Dear forum,
I recently came across this company website and their product UWB3000F27. Does it really include official Qorvo parts? Is it possible to buy a similar device (Dw3000 + Power Amplifier) directly from the Qorvo site?
Thanks and best regards,
Christian
It does seem like it.
Adopting Qorvos chips isn’t something new in particular, there are many adaption boards for special use cases out there. As far as I know, Qorvo sells seperate amplifiers in their shop, but no combined module.
On the other side, you’d probably have to check your countries regulations on the maximum power for UWB applications, as the 500mW mentioned seems to be quite high for most countries in non-emergency use-cases.
Kind regards
Fhilb
The 500mW are definitely too high. It is interesting, so far I was going with Makerfab’s ESP32 DW1000 based modules. In my use case (indoor navigation wo GPS), I really saw a difference when I switch to the Pro variant, which apparently has an inbuilt PA. It would be nice to have an off the shelf variant with PA, like NiceRF does, which safes me some integration effort.
Anyway thanks for confirming Fhilb.
I do like that their spec is clearly just a copy and past of the underlying UWB chip specification.
They include the lines:
- Worldwide UWB Radio
- Regulatory compliance
I’d love to see their definition of regulatory compliance. It probably involves never transmitting.
Looks like they also sell a version with a small processor added that can perform TWR and basic localisation. The claim ranges up to 1 km.
Meanwhile in for those of us who need to comply with the regulations generally the parts are easily capable to hit the FCC limit without needing an external amplifier. So unless you are using a very poor antenna an amplifier on the transmit side is pointless if you need to stay under the -41.3 average/0 dBm peak limits.
Where an amplifier can help while staying within limits is on the receive side. You can boost the sensitivity of the receiver that way. But of course amplifying somethings that’s getting lost in the noise is both harder (you need a very low noise amplifier) and also less beneficial to simply blasting the signal louder to start with.
Hi Andy, thanks for your feedback. That’s what I feared to hear. My measurements I ran so far, with DW1000 dev boards from Makerfabs as well as the Decawave DW1000M eval boards, had some significant communication loss in cluttered environments ( I performed measurements in a park with trees and bushes, those weren’t in the LOS though). I barely managed to get above 30m with both boards, although the amplifier definitely helped and I could still crank up performance I guess.
Would you have a recommendation for me on the hardware/software side to increase range? I believe amplifying the receiver side could be a nice solution, but is there anything like this I can buy off the shelf?
The boards of above’s manufacturer might solve the problem, but as you said, in an official product it will most definitely hard to avoid the limits.
We managed to get 60m fairly reliably and more intermittently or in more controlled environments. That is assuming LOS. We want the best accuracy possible (we get position accuracy 3cm from truth), non-LOS situations have less accuracy so we don’t care about them. The theory is that if you have something in the way then move things until you don’t.
We got that range by using 850k data rates, the 16 bit decawave SFD and a short packet and preamble. I know they claim a longer preamble helps but I found that benefit very minimal and it slows things down making you more vulnerable to intermittent bursts of noise causing issues.
On the physical side we got the antennas away from the electronics as much as possible (we have a physical antenna above a ground plane rather than a chip or PCB trace antenna sitting on the board). We get all the antennas at least 2 meters above the ground, ideally more.
And while the antennas have a fairly uniform gain pattern we try to mount things so that the variations that do exist are working in our favour.
The DW3000 we didn’t get as good range as the DW1000, that’s probably partly the different frequency and partly we tried a different antenna since the current one wouldn’t work on channel 9 and it wasn’t as good.
Hello, did you get “60m fairly reliably” with a pair of DW3000s, or a pair of DW1000s, or one of each? We are trying to get the best range we can with a pair of DW3000s - so far seems to be limited to about 10m (with the relative orientation Phi, XZ plane, table 10, page 17 of the datasheet). We are using channel 5, maxed-out transmit power. The DW1000s led us to believe that great things were possible with the DW3000 - but the range of the DW3000s seems worse, not better.
If you have been able to get 30m or so - reliably - with a pair of DW3000s, could you share a complete set of settings?
