Long Range, Low Power Wireless module for IOT: AWESOME!!!

[usgine2003]

I don't know if you guys are into IOT, but this module is the BOMB!
More than 4miles range, 5years on 2AAA batteries, Serial, Digital and Analog ports, works with lots of sensors:
Check it out here on Kickstarter

[tonyarkles]

Hmmm... New user, talking about a Kickstarter... Methinks you might have something to do with it

At any rate, this *is* pretty interesting to me.

[LukeW]

If you're going to post here to self-promote your crowdfunding project, at the very least you should post in the correct crowdfunding forum.

OK, so, you're working in the sub-gigahertz UHF ISM spectrum... probably using something like a Semtech SX1272 or similar?

My first question would be... why are you using RPSMA? IMO, you should be using SMA, which is a more standard connector for RF systems - RPSMA was introduced on consumer-grade WiFi devices as a deliberate "standard breaking" move to comply with FCC requirements to stop end users from swapping the antenna for a higher-gain one. If you were marketing consumer WiFi routers then RPSMA would make sense as "the standard" within that narrow market but I don't think it's the best choice here, since you want to be able to connect antennas or cables for 800/900MHz ISM use. Yes, RPSMA connectors and adapters are readily available these days, but I think it's just an extra level of inconvenience you could easily avoid.

[LabSpokane]

You know what's not awesome?

- shilling for your kick starter in the wrong place, twice.

- not specifying the operating frequency. I can only assume that this runs in the 902-928 ISM band.

- trying to sell a radio with bullshit/obfuscated specs. Four miles typical range with a 100mW radio?  Under what conditions?  Using what antennas?  With what fade margin?  Nobody it's going to get 4 miles using the 0dB omnis you show except under extraordinary circumstances with unobstructed line of sight. Foliage and buildings attenuate 900 MHz. It does not follow terrain like lower frequencies do.

- -160 dB link budget?!  Really?  That's better than a $1500 Calamp/Dataradio.

- not providing a proper spec sheet for your radio. There's plenty of them for the ISM band. Pick one and use it as a template.

- then there's the power budget. I'll save that for another day.

[Psi]

i really hate RPSMA with a passion.

[LabSpokane]

Oh, and by the way usgine2003, change your certification list to "pending" since you apparently have exactly none of them approved yet.

Oh and "unbelievable" really sums up your claims on transmission range.  That's the one accurate spec I could find on your kickstarter page.

[dr.diesel]

1 of 7 selling features is a unique MAC?   

[LabSpokane]

1 of 7 selling features is a unique MAC?   

Only made more bizarre by the fact that it is not an IP-capable device. 

[hamdi.tn]

Well Xbee PRO claim to have 25 miles ... so nothing special about 4 miles (outdoor)
As many said , those certification clearly still not approved .
nice work though 

[colotron]

...More than 4miles range, 5years on 2AAA batteries, Serial, Digital and Analog ports, works with lots of sensors:

5 years on 2AAA batteries ... didn't AAA batteries self-discharge in a quite shorter period of time? (the low-self discharge ones promote 30% in 1 year: link)

[LabSpokane]

Well Xbee PRO claim to have 25 miles ... so nothing special about 4 miles (outdoor)
As many said , those certification clearly still not approved .
nice work though 

The Xbee Pro is a 24 dBm device.  This device is 20 dBm. So the Xbee has 2.5X the power output. Also, the Xbee specs clearly state that their range claims are for outdoor, clear line of sight, with high gain antennas.

The Xbee's urban specs are properly reduced to 2000 feet or less - which is how most of these devices will be used.

And the Xbee has a proper data sheet that states a maximum sensitivity of -110 dBm, not some crazy crap about a 160 dBm link budget.

[JBeale]

> 160 dBm link budget
I wonder if that is possible with suitable processing gain. As I recall the GPS signal enjoys +43 dB gain from a despreading factor of 20460 (2.046 MHz down to 100 Hz). So if you were willing to have extremely low data rates...

[magetoo]

The kickstarter says "last October, we were introduced to a new radio technology called LoRa that boasted unbelievable range performance."

That seems to be this thing: HopeRF LoRa modules.  You can get some HopeRF transceiver modules from Sparkfun/Adafruit (pretty cheaply too) so that part doesn't seem too crazy at least.

[LabSpokane]

The kickstarter says "last October, we were introduced to a new radio technology called LoRa that boasted unbelievable range performance."

