Loading Course on Edge 820 Very Slow

Nickdodo,

the directions for me don’t come up when the device is calculating the route. I got to 8 miles into route whilst it was still calculating & received no turn / directions. I had to stop and get google maps on phone 

would be interested in your settings for courses if you wouldn’t mind sharing. If I get directions I don’t care how long it takes to calculate route 

thanks 

I've waited about 20min without a luck to calculate a route in the park today. It was 21km long route created in GC from dowloaded gpx. Every time it went to 34%, started from the beginning. Finally I was frustrated and I started a route. I was able to see only a path, no turn by turn navigations and distance to turn. After about 40min of riding I had 100% calculation and I had turn by turn navigation. I lost any hope that this device could be usable as navigation. Recalculating a MTB route when you miss a turn - for you it would be better to back than waiting.

It may not be much consolation, but many of us do not have issues with long loading times.   What that suggests to me is there may be an accumulation of "errors" on your device and it might be worthwhile trying a full factory reset.  If you opt to do this, save your rides and courses first, and save your setting so you can re-install them after the update.

Here's some info on that:

For factory reset:

1. Turn device off
2. Hold Lap/Reset and Start/Stop
3. Press and release the power button
4. Continue to hold only Lap/Reset and Start/Stop
5. Release Lap/Reset and Start/Stop when the Garmin Logo is cleared
6. The unit will restart and clear non-vol

Saving settings for before reset

Exactly what I am getting !

the lady in support did say gpx will take longer and need to down load / create as fit 

wish they would sort it out, these devices aren’t cheap !

I’ll see if I can put my settings on here later when I have some time. 

looigi - thanks for this detailed reply - I think a factory reset is probably a good idea - to eliminate possible sources of problems one by one.

My gut feeling (because the progress % gets stuck at particular values - different for each course - that the software is hunting for solutions and maybe only moves on at some timeout) is that the problem is something to do with the relation of the unit's maps to the course file's track - it can't follow / work out which track or road a point is on for some reason - eg the two sets of co-ordinates are mismatched or roads / tracks do not follow the exact same meanderings (route creation software may be more or less good when told to 'follow roads'). All of this should be testable.

Looking at other entries in the thread  I think my calculating times are OK (5 min).

Because Garmin's map updates require a Windows or Mac computer I can't use them (I use Linux). For this reason I download OSM maps from http://garmin.openstreetmap.nl/ using Routable Bicycle to get a gmapsupp.img file (instructions here http://www.openfietsmap.nl/installation). I plan my routes using ridewithgps with the map used as OSM. I know that the map on my unit & the map in ridewithgps may have different dates - and differ in detail as a result - and that there may be  a general co-ordinate mismatch too, but this method seems to be good enough for a 5min calculate.

My guess is that using your own downloaded OSM map, with the excellently accurate ridewithgps will give you as good as it is possible to get. I notice that ridewithgps can satisfactorly import a gpx file and make it into an editable route - something Strava's (very inferior) route planner can't do & make editable.

It is really mad that the 820 does not save the calculation for reuse - all it needs to do is check that both the course file and map files are unchanged since the calculation (retired programmer talking here).

I will now try a few different possibilities: a very short straight route through to more complex routes. I'll also see what happens when roads aren't followed well (by plotting lines). It would be worth fully zooming in to see if the planned route on ridewithgps truly and totally accurately follows the roads/tracks and how this impacts calculate times.

Early result:

Using OSM from openfietsmap, OSM on ridewithgps route planner, fit files downloaded into the unit's New Files folder:

1. 70k route (file size ~25kB), created with only two clicks - ridewithgps makes whole route. Calc time on 820 was 6min27s

2. 70k route (file size ~25kB) created by many mouse clicks to gradually make route. Calc time 1min23s

3. 36k route made with 2 clicks: 2min 39s

4. 33k route made with many clicks: 4min29s

So it seems unrelated to file size or whether few or many clicks were used. However I noticed in route 4 that the main pauses in % calc - assuming that the % refers to distance calculated - were where the route went down a little side track for a bit (ridewithgps did this - i wouldn't have done it), and later when passing through a village.

Basically inconsistent - though there really must be a rationale here somewhere...

Jacqui,

Here is a note about how I set up my routes and Garmin Edge 820. I hope it helps. Any questions ask away.

Nick

820 settings.pdf

Some more tests:

 1. 50k route, calc 2min49s, repeated 2:49s so consistent, big delay at 4% then steady

2. 49k route (the previous reversed - some detail different), hardly halted much, calc 1min25s

So though it would be natural for a longer route to need a longer calculation time the problematic slowness is not related to route length.

Assuming that the 4% above refers to distance - the difficulty the algorithm is having is about 2km into the ride. When I examine this part os the route it is along a straight road with cycle paths both sides. immediately followed by a complex few bits: a little side road stub (my fault) and zig zag with tunnels under a major roundabout on a dual carriageway. The route 2 also covers this is reverse and has a leftover stub.

Turning off Avoidance Unpaved, Narrow & Ferries produced an improvement (all Avoidance OFF) - route 1 now takes 2min17s.

(all above results repeated to check for consistency - with the same results)

I'd love to know what it is in certain courses that can cause the algorithm to slow down!