Thanks,
Greg Huey
We normally use DW1000 but have also tried DW3000. The 3000 has a lower maximum transmit power so your maximum range if you dial the power up to max will be less. They don’t seem to have worried too much about the maximum power as long as it’s over the regulatory limit.
That said we can get several hundred meters on a DW1000 with the power dialled up to maximum. That does however require you raise everything up off the ground a reasonable amount.
Also the antennas on the DWM3000 modules aren’t great. We use our own antenna but for testing I did modify a couple of modules and replace their antennas with SMA connectors, that allowed us to use a better antenna and also separate the antenna from the electronics to get a better pattern.
In terms of settings we use 850k with the decawave 16 bit SFD pattern and 64kHz PRF. But we also use an absurdly short preamble length of 64 which everything says not to use. Increasing the preamble would probably help range and accuracy but would also slow things down a lot. For my application I’d rather have the shorter packets.
Hello!
Thanks for the quick reply.
I had foolishly assumed that the DW3000s would have a max range that was atleast as large as the DW1000s that a friend tested. I don’t actually have any DW1000s. I have a handful of iLabs RP2040 DW3000 Challenger modules. The kicker is that I don’t really have a regulatory limit to worry about on the transmit power - I set txcfg.power = 0xFFFFFFFF; and I can barely get 10m range. Are there any tricks to increasing the transmit power further? I set up some debugging data collection, such as getting the received signal power (using the formula in sec 4.7.2 of the DW3000 user manual). At 5.6m distance I have -82 dBm to -85 dBm. The latter is when the two modules face each other with the relative orientation Phi, XZ plane, table 10, page 17 of the datasheet. The former is when the two modules are azimuthaly rotated 45 deg & -45 deg relative to the line connecting the two. This asymmetry in the beam function increases when I double the distance. At 11.2m separation I see -88 dBm and -93 dBm. With the settings we have been using, I start to see significant packet loss when the received signal strength falls below approximately -90 dBm. This is with the modules on the ground.
So, I’m wondering what you have done to improve your range. From what I can tell, I don’t think your settings increase the transmit power? Or do they? You say you have desoldered the ceramic patch antennae that come with the iLabs RP2040 DWM3000s, and replaced them with SMA connectors? And then what type of antennae are you using?
Earlier you said that you keep your antennae about 2m above your electronics? In our case this will never be more than perhaps 200mm at most.
So, in summary, to get your increased range with the DWM3000s, did you increase the transmit power? Or increase the sensitivity at the receiver? If the latter, do you see the same received signal power as I quoted above?
Thanks,
Greg Huey
You can’t increase the power further. In fact in some situations maxing the power out can make things worse. If the power supply isn’t clean and low impedance enough turning the gain up to max can distort the signal and end up costing you more range than the last little bit of power gives you. Maybe try dialling it back a little and see what happens. If you look at the chart in section 8.2.2.21.2 of the user manual you can see that changing the Coarse gain from 2 to 3 really doesn’t make any meaningful difference.
Changing the antenna has two benefits - 1) the built in one is -2dB gain at best. Other antennas are better. Secondly it gives more separation. Distance from the electronics helps simply because it gives a cleaner antenna pattern, anything conductive within 30cm or so will impact the antenna pattern. Not necessary make it worse but make it less uniform. Part of the reason we used more distance was to get a more uniform pattern to reduce the orientation specific effects. We were using some generic off the shelf SMA UWB antennas, I can’t remember the exact part. For the final product we got away from the cables, they were a delay calibration nightmare, and have a custom monocone antenna sitting above a ground plane with all the electronics on the PCB below it. This gives a very uniform pattern about the vertical axis and far more gain than we get from a little chip antenna. The down side is that it’s a lot more complex from a mechanical and assembly perspective.
Antenna distance above the ground helps. Sitting right on the ground will kill your range. How far above the ground you need to be depends on the range but aim for at least 1 m. At under 2 m I start to see odd effects and small dead zones for ranges over around 30 m. Raising things up removes these or at least moves them further out.
Clock differences between the two can hurt your receive sensitivity. If you are using modules the XTAL_Trim value in the OTP should have already be set for you at the factory, make sure you are applying that value. In theory this should get loaded on start-up but it does depend on the firmware.