That seems to be this thing: HopeRF LoRa modules.  You can get some HopeRF transceiver modules from Sparkfun/Adafruit (pretty cheaply too) so that part doesn't seem too crazy at least.

No.  It just keeps getting worse the more one reads. The output power is only 20 dBm in *single* frequency mode with a special setting for the spectrum spread. For a radio that could possibly be licensed in the US, which would have to be frequency hopping spread spectrum (FHSS) the output is restricted to 13 dBm, or 50 mW.

Since the Kickstarter is US based and plans to license there first, this is a valid criticism. And the fantasy link budget of 160 dB was overstated by nearly double. The spec sheet lists it as 157.5 dB.

Talk about cherry picking your data... Nothing says fraudulent quite like quadrupling the actual output power (3dBm + 2.5dBm of rounding up)  while using stats from a mode one can't get licensed in.

Usgine2003,

Do everyone a favor and end the campaign now.  If this garbage is what you're submitting to the FCC, you'll never get approval. Nor should you.

[magetoo]

LabSpokane, where do you get the 160dB figure?  I'm only finding it in the "Technical Overview" (which doesn't really talk about the radio side of things at all).  I don't think you can grab tech specs from those slides.

I'm looking at their datasheet (under "Whisker Engine Technical User's Manual" on the Kickstarter), and I'm seeing a figure of 132-139 dB for the link budget (also 13-20 dBm transmit power).  And HopeRF has a third set of wildly optimistic numbers for their modules.


None of that is important, though - they're selling their idea on the claims of "4 miles range" and "5 years battery life", not the details of the technobabble for a component they are not even building themselves.

I'm not an RF person, but four miles on low power and modern error correcting codes doesn't seem too crazy.

[LabSpokane]

Magnetoo,

I used to design ISM band mobile networks for a living.   What they are advertising bears no resemblance to the actual performance of such a radio in that band. If this was 150-174 MHz licensed band, I would have a completely different opinion. I put in 1W (30 dBm) radios with antennas 100+ feet up on a tower with high gain antennas and these propagation claims are just silly.

I pulled the 160 dBm off their presentation. If it's not a real number, the shouldn't claim it on a document with their logo on it.

The battery life is another pile of horse pucky. For such a radio to have that kind of battery life, it is virtually a simplex device since it has to stay asleep. That radio consumes 10 mA in order to detect an incoming carrier wave and at least 90 mA to transmit. And that fact is conveniently omitted when talking about the battery life. There's 8760 hours in a year and if they have an 87.6 amp hour set of AAA batteries, they should get out of radios and start a battery company. Maybe that radio's uC can sleep for 5 years, but send a meaningful amount of data 4 miles?  Hardly. If it worked for six months to a year on a set of batteries, it would still be spectacular and possible.  But online campaigns are the land of silly promises, so here we are in la la land.

[magetoo]

I used to design ISM band mobile networks for a living.

Ok, I'll go with your take on the radio part of this project then.

I pulled the 160 dBm off their presentation. If it's not a real number, the shouldn't claim it on a document with their logo on it.

It's obviously inflated by marketing need for impressive figures, naive arithmetic, and boundless optimism.

As for the ultimate source, it's not really spelled out anywhere I could see, but I'm assuming it is the best case scenario at maximum everything + whatever extra headroom the error correction gives. (So 20 dB or so?  Considering what goes into reading blocks of data off spinning rust and talking to robots on other planets, that doesn't seem implausible.)

The battery life is another pile of horse pucky.

No argument there.

[yramgu]

Reading this thread i'd like to clarify some things about LoRa modulation (I hope i'm not too off-topic):

I've been working on it for a few months for my company and yes, 160dB link budget IS possible (you can get a -140dBm sensitivity @SF12/125kHz/CR4). LoRa modulation is not like your standards GFSK. It's also not a real spread-spectrum modulation (unfortunately details are confidential) and it's REALLY resilient to interference.

During my tests (at 868MHz as i'm in Europe) we were able to achieve a ~12km range with an output power of 14dbm (25mW) and a 1% PER. To achieve this performance our radio wasn't even placed on high ground: we were holding it in our hands.

