PDA

View Full Version : Satellite Locking



ALAALA
07-24-2009, 06:46 AM
This morning I, as usual when I wear both, turned on the 310 and the 305 virtually simultanously and put them together on the windowsill while getting ready for a run. I turned both of them on using down arrow + on (though I'm not sure it matters for the 310).

Anyways, the 305 took its usual time (not very long), but the 310(!) took a long time, and finally complained and wanted to now whether I was indoors or had moved far away since last. I said no to both (technically I was of course indoors in my apartment, but that has never been a problem for neither the 305 nor the 310).

This was a disappointment. I certainly don't expect the 310 to be worse than the 305 in locking satellites.

Perhaps one should NOT hit down arrow while powering on the 310, and let the surf chip ride the those waves freely without intervention?

NDURANCERIDER
07-24-2009, 06:55 AM
Using the down arrow when turning on the unit will dump the satellite cache and force it to find satellites from scratch. Same for the 310 and the 305.

Strange that the 310 took longer though.

April

KANSAIBEN
07-24-2009, 08:37 AM
Is there some reason you clear the sat cache on a usual basis or was this some kind of test you were performing? Kind of defeats the purpose of the sat cache (speedier sat locks)...

SEILOGRAMP
07-24-2009, 08:53 AM
I've only done the satellite reset once when comparing the 205 to 310. That was only because of issues the 310 was having in tracking my activity. Since then it has been fine and I have not reset it again.

From the old forum (erroneous question marks included, just for old times sake ;))


Quote from: seilogramp on June 24, 2009, 08:08:28 pm
Quote from: fourbeer on June 24, 2009, 08:05:19 pm
That 310xt track does not look very good.? I was hoping for at least 205 performance.? Bummer.

Yeah, I'm hoping this can be improved with a reset of the satellites on the unit. Also, I'm not sure if the GPS chipset can be tweaked for better performance.

Today was much better for the 310XT. I cleared the cached satellites on both units, not being totally sure that the same reset combination (hold down arrow while powering on) would work on the 310XT. Both units registered roughly 15 - 16 feet GPS accuracy. Will report details later.

Also, from the old forum...


I just now played around with sat acquisition with the 310. Not sure how you had your 310 resting. As you know it's hard to get it to balance with the band in the way. If you have it just resting on its own, it will not be at its best angle for reception. The x05 is at a better angle just resting. With the face of the 310 parallel with the earth I got 18' accuracy, with it perpendicular to the earth I got 20', and with it at a 45 degree angle I got 15' accuracy. This is good, since the normal running position is going to be with the watch tilted at ~45 degree angle.

DW152
07-24-2009, 09:30 AM
Anecdotally, my 310 seems to acquire satellites faster than my 305 did, but I don't do the side-by-side comparisons that you do so I can't say for sure. If I have time this afternoon I could do a comparison.

So far, I've only had one bad experience with losing satellites:

http://connect.garmin.com/activity/7934318

I'm thinking that when it lost signal, it was only the third or fourth time I had used it, so maybe I had never cleared the cache, but I don't remember.

Since then I've run through more tree-covered paths without any problems. I've also started monitoring my GPS accuracy while I run, and it varies from ~14 feet (wide open skies) to ~30 feet (tree covered trails), which the typical range being ~22 to 28 feet. I don't know how this compares to my FR305, but it's better than I get with my Edge 305, which is typically ~30 feet but will often creep up to 60 feet or higher if I'm riding my road bike down a tree-covered road while in the drops with my body blocking the unit.

DW152
07-25-2009, 07:53 PM
I repeated your experiment, but I was outdoors instead of indoors. I did it two or three times, and in each case, the 310 locked onto the satellites faster than the 305, and had reached better accuracy than the 305.

