What radio i should aim for? suggestions

Wireless & IoT
1

Hello,

I have the following scenario:

  • Multiples (around 4 or 5) RC boats within about an area with 400m of radius.

  • I have a ground station within the previous area.

  • I want a multi-point communication, avoiding an undiscriminated broadcast.

  • I can’t not use 900MHz.

  • 100kbps baud rate.

  • UART interface, and if it would be possible, the protocol between micro-controller and radio should be something like that “Bit start, Radio Address, Payload”.

  • Use and RP-SMA antenna, i can use an adapter if it is necessary, just the antenna is placed on top of a mast.

Actually I have old radio-modems for that duty, but they are expensive, heavy and so big. And i was thinking on moving to this kind of devices:

http://www.sparkfun.com/commerce/produc … ts_id=8710

http://www.sparkfun.com/commerce/produc … ts_id=8768

So… I’'m little confused between version 1? 2.5? what’s the best choice for my application?

I was reviewing the data-sheet of version 1, and i think i can use it with the communication protocol that i described before.

But first i want to know your opinion xDD.

About the project, it is a hobby project to do cooperation between boats, but the electronics I’m currently using are from my UAVs and i don’t need many of this stuff for boats.

Thank you very much in advance.

2

I have used the Xbee’s and found them very simple to use, but very volatile (had over 6 out of 8 die in days (3 were new replacements and one died on the dev board provided by Digi themselves) before scrapping them totally and moving to Nordic.

Digi release a new version every day ending in ‘y’, but behind it all there are only 2 hardware versions to be concerned with, 1 and 2.5. 2.5 has meshing support so it scraps the idea of co-ordinator and endpoints/routers, and instead the only node in a network is a universal mesh node.

I personally only ever used 2.5 but never used the meshing architecture, it can be used almost identically to the v1 if you wish, just using normal AT Co-ord and Endpoint firmwares.

However, I am just in the process of replacing a Xbee setup with a nordic nRF24L01+ setup, all ready for implementation tomorrow if all goes well. Learning the ANT protocal for the nordics is a lot more long winded than learning to program a Xbee in 5 minutes, but you also gain a lot more understanding and power of flexibility with the nordic chips, and you are at a much lower level and can understand exactly what is going on.

Either way both of these will suit. Perhaps start with Xbee like I did to get the system up and running, and then when you know everything is working, replace out the wireless section of the system with the nordics. They are also much cheaper once you get to bulk orders than the Xbee, and smaller size.

3

angelsix:
I have used the Xbee’s and found them very simple to use, but very volatile (had over 6 out of 8 die in days (3 were new replacements and one died on the dev board provided by Digi themselves) before scrapping them totally and moving to Nordic.

Digi release a new version every day ending in ‘y’, but behind it all there are only 2 hardware versions to be concerned with, 1 and 2.5. 2.5 has meshing support so it scraps the idea of co-ordinator and endpoints/routers, and instead the only node in a network is a universal mesh node.

I personally only ever used 2.5 but never used the meshing architecture, it can be used almost identically to the v1 if you wish, just using normal AT Co-ord and Endpoint firmwares.

However, I am just in the process of replacing a Xbee setup with a nordic nRF24L01+ setup, all ready for implementation tomorrow if all goes well. Learning the ANT protocal for the nordics is a lot more long winded than learning to program a Xbee in 5 minutes, but you also gain a lot more understanding and power of flexibility with the nordic chips, and you are at a much lower level and can understand exactly what is going on.

Either way both of these will suit. Perhaps start with Xbee like I did to get the system up and running, and then when you know everything is working, replace out the wireless section of the system with the nordics. They are also much cheaper once you get to bulk orders than the Xbee, and smaller size.

Thank you very much for your answer angelsix.

I checked out the Nordic chips and their features are amazing, i will test them in another future projects, (an estimation about its operation range?) But at this moment the UART is mandatory as i want to use an UART for radio-link (because i have a DMA implemented already), and I would prefer don’t spend more time on SW.

What’s the reason for this lack of reliable in X-Bee? overheating? or just they die without a reasonable explanation.

Did anyone have a bad experience too?

