Garmin 轉了新GPS晶片後 (Fenix/Tactix, Fenix2, FR10, FR220, FR620)

bunbun

Garmin 轉了新GPS晶片後 (Fenix/Tactix, Fenix2, FR10, FR220, FR620)
https://forums.garmin.com/showthread.php?87511-The-worst-GPS-accuracy-performance-on-the-market

The worst GPS accuracy/performance on the market

Since Garmin has decided to replace SiRF Star chipsets with their no-name M4/M4xx GPS chips (MediaTek maybe?), the accuracy and reliability drastically decreased.

I owned several Garmin models and while Forerunner 110/210, 310XT,410, 610 and 910XT are reliable and accurate devices (all with SiRF chipset), then "new generation" with cheap Mxx chipsets (Fenix/Tactix, Fenix2, FR10, FR220, FR620) are far far away from the latest and modern devices made by rival companies..

I bought FR220 several days ago and after several runs I've decided to return it back. It has the same problem under trees as Fenix had, loses signal, the track is chaotic, unstable.. Absolutely unacceptable. The same results has FR620 because it's simiral device as FR220 (difference is touch screen and firmware, chipset is the same).

It was my last and final purchase of Garmin product. It's a shame how can degrade a company which used to be so respected in Navigation/GPS branch.

keithlky

我比較信賴 sirf star IV (MW-XT), sirf star III (first gen of 910XT)chipset

Ziramas

Ok, let me have a run with both my 910xt (s0ftware *3.10) and Fenix 2 (software 3.80) to compare.

My old 910Xt is broken and I got a new one with software 3.10

bunbun

期待你的分享

網上一篇比較的文章
http://captainchas.com/2014/05/07/fenix-2-vs-910xt-ironman-testing/

bunbun

印象中, 網上見過一張 910 xt vs Fenix 2
upload上 garmin connect後的路線map

好似是 910xt, record的路線map 會較為smooth

bunbun

Ziramas 發表於 2014-9-10 23:09
Ok, let me have a run with both my 910xt (s0ftware *3.10) and Fenix 2 (software 3.80) to compare.

M ...

話時話, 你有910 , 又有Fenix 2, 會否太多錶呀?

bunbun

之前提到910 vs Fenix 2的路線map:

https://forums.garmin.com/showthread.php?79470-Anyone-else-having-problems-with-the-GPS-accuracy-and-stability-of-signal-of-Fenix-2

Ziramas

bunbun 發表於 2014-9-12 16:14
話時話, 你有910 , 又有Fenix 2, 會否太多錶呀?

I am selling my Fenix 2 at e-bay now

bunbun

有人做咗個咁嘅實驗:

http://www.dcrainmaker.com/2014/03/garmin-fenix2-multisport.html#comment-741868

Fenix 2 ACCURACY TESTS ON 400 meter track
There's been some question about the Fenix 2's GPS accuracy. Part of that question can be answered by testing it against a known distance, the easiest being a 400 meter track. On 8/28/2014 I did a 6x400 m workout with the Fenix 2 (Firmware 3.80). Garmin Connect rounds all the values off, but if you import the workouts into Garmin Training Center and view them as metric units, the unrounded values recorded by the watch are shown.
Here's what my Fenix 2 recorded:
rep 01 -> 394.98 m
rep 02 -> 401.97 m
rep 03 -> 404.42 m
rep 04 -> 391.97 m
rep 05 -> 397.98 m
rep 06 -> 391.70 m
Total  -> 2,383.02 m recorded, 2,400 actual track distance, 2383.02/2400 = 99.29% accurate!
Garmin Connect link, note: the route going into the bleachers is not a mistake, I did the recoveries under the bleachers to get out of the hot sun:
link to connect.garmin.com
Since that seemed accurate, I decided to take it a step further, following DC's excellent articles on GPS accuracy here:
link to dcrainmaker.com
So I did several walk/run tests on the 400 meter track summarized below with links:
WALKING:
01    WAAS, 1-sec, walk 800 m->   785.83 (98.23% accuracy)
link to connect.garmin.com
03    WAAS, smart, walk 800 m->   807.84 (99.02% accuracy)
link to connect.garmin.com
05 NO WAAS, 1-sec, walk 800 m->   799.84 (99.98% accuracy)
link to connect.garmin.com
07 NO WAAS, smart, WALK 800 m->   797.29 (99.66% accuracy)
link to connect.garmin.com
   WALK SUMMARY (3,200 m) -> 3,190.80 recorded (99.71% accuracy)
RUNNING:
02    WAAS, 1-sec, run 800 m->    805.37 (99.33% accuracy)
link to connect.garmin.com
04    WAAS, smart, run 800 m->    802.95 (99.63% accuracy)
link to connect.garmin.com
06 NO WAAS, 1-sec, run 800 m->    799.14 (99.89% accuracy)
link to connect.garmin.com
08 NO WAAS, smart, run 800 m ->   807.05 (99.12% accuracy)
link to connect.garmin.com
    RUN SUMMARY (3,200 m) -> 3,214.51 recorded (99.55% accuracy)
    ALL SUMMARY (6,400 m) -> 6,405.31 recorded (99.92% accuracy)
CONCLUSIONS: As far as the 400 meter track goes, the Fenix 2 (firmware 3.80) is very accurate. The 'ALL SUMMARY' above shows that in about 4 miles on the track the Fenix 2 was only off by about 5 meters! That gives me confidence for the road because normally you're not turning near as much as on a track. It would be interesting to have DC's other GPS accuracy tests for the Fenix 2 (wooded trails, etc.) that he did in his GPS accuracy tests. Hope this helps answer some of the questions about GPS accuracy for the Fenix 2.

bunbun

https://forums.garmin.com/showth ... p=387859#post387859

Here’s my quick side by side comparison of the 910xt & F2 with the latest firmware (3.9).

I did a simple run around a suburban block near my home, 3 loops, 2.35 miles total, one device on each wrist at the same time.

910xt: http://connect.garmin.com/modern/activity/588745803
Fenix 2: http://connect.garmin.com/modern/activity/588745892

From my perspective, looking at the map, both devices seems to be similarly inconsistent. From previous posts in this thread I would have expected the 910 to really outshine (firmware v 3.0). I stayed on the sidewalk the entire time, hitting the same exact path. The track are all over the place. Both devices are set to Smart Recording. Moderate tree cover on only one section. WAAS is enabled in the F2, but I haven’t yet seen the “D” (I just enabled it shortly before my run). I’m just outside of Boston Mass so I assume I’d be able to get a WAAS signal. I believe on the 910 it’s always enabled by default (I can’t find the setting). The overall distance is only .01 miles off, along with 2 seconds on the timer (my button pushing delay) – that seems reasonable. Before I started my run the accuracy read ~15ft on the 910, and ~40 on the F2, which seems to be consistent with these two units for me.

The elevation is also slightly different, the gain/loss is pretty close between the devices. But the actual is ~70 feet lower on the F2. I’m not sure if the chart is using the GPS or the altimeter to calculate that. I have the Altimeter set to auto calculate = continuous. And I manually calibrated it once about two weeks ago at an AMC lodge in the White Mountains (Madison Spring Hut) – a sign posted said it was 4805 feet. Not sure if manually calibrating as this high altitude lowered its accuracy at sea-level or not – any thoughts?

The speed/pace looks pretty erratic on the 910 compared to the F2.

Slightly off topic: I’m glad the cadence on the new HRM with the F2 is nearly as accurate as the 910 with the footpod.