Author Topic: indigogo : NavSpark - gps and a risc on a tiny board for 15$ (if you buy 2..)  (Read 5536 times)

0 Members and 1 Guest are viewing this topic.

Offline PolossatikTopic starter

  • Frequent Contributor
  • **
  • Posts: 295
  • Country: be
just came across this

http://www.indiegogo.com/projects/navspark-packing-32bit-features-and-gps-gnss-into-8bit-size-and-price/x/533586

dev is done by SkyTraq which is ,according to this blog, a fabless semiconductor company, which develops GPS chipset for consumer navigation and tracking applications
they have 3 boards

15$ "NavSpark" advertised as  "low cost GPS / GNSS board with a 32-bit LEON3 SPARC V8 MCU that can be programmed with the Arduino IDE. "
17$ "NavSpark-BD" is  GPS + Beidou
17$ "NavSpark-GL" is GPS + GLONASS

difference seams to be in the binaray blob they use, they claim to "NavSpark hardware design files (schematic, board design, and bill of materials) will be released and made available to users. Source code of the customized Arduino IDE will also be made available; the GPS / Beidou / GLONASS navigation kernel library integrated will remain in library file format. Users are free to extend, adapt and redistribute the hardware and software for any projects, commercial or non-commercial. "

anyway, even if they are still a long shot away of funding and not sure if there is much "headroom" left in the main "100MHz 32bit RISC/FPU," after all the gps processing,
but hey,  for 17$ it might be a nice thing to dabble with if this comes of the ground.

they are bit stupid, the one piece items are only "early bird" solde out and next up you need to get 2
not sure why they don't add perks for 20usd/ 22 used or so for 1
« Last Edit: January 03, 2014, 10:07:24 pm by Polossatik »
Real Circuit design time in minutes= (2 + Nscopes) Testim + (40 +120 Kbrewski) Nfriends

Testim = estimated time in minutes Nscopes= number of oscilloscopes present Kbrewski = linear approx of the nonlinear beer effect Nfriends = number of circuit design friends present
 

Offline fake-name

  • Regular Contributor
  • *
  • Posts: 75
I've worked with the Skytraq Venus GPS modules before. They're definitely a real company.

Hell, you can buy skytraq parts from sparkfun.

I'd suspect the difference between the different versions is in the main chip, which I suspect isn't a single-die part, unless they can fit a GPS LNA (though this has a external LNA as well. the Skytraq Venus has a internal LNA) and a decent CPU into the same fab process without driving their costs through the roof.

I'd guess they have different chips they stuff the boards with, rather then just different firmwares.

---

For anyone who hasn't heard of Beidou, it looks china has remained the COMPASS satnav system Beidou.
 

Offline JBeale

  • Frequent Contributor
  • **
  • Posts: 298
Looks to me like this SkyTraq NavSpark GPS is a two chip solution. The GPS LNA & receiver is the Skyworks SE4150L  http://www.skyworksinc.com/uploads/documents/202445A.pdf and the baseband processing is done by the Venus 822 which uses the open-IP 32-bit CPU design, LEON3 Sparc-V8 with FPU.

Possibly as a result of my request yesterday, they just announced a new feature today, time-stamp an external input signal:
Quote
Oliver Huang said 14 hours ago
Regarding time stamping on external trigger, we’ll modify hardware to support this. For NavSpark time accuracy is ~60nsec due to 16.368MHz clock. For -GL, -BD, time accuracy is ~40nsec due to 24.552MHz clock.
The 40 or 60 nsec resolution timestamp is not amazing, but it is linked to GPS time so it will not drift relative to the international UTC timescale.  This part would be the cheapest way to achieve this, that I'm aware of.  They achieved their funding goal, so presumably the project will now actually happen.
« Last Edit: January 22, 2014, 08:51:16 pm by JBeale »
 

Offline JBeale

  • Frequent Contributor
  • **
  • Posts: 298
There is quite a bit of detail about the board that is scattered through the "comments" section on the Indiegogo page. I made an unofficial FAQ summarizing what has been explained there, for easier reference.

NavSpark FAQ: http://pastebin.com/1f5iXwiT
 

Online amyk

  • Super Contributor
  • ***
  • Posts: 8232
How many channels is the GPS receiver and what's the expected TTFF?
 

Offline JBeale

  • Frequent Contributor
  • **
  • Posts: 298
from http://www.indiegogo.com/projects/navspark-arduino-compatible-with-gps-gnss-receiver/x/6094574
"The Venus 8 chip is built with enough MIPS and memory to process signal from combination of 34 GPS, GLONASS, Beidou, and Galileo satellites in parallel simultaneously. It is the reason why it has more memory and computation MIPS than most popular low-cost embedded microcontrollers on the market."

Of course, in practice you are limited by how many satellites are above your local horizon. The web page has a screenshot showing tracking of 12 GPS and 8 GLONASS satellites.  The same screen capture shows TTFF: 18 but I don't know what that means exactly, and if it is cold start, hot start or what.

I wasn't expecting amazing performance given that it is, as far as I know, the cheapest GPS available anywhere. That might be the biggest selling point.
« Last Edit: January 29, 2014, 05:12:49 am by JBeale »
 

Offline JBeale

  • Frequent Contributor
  • **
  • Posts: 298
Oliver replied to the question:
Quote
NavSpark TTFF is 30sec average under open sky. 34 tracking channels not fully used. Some GPS claims 48 tracking channel while showing no more than 16 signal; what difference between 16, 34, 48, or 100 tracking channels to an end user? :-)
 

Online amyk

  • Super Contributor
  • ***
  • Posts: 8232
 

Offline JBeale

  • Frequent Contributor
  • **
  • Posts: 298
The GlobalTop Gmm-g3 looks like a nice module. It is slightly more expensive than NavSpark and it does not come with an antenna. Not clear if it is user-programmable, and for one-off hobby projects it is a more difficult to connect to, as the connections are only SMT pads.  But if you're building your own PCB, that module could give a smaller total design.

Note from Indiegogo site: apparently NavSpark is not meeting their volume expectations of "Digispark" proportions (although it did hit the nominal funding level) so the price will go up after the funding campaign.
« Last Edit: January 30, 2014, 01:45:43 am by JBeale »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf