Vista C vs HCx

[Euan Uzami]

re. the handlebar mount , the way to stop it rattling about on there is to put some insulation tape on the base surface of the mount itself (avoid the latch bit). Each successive layer makes it hold it slightly tighter, i find 2 or 3 layers and it is just perfect.

I've read reports of the HCx sometimes being TOO sensitive, but i've never experienced this.

[frankly frankie]

I've read reports of the HCx sometimes being TOO sensitive, but i've never experienced this.

It can get a fix using weak signals, and these will often include some reflections so the fix will be inaccurate.  
However presumably a poor fix is still a lot better than no fix at all.

The question is, where there are sufficient strong signals available, do any weak ones still go into the mix - if they do, that might make it less precise than a GPS of lower sensitivity in the same situation.  
But you'd never notice, on a moving bike - a lot of the 'accuracy' angst is about getting stationary fixes and doesn't relate to following a route or recording a track.

[andrew_s]

I think the weak signals do get into the mix, and that this is why walking pace tracks are so much worse than cycling pace tracks. If you are moving, you don't get any one reflection for long enough to get a lock onto it and for it to render your position less accurate.
(supposition, not information)

[drossall]

Slight sideline, but I'm planning to follow a downloaded track on Saturday at the Golden Tints. What's the easiest way to make a Vista C "TracBack" over a whole track? Obviously I can select the last point, if I can find it on the screen, but that's fiddly. Isn't there a shortcut to tell it just to follow the whole track?

[Sergeant Pluck]

Slight sideline, but I'm planning to follow a downloaded track on Saturday at the Golden Tints. What's the easiest way to make a Vista C "TracBack" over a whole track? Obviously I can select the last point, if I can find it on the screen, but that's fiddly. Isn't there a shortcut to tell it just to follow the whole track?

The Backtrack function will allow to you go back along the route you have just taken (rather than the downloaded track). So if you have, on the day, followed your downloaded track in one direction, the device will have laid a "breadcrumb trail" as you went, and it is this that the backtrack function uses.

IYSWIM

[drossall]

I think so, but I've got a saved track (by uploading it into the GPS). Now I just want to follow it from start to end. I can't see an option to do that on the Vista C.

[Sergeant Pluck]

The track you've uploaded to the device is simply a line on the displayed map. Just click on the appropriate track, ensure "display track" or whatever selected, and when you are in the area of the start (or any point on the track) you'll see the line on the map. Then it is simply a matter of following it - there are no direction instructions etc.

[simonp]

The track you've uploaded to the device is simply a line on the displayed map. Just click on the appropriate track, ensure "display track" or whatever selected, and when you are in the area of the start (or any point on the track) you'll see the line on the map. Then it is simply a matter of following it - there are no direction instructions etc.

The Trackback facility allows the GPS to follow a pre-recorded track and give you a direction arrow (on the compass display) and it beeps when there are turns in the track.

The Vista HCx has a very poor UI for turning this mode on though, which is exactly what drossall is describing.  It's a pain.  I don't know of any other way to turn Trackback on.  Idiots.

[drossall]

Thanks Simon. One other point is that Etrexs are designed for walkers and cyclists, who tend to travel in circles back to their point of origin (back to their cars or railway stations or homes). If I ask to track back to a point near where I am, how can I be sure that it won't take me the short way?

On this occasion, it's two loops, which is even worse! Will it go half way and stop?

My old Etrex Summit was better - it just offered to follow a track to end or start, like a route. I may try that, but I'll have to downsample further - I think it has a limit of 250 points.

[Sergeant Pluck]

On this occasion, it's two loops, which is even worse! Will it go half way and stop?

Is is a track or a route that you have uploaded to the device?

If it is a track, you won't need to use Trackback - forget about Trackback. Simply follow the displayed track as a guide to where to go at junctions etc. It will not "take you the short way" as you are simply following a line on the displayed map.

[andygates]