4

Ha no the second reason. I got 2 replacements a day later, plugged it into their dev board, connected and 20 seconds later that died too. A few survived and are in use now but I lost all faith in them as a module. You only have to google “Xbee bricked” to get 100s of results.

As for USART, once you have done your connection type stuff with the nordics, you can do UART in one line of code and a PIC in between them, as all you do is use the PIC to setup the nordic, then listen on its USART pins (connected to your own PIC) and pass that USART data to the TX payload, and any incoming from the nordic pass back up the USART to the PIC, its rather nice, and what I plan to do tomorrow. I have just been spending the last few days writing and debugging my own assembly library for the nordic chip as the only ones available are C at present. I plan to release my library once tested so your welcome to that.

5

Hmm. I don’t believe that XBees are nearly as destructible as you’re making them out to be. I haven’t had any trouble with them, and heck, I fly 'em in rockets. I’ve had one survive an 80+ MPH crash into the hard ground.

In addition, almost all of the Google results for “XBee bricked” have to do with XBees made useless by attempts to change the firmware, or by incorrectly messing with the settings. While that’s disturbing, it’s very different from XBees being simply destroyed by being plugged into circuits.

If your XBees (angelsix) simply stopped working after being plugged into a board (with no attempt to program them), I’d think it much more likely that you got a defective dev board, or there was something wrong with your power supply.

I don’t know that 2.4GHz XBees are going to give you (Noether) the 400 meter radius range you’re after, but other than that, I think they fit the bill pretty well. I do think you’ll be happiest with Series 1 XBees, but maybe that’s just my simple thinking.

6

noether:
… snip …

  • I can’t not use 900MHz.

  • 100kbps baud rate.

… snip …

What does first criteria mean? You cannot use 900MHz - or you’re forced to use 900MHz? XBees are available in either frequency.

Regardless, I’m not sure you can get 100kbps between XBees - most I’ve heard is 57k - and that might be wishful thinking. I may be wrong but I’d check further into this limitation if it is a sticky point for your project.

I don’t know about the Nordics or others.

7

mitchind:

noether:
… snip …

  • I can’t not use 900MHz.

  • 100kbps baud rate.

… snip …

What does first criteria mean? You cannot use 900MHz - or you’re forced to use 900MHz? XBees are available in either frequency.

Regardless, I’m not sure you can get 100kbps between XBees - most I’ve heard is 57k - and that might be wishful thinking. I may be wrong but I’d check further into this limitation if it is a sticky point for your project.

I don’t know about the Nordics or others.

900MHz is forbidden in Spain (GSM Band) xD.

57kbps could be acceptable without a exhaustive log from the Ground Station. Thank you very much for that point.

angelsix:
Ha no the second reason. I got 2 replacements a day later, plugged it into their dev board, connected and 20 seconds later that died too. A few survived and are in use now but I lost all faith in them as a module. You only have to google “Xbee bricked” to get 100s of results.

As for USART, once you have done your connection type stuff with the nordics, you can do UART in one line of code and a PIC in between them, as all you do is use the PIC to setup the nordic, then listen on its USART pins (connected to your own PIC) and pass that USART data to the TX payload, and any incoming from the nordic pass back up the USART to the PIC, its rather nice, and what I plan to do tomorrow. I have just been spending the last few days writing and debugging my own assembly library for the nordic chip as the only ones available are C at present. I plan to release my library once tested so your welcome to that.

Yeah, i thought about it, but i want to keep as simple as possible all the electronics and software. Currently I’m using ARM7 controllers, I’m not used with PIC, and another ARM7 for only SPI<->UART is wasting resources imo.

sylvie:
Hmm. I don’t believe that XBees are nearly as destructible as you’re making them out to be. I haven’t had any trouble with them, and heck, I fly 'em in rockets. I’ve had one survive an 80+ MPH crash into the hard ground.

In addition, almost all of the Google results for “XBee bricked” have to do with XBees made useless by attempts to change the firmware, or by incorrectly messing with the settings. While that’s disturbing, it’s very different from XBees being simply destroyed by being plugged into circuits.

