GPS Module Insight

JC · Post by JC » 03 July 2006, 8:49 AM

Any thoughts on GPS SiRF II vs III vs other Modules?

I have a ZX reading NMEA GPS data, without difficulty.
I am presently using a (surplus) SandPiper GPS module, whose board says "Powered by SiRF" on the silkscreen, (SiRF I ???). Rx Sensitivity: -143.5 dB-Hz.
I am not pleased with the Start Up time. It is spec'd at Hot Start 8 Sec, cold 42 Sec. In practice, outside, with a good antenna view of the sky, recently on, Hot start works within several seconds, usually. Masking the (active) antenna, even just inside a car's dash/windshield, and it can take 5 Min to obtain a GPS lock, or longer, with a current, known, sat data table.
I'm using an active antenna, and have power to the back up for storing the last sat. table info.

I'm looking at SparkFun's SiRF II ($65/$72) and SiRF III ($80) GPS modules, and wondering if I am apt to see much improvement, in practice. (Both -159 dbm).

I wondered if anyone has actual experience playing with them?
My small bluetooth GPS, (size of 2, 9v Batteries, including it's own antenna), locks on within 2-3 Seconds, always, with a stronger signal, and more Active Sats, but I do not wish to destroy it to look at it's chip set.

Anybody's thoughts on various GPS modules, before I spend more money..., would be appreciated. Although I'll make a board for some prototypes, I'm not up to RF boards, or design, or ant matching, etc. Hence I have been looking primarily at canned modules. Eventual quantity is small, very small. Thanks!

JC

spamiam · Post by **spamiam** » 03 July 2006, 10:35 AM

I, too, have used a GPS module. I think it is the sandpiper.... It seems to be an old design. It draws about 70ma and the active antenna draws more. It has a PITA 1mm type pin header for which I had to get some ribbon cable, and an IDC connector. Not especially cheap.

I have run it in my home office and I have the antenna sitting on my windowsill so it never sees more that half the sky, and probably considerably less. It takes a very long time to get lock. The elevation number always jumps around even with a stable lock. I have not used it in the wide open (yet).

I want to try to use it to make a homing robot, which can will "see" and manage obstacles. It has the capability to climb 8-10" curbs on my residential street! The suspension is similar the the Mars rovers.

I have seen the sparkfun GPS devices, and they seem better. Less power draw, for one thing.

I am not sure if the sandpiper GPS gets any extra accuracy from multiple satellite locks vs just 3 or 4. I have a Garmin Etrex and it never has a better lock than 20 feet or so, and never seems to have more than 5 or 6 satellites locked. I also have a Garmin GPS60C and it gets a lock that it says is within a few feet, and often reports nearly all channels locked in the same situations as the Etrex. Maybe it is that WAN (or whatever it is called) capability showing.

I think the SIRF-II capabilities give better accuracy... Through the WAN capability?

I am not sure what you plan on as your application, so I could not say whether it would be worth an extra $100 for the Sparkfun stuff with the supporting devices.

Another possibility is to use a handheld GPS as the data source and feed the ZX from that with NMEA output. My Etrex works really well for that! Better than the sandpiper and it does not draw down the batteries running the robot.

-Tony

GTBecker · Post by **GTBecker** » 03 July 2006, 12:29 PM

Anecdotally,

I've used SiRF-II-based RoyalTek(.com) modules in two projects. One had
a micropatch antenna and the other used an active roof-mounted Garmin
marine antenna. There is no comparison; the antenna (and its
visibility) are critically important to the receiver's performance.
Numerically, the Garmin is a lower-gain antenna and it's 30 cable-feet
away, but with its clear view it slaughters the patch - unless it was
also on the roof, I imagine. Expecting the same performance from a
patch on a window ledge might be unfair.

That said, I recently purchased an Itronix ix260+ laptop with an inboard
GPS. The GPS module is a Leadtek(.com) and is SiRF-II-based. The
module and its helical antenna are buried in the LCD lid. This GPS is
the only receiver I've ever used (and I've used many) that can track
while inside my home (if the tile roof is dry); its sensitivity is stunning.

