually need to sync the clocks and have them tx in sequence.
Modules don't seem very useful in their current form, but interesting video nontheless. Didn't know this was (one of?) the technologies behind indoor positioning... seems like if implemented correctly it would be great for quadcopters, or even autonomous vacuum cleaners ;) The LEDs are showing something useful.. devil is in the details. From left to right... LED1 was showing when the module had finished transmitting something, LED2 was showing when the receiver was active, and LED3 was showing when the receiver had heard something valid... so really for the anchor you probably just want the TX LED (LED1), and for the key RX_OK (LED3)... although RX (LED2) would be good to have also... so you can tell that it is still listening but not hearing valid info. Hope you have luck with them in the future!
Regarding the range: u should look at the datasheet and see if u can't modify a modem bitrate register or smtng. If u turn down the bitrate, u should have drastically improved range. Datasheet said 290m range max, and ur in the middle of a field-- the range ur getting otherwise makes no sense.
have you seen this ? The Daatasheet claims 290 meters @ 110 kbps, page 29 Table 25 states the operation modes, also there are some pages about TX power ... maybe this thing is just not "switched on" properly ? ^^ www.decawave.com/sites/default/files/resources/DW1000-Datasheet-V2.12.pdf i guess the stuff you have to send over spi is in there too ...
Would there be any point in using external, directional antennae? Defeats the purpose of omni-directional UWB and position triangulation but you may get your 300m!
Hi @iforce2d, it would be great if you could share your code? And do you know any code written in Python which uses UWB to get distance as well?? thanks
Cool project! That is surprisingly low power transmission if you are only receiving -80dBm when you are right next to it. There must be a higher power transmit mode you can enable somehow. When you were out at 8m, with -100dBm I'm not surprised it was dropping out. Might need to dig into the settings to figure out how to configure higher power.
This technology is very exciting! Granted, it has some issues for now. I first became aware of this technology when the latest iPhone added it last year iirc. It is used to sense the direction and location of other iPhones and is rumored to be used to find tags soon (keychain tag etc). Have you heard of AGPS? When GPS is combined with one very accurate distance measurement to a fixed base station, it can be used to make GPS much more accurate. I think it's used in marine applications. Your videos are very interesting / entertaining. Especially during this time where we all have to stay home a bit more than usual. Cheers! PS: Wait, didn't you even make a video about the "assisted gps" a while back? The video with one stationary gps receiver? I must be mixing it up with something else
I was working on the same project, but for the processor I am using SAMD21G18. I am also getting range of around 8 meters, but the datasheet says it is 300 meters. Have you found any solution for that? Can I have your contacts to share the progress on the project and perhaps help each other. Thanks!
@@iforce2d They are indeed fairly spherical, but because of that that antenna has no real gain in any direction, basicly 0dB in all directions and even negativ dB in the main directions (fig. 7 elevation plane 2) (when modules lay flat and facing each other each other). I would bet the pcb antenna which get +100m has at least a little gain in the horizontal plane when holding upright (probably some donut shaped antenna pattern instead of a omnidirectional (ball shape) pattern)
I have a couple of other boards on the way, I actually ordered them in early January but they're only just finishing production - got held up first by Chinese New Year and then corona virus... super bad timing.
Thank you for a good video. I set the DW1000's Mode to LONGDATA_RANGE_LOWPOWER, but if I get more than 7M, communication is cut off. Can you guess why?
Yes, if you have enough space and you're landing under open sky it would be fine, especially if there was an RTK base station at the port. This UWB stuff might be useful if you want to have the craft fly right inside a building.
$25 doesn't seem like a good deal for something that doesn't work. According to the datasheet section you showed us, RX indicates signal is being received at the RF level, akin to putting a light on an RX pin of a wired bus. RXOK indicates that a data packet was completely received and was decoded without error. So RX is a physical layer (layer 1) receive, while RXOK is a media layer (layer 2) receive.
You know what's unreliable? Assuming things without watching a video properly. Nobody said anything about signal power, and I'm surprised you're suggesting the accuracy demonstrated here would be achievable by power measurement. Plus, why would we need UWB for that? Please watch at 5:51 or just look around one some of the sites I linked to in the description, eg. www.pozyx.io/technology/positioning-protocols-explained
This is your second comment saying the same thing, but with no useful details. Aligned how? I tried a lot of locations and orientations, just didn't film all of it. If you watch the other guy's video I showed, orientation doesn't seem to be super important.
Would enjoy a pcb creation and build video. You should add a gps to your 328p pcb design, the Quectel L70 is a very low power module with great performance (tracking 15 mA) but you have to add an antenna, the L80 module has a antenna, slightly more current but also good performance. Both are easy soldering surface mount modules. Here is some interesting reading in detailed gps performance tests. github.com/LoRaTracker/GPSTutorial/blob/master/GPS%20performance%20comparisons/Readme.md I think you will find it interesting. (Not my own work to add) Thank you, your videos a few years back sparked my interest in arduino and gsm modules and I now have a few projects of my own. 👍🏻
Yeah I'm thinking about doing a few videos on the EasyEDA process sometime. Seems like those type of videos are not very popular though, everyone wants to see a plane being made.
Hopefully you will get a chance to do a few videos on EasyEDA, I have made a few parts library’s on there, fairly easy to use once I spent a bit of time playing around with it. Not made my own 328p board yet though, I am in the middle of making a Pro mini - GPS - Sim800L carrier board, all the modules are cheap and allows easy soldering. I enjoy the plane builds/tests too, they are incredibly detailed, hats off!