If your XBees (angelsix) simply stopped working after being plugged into a board (with no attempt to program them), I’d think it much more likely that you got a defective dev board, or there was something wrong with your power supply.

I don’t know that 2.4GHz XBees are going to give you (Noether) the 400 meter radius range you’re after, but other than that, I think they fit the bill pretty well. I do think you’ll be happiest with Series 1 XBees, but maybe that’s just my simple thinking.

Thank you very much for you feedback. Yeah, I think so that the problem with the angel’s X-Bees are not the radio themselves, and maybe the problem is in supply and/or dev board.

8

Lets just call it bad luck but I had the entire dev kit replaced twice from Digi themselves and only 2 of 10 survived, but like I say those 2 are working to this day, but my confidence in them is long since gone. Not had any issues with nordics at all yet.

Had so many issues with Xbees and even this forum if you search for problems a lot of people have them, people reply presuming its the firmware issue, and then the original poster never replies as to whether it was or not (probably gone off and brought something else).

Hard to see 2 dev kits and boards from Digi being faulty. Perhaps bad USB ports on my laptop to blame maybe? But I use a PICKit 3, a lot of home made SMD breakout boards powered from my USB when testing, and other sensitive wireless modules, and not one of them has ever had a single issue, the only common place is the Xbee module, so thats enough to put me off.

The names Luke by the way for all those wondering :stuck_out_tongue:

9

angelsix:
Had so many issues with Xbees and even this forum if you search for problems a lot of people have them, people reply presuming its the firmware issue, and then the original poster never replies as to whether it was or not (probably gone off and brought something else).

I understand that having several XBees stop working will put you off of them. At the same time (and not to be argumentative, but…) I’m not seeing those problems that you’re talking about, either in the Sparkfun forums or through a Google search. I tried “XBee bricked”, “XBee dead”, “XBee problems” and “XBee fried”, and didn’t find anything at all outside of the programming problems I mentioned.

I use XBees a lot, and if they really are that fragile I’d like to know about it, but I can’t find any posts other than yours that suggest that XBees are dying randomly. I do still think it’s a problem in your specific setup.

I do find it disturbing that XBees can so easily be programmed in ways that brick them (or require elaborate and cryptical incantations to restore them to usefulness*), but frankly I don’t see any sign that XBees are unusually prone to just dying for no reason. If you’ve had that problem, of course I don’t blame you at all for not trusting them again. I’m sure I’d have decided they were defective too, if I were in your shoes. But I can’t find anything that would make me think it’s a widespread problem.

  • On the other hand, I imagine that making them immune to this problem would mean preventing us from using much of their functionality. I think it’s better to give us control even if it means we might break 'em.
10

Yes I agree if they were so fragile that certainly wouldn’t be selling many. Perhaps I just had the last kits of a bad batch, or I happen to be very unlucky. I just can’t see the environment or my “setup” being an issue when the only non-Digi objects were a USB port alone, everything after the power coming from a USB port was Digi boards and modules.

They certainly didn’t hassle about sending a new kit out for the old, and customer service was not an issue, its just like many say, once bitten. But on the other hand I’m glad it happened because in comparison the nordics are far more powerful (or at least expose you to far more control) than that or a Xbee, and at a cheaper cost, smaller size, and no daft 2mm socket pins.

I’m going to hopefully finish my nordic wireless test off tonight so will let you know the results and do some range testing with the SMA antenna version compared to the chip antenna etc…

11

Fair enough, and you’ve raised my interest in the Nordic RF stuff enough that I’ll probably play around with them early next year. It’s always good to have more options.

12

I have 15 or so XBee series one - used them, seriously abused them physically, never applying the wrong voltages. Never have had one fail, in years of use.

My issue with Nordic is that it’s proprietary. Not everyone cares. If you need to change suppliers, you can’t, whereas with 802.15.4, there are many module suppliers, with already-type-certified modules.

13

Does nordic not use the ANT protocol which is very well backed and supported, and in fact it is the Xbee that suffer worse from this if they went bump? Thats what I read and understood to be true and another reason I decided to look at nordics over Xbees, not the other way around as you say?