Tom

stevech · Post by **stevech** » 03 July 2006, 23:40 PM

I've used SiRF GPS chips at work - they're OK but they had an unadmitted bug with the 1PPS output - if you use it.

for a standalone GPS receiver with integral LNA/Antenna, I like the Garmin OEM series
http://www.garmin.com/oem/

I used the '35 in a past project. Run on 12V, RS232 out.

victorf · Post by **victorf** » 04 July 2006, 2:53 AM

To get a wider base of views as to the performance of the SiRF-II modules, you might find some discussion of them in one of the Sparkfun forums. It is worth joining those forums.

Just my 2 cents.

Vic

GTBecker · Post by **GTBecker** » 04 July 2006, 4:49 AM

>... they had an unadmitted bug with the 1PPS output...

What bug is that? Link?

Tom

JC · Post by JC » 05 July 2006, 5:07 AM

Thank you, everyone, for the feedback. The laptop info above, and my BT unit, show what an optimal patch antenna design can do, indoors! Incredible.
JC

stevech · Post by **stevech** » 05 July 2006, 9:42 AM

GTBecker wrote:>... they had an unadmitted bug with the 1PPS output...

What bug is that? Link?

Tom

I have to check the "NDA" that we had to sign.

stevech · Post by **stevech** » 05 July 2006, 9:47 AM

JC wrote:Thank you, everyone, for the feedback. The laptop info above, and my BT unit, show what an optimal patch antenna design can do, indoors! Incredible.
JC

We did a lot of testing of SnapTrak's assisted GPS scheme. This is a receiver in cell phones with clues from the network as to which PNGs are likely in view "now" - since the network knows which cell site is serving the handset and knows where that cell site is.

The raw pseudo-ranges are sent to the network where the lat/lon is computed at servers using differential correction data and other techniques. This is for E911 compliance for CDMA (Verizon/Sprint) cell phones. The techniques yield decent 2D location estimates on the bottom floor of a 3 story office building. Sometimes.

"Post processing" is the technique - meaning deferred calculation (by and at servers) of the location given the pseudo-range data from an earlier time and the correction data for that time, be it seconds or days. The computational load on the handset is greatly reduced and the GPS "receiver" becomes simpler.

Many generic GPS receivers have a special mode in which you can get the raw pseudo-range data. The format/protocol is proprietary to the vendor.

GTBecker · Post by **GTBecker** » 05 July 2006, 10:01 AM

> ... the "NDA" that we had to sign on this.

SiRF has you silenced about a very popular product's unrevealed
performance failure? You're not thinking about 100mS slips of NMEA
data, are you?

Tom

GTBecker · Post by **GTBecker** » 31 October 2006, 19:12 PM

I've now spent a few weeks using two SiRFStarIII-based GlobalSat receivers alongside two SiRFStarII modules. Like night and day.

Three of the four modules use patch antennas; one is a helix. All exhibit great sensitivity but the II-based patch can't track under my roof; the others can.

More important to me, the dynamics are profoundly different. At ~3MPH, a 90-degree turn is fully reported by the III-based patch modules in two seconds (outdoors); the II-based modules can take 10 seconds to gradually show the turn at that speed, way late.

With WAAS, two receivers can provide useful differential GPS with only a few meters of separation, though certainly not at RTK precision.

SiRFStarIII is very impressive.

stevech · Post by **stevech** » 31 October 2006, 20:01 PM

WAAS - did they ever get the mid-continent (US) satellite on the air? Without these, the elevation angle to the pacific and atlantic WAAS satellites (geos) is such that they're easily obscured by mountains or buildings. Some receivers didn't seem to be forthright as to revealing when the position was based on stale WAAS data.

GTBecker · Post by **GTBecker** » 31 October 2006, 20:24 PM

WAAS - did they ever get the mid-continent (US) satellite on the air?

PRN 122 (Inmarsat 3F4) is high to the southeast here in southwest
Florida, visible all the time, but PRN 138 (an Anik, I think) is only
about 20 degrees above the western horizon, easily blocked.

It's unclear to me if PRN 122 is in the process of moving west, but it
is easy to see from here; sometimes corrections aren't being sent,
though. Still in testing, I guess.

stevech · Post by **stevech** » 31 October 2006, 21:06 PM

which WAAS GPS receivers flag whether the position was corrected using WAAS? Is this flag in NMEA formats now?

GTBecker · Post by **GTBecker** » 03 November 2006, 8:11 AM

SBAS/WAAS is just another method to deliver differential corrections to the GPS receiver. A WAAS-corrected satellite looks the same as a DGPS-corrected one.

ZBasic Forum