EDIT: tests done in a non-line-of-sight situation. So a LOS situation woud be even better

So to get back to the product that was advertised:

- 4 mile range with 20dBm : that's a very poor performance for LoRa
- 5 years autonomy with 2*AA batteries : yeah right.
- they claim to have achieved maximum sensitivity (maximum link budget): unfortunately we can only take their word for it.
- BEWARE of the +20dbm setting: the datasheet of sx1272 / sx1276 specifies that you must have a VSWR < 3:1  otherwise you burn the chip's PA...

That was just some information for the people saying a 160dB link budget is impossible/range specs aren't realistic, etc... Technology is evolving... Sigfox for example achieves similar performances with a completely different approach.

[LabSpokane]

Reading this thread i'd like to clarify some things about LoRa modulation (I hope i'm not too off-topic):

I've been working on it for a few months for my company and yes, 160dB link budget IS possible (you can get a -140dBm sensitivity @SF12/125kHz/CR4). LoRa modulation is not like your standards GFSK. It's also not a real spread-spectrum modulation (unfortunately details are confidential) and it's REALLY resilient to interference.

During my tests (at 868MHz as i'm in Europe) we were able to achieve a ~12km range with an output power of 14dbm (25mW) and a 1% PER. To achieve this performance our radio wasn't even placed on high ground: we were holding it in our hands.

EDIT: tests done in a non-line-of-sight situation. So a LOS situation woud be even better

So to get back to the product that was advertised:

- 4 mile range with 20dBm : that's a very poor performance for LoRa
- 5 years autonomy with 2*AA batteries : yeah right.
- they claim to have achieved maximum sensitivity (maximum link budget): unfortunately we can only take their word for it.
- BEWARE of the +20dbm setting: the datasheet of sx1272 / sx1276 specifies that you must have a VSWR < 3:1  otherwise you burn the chip's PA...

That was just some information for the people saying a 160dB link budget is impossible/range specs aren't realistic, etc... Technology is evolving... Sigfox for example achieves similar performances with a completely different approach.

Hi yramgu,

That type of performance is truly extraordinary.  And I'm ready to stand corrected, because that would be an incredible improvement to what I've used thus far.  Can I ask some questions?

- What is the fade margin on on your 12 km link? 
- What mode was the radio operating in?  Single channel, or FHSS?  (In the US, we use FHSS between 902 and 928 MHz in order to minimize interference and that degrades the link performance.)
- Was the test area an urban area or rural? 
- How long was the link active?  Hours, days, months? 
- Was there an elevation difference between the radios? 
- Was either radio near an object such as a metal building that might create constructive interference? 
- Was the NLOS link obstructed by structures?  Trees?  Was it summer or winter (when there is less foliage to attenuate the signal)?
- What type of antenna and what gain?
- What is the noise floor of the area you were testing in?

Thanks!

[Marco]

LoRa modulation is not like your standards GFSK. It's also not a real spread-spectrum modulation (unfortunately details are confidential)

Why is it confidential if the patent is already published? (Strangely enough I can not find any patent from the Cycleo time frame though.)

[LabSpokane]

I'm looking at SigFox right now.  And it does look like a nice solution, but it's really not a comparable device and that's how they're getting such good range and performance.  It operates at 100 bps and has a limit of 140, 12 byte messages per day over a proprietary network. 

I should have noted the devices I was working with were 256 and 512 kbps devices, so obviously they would suffer from attenuation more severely than a very low bit rate device. 

[Howardlong]

Folks, if you're going to start quoting x dBm you're going to need to quote channel bandwidth too, otherwise it's of little relevance when talking about link budget. A really low bandwidth signal can go an awfully long way without much power when the channel bandwidth is tiny, although assuming this is using consumer grade parts the frequency tolerance won't support channel bandwidths less than about 10kHz at the usual ISM frequencies.

FWIW, line of sight, 20dBm/100mW EIRP with a 10kHz NBFM channel, 1000km is easily possible on analogue voice. It's done regularly by radio hams operating through satellites, and I've frequently done it myself. Indeed 100mW is the low power beacon on a couple of satellites where I developed the coding scheme, and you can pick that up with a 0dBi antenna, the channel bandwidth is about 1.2kHz and the final bit rate once you've stripped the FEC off it is 400bps, typical distance to the satellite being 1000-2000km.

Having said that, I would imagine the alleged range is because it relies on a mesh network. And yes, I do believe that the reality is that the product might just be able to do all that's claimed, but not necessarily at the same time, nor reliably, and even then there will need to be just the right tongue angle.