After doing that test, I went for a trail run. I set one of my screens to monitor GPS accuracy. Almost my entire 7-mile run was in dense forest, and I was down in some low spots with small hills around me. GPS accuracy was was usually better than 23 feet, and I don't think I ever saw it greater than 26 feet. So, ever since I had my bad experience the second or third I used my 310, it's been solid ever since.

One thing I've noticed is that when I've had GPS failures, it's occurred shortly after I received the unit. This actually happened to me yesterday during a mountain bike ride with an Edge 305 I had just received as a warranty replacement for my other Edge (which had developed a broken lap button). After my ride, I did a complete reset, so hopefully it will work better next time. Anyway, I'm starting to form a theory that whenever we receive a new device, we should first do a complete reset to clear everything.

SEILOGRAMP
07-25-2009, 08:00 PM
I repeated your experiment, but I was outdoors instead of indoors. I did it two or three times, and in each case, the 310 locked onto the satellites faster than the 305

It should. The 310 uses SiRFInstantFixII (http://www.sirf.com/products/location2_services.html).

ALAALA
07-26-2009, 07:59 AM
Same problem today. 305 locks normally, 310 looks, and looks, and looks, and finally gives up and asks me if I'm indoors, etc.

I'm doing a hard reset today. No problems, since there's nothing in memory that I haven't got elsewhere.

SEILOGRAMP
07-26-2009, 08:24 AM
I'm doing a hard reset today.

Is this your first hard reset since getting your 310? Have you just been doing the satellite reset?

KANSAIBEN
07-27-2009, 02:40 PM
What firmware version and GPS version does your unit say it has on it? (Settings//About Forerunner)

ALAALA
07-27-2009, 11:41 PM
Is this your first hard reset since getting your 310? Have you just been doing the satellite reset?

Yes, it became the first time I did a hard reset (Mode+On).
This last time I didn't hit down button on either the 310 or the 305.

ALAALA
07-27-2009, 11:43 PM
what firmware version and gps version does your unit say it has on it? (settings//about forerunner)

2.40/2.10

KANSAIBEN
07-28-2009, 04:38 PM
If the hard reset didn't work, you'll want to be in contact with Garmin CS as your software numbers look current. Could be you may have corrupt GPS software and a reinstall may help, or you may have a faulty antenna and need a replacement.

ALAALA
07-29-2009, 12:50 PM
Well, yesterday, and today, the 310 at least didn't look blindly, but I must say that it isn't much, if at all, better than my 305 either in finding sats fast or keeping them.

In fact, I've noticed that the 310 seems to lose itself more often than the 305. The units go side by side, so to speak, for some kms but sometimes the 310 has jumped 40-50 meters ahead in quite a short time. I'm not so sure it's about how the GPS track is laid out as much as how the internal correction algorithms works.

I'd be interested in knowing if anyone has found something similar?

I certainly hope its a software problem (if it is a problem at all, and not just pure coincidence due to the aligment of the stars 2543,57 light years to the left of the far side of Endor), because then it should be possible the fix without messing with returns.

SEILOGRAMP
07-29-2009, 04:41 PM
I'd be interested in knowing if anyone has found something similar?

Nothing quite as bad as you've experienced. I did not get the 310 expecting it to be better at tracking. In my comparison tests, it is at least as good as the 305. I did expect it to lock on faster. It at least does that.

ALAALA
07-30-2009, 06:58 AM
Apropos the 310's GPS track layout. This image is of the unaltered GPS track from today's interval training session at a certified 400m track (of course lane 1, slick to the inner edge). All workbouts and rest intervals started/stopped either exactly at the start/finnish line or the 200m mark.

As you can see the laps drift farther and farther away. The 310 (GPS on and set as speed source, footpod present for cadence) measured 6,10 km (should be 6,00 km (15 laps)), which isn't atypical from what the 305 use to measure.

But, the GPS track itself looks better, and more true to the actual path.

Furthermore, Sporttracks interpreted the raw GPS track as 6,04 km. If you look at the picture, you'd probably assume that the GPS track should measure short, since it seems like it cuts corners rather than "forgets to turn". This again, leads me to believe that the 310 correction algorithms, tend to overshoot, which is worse than if it was conservative.

I hope someone at Garmin's Forerunner 310 dept. is listening!

T_SMIT
07-30-2009, 11:55 AM
From the picture it's not especially clear what arrow or dot belongs to what lap. It would be expected that the position solution would tend to drift over time as the satellite constellation changes relative to your position. It doesn't look to me like any of the positions has more than maybe 1.5 meters error relative to the inside lane of the track... that's pretty impressive considering that as you run, you're constantly shadowing part of the sky with your body, and as you round the corners you quickly change what part of the sky is blocked. From that point of view, track workouts are probably a worst-case scenario as far as GPS distance accuracy is concerned.

Do you have a record of the GPS solution accuracy during your run? That would help to determine if the 310 was working as well as possible or if there is some reason to think it's position solutions are off.

BBECKER39
07-30-2009, 01:48 PM
The 310 calculates distance using its GPS location once per second. Since the track is recorded with Smart Recording, recalculating distance from the track may 'cut corners' more than the original calclation, resulting in slightly less distance. There is not a "correction algorithm" being applied to the recorded distance.

ALAALA
07-31-2009, 01:59 AM
The 310 calculates distance using its GPS location once per second. Since the track is recorded with Smart Recording, recalculating distance from the track may 'cut corners' more than the original calclation, resulting in slightly less distance. There is not a "correction algorithm" being applied to the recorded distance.

Aha, that's very interesting to hear. Of course I knew that the Forerunners always samples per second even in smart recording mode.

What I didn't know (and have never seen stated anywhere, despite having read most Forerunner forum threads and information pages available on the internet since man invented the wheel) was that the 310 (and the 305?) always uses each and every "one second solution" to calculate the actual distance, but only stores to history the GPS points it found reasonable (depending on the GPS accuracy that second, I guess).

When working with a movement pattern such as that created running on a track, does that mean that whenever the history shows a cut corner, then the 310 (and 305?) would in fact always have used a non-saved GPS location solution, for the distance (and speed/pace) update, that laid in the tangent somewhere (lane 2,3,4,5...).

That would certainly explain, wouldn't it, why the 310 (and 305) always ends up with too much distance registered when measuring track work, and that post-processors such as Sporttracks would always end up with a lower figure than the 310 (305), since they will not calculate using those "dark" GPS points never saved to history.

Or am I seriously mis-thinking here?

ALAALA
07-31-2009, 04:42 AM
The Forerunner results of today's easy trail run intrigues me.

Both the 310 and 305 was started and stopped at the same time (within half a second or so, but the 310 first on both occasions). Both had 10+ satellites locked. It was the first time I did this particular trail with the 310. The 305 averages 9,68k (usually 9,65-9,70). The trail has, btw, been measured using a measuring wheel to 9,76k.

Today:
Distance on device & in GC: 310/9.67k (http://connect.garmin.com/activity/10194213), 305/9,63k (http://connect.garmin.com/activity/10195558)

Sporttracks, OTOH, calculated, from the tcx-file, like so (in the image red is 310 and white is 305) :
310: 9.64k, 305: 9,69k.

So, while the units differ by 0.4 with the 310 having the higher number, Sporttracks had the units differ by 0.5 but with the 305 having the higher number.

I have not seen this "turn-around" before AFAIR when I've used two devices the same time (305/305, 305/405, 50/305).

FULMAR2
10-21-2009, 12:23 PM
ALAALA -
I have observed the same thing with smart recording - going around a track. The 205 and 305 let you set the track to record every 1s - and you get a precise look at your track workout - usually in lane 1 or 2 when you're in lane 1. Good enough for me. The new smart recording in the 310 results in poor accuracy, and I really hope that they allow for a feature where you can select 1s recording in this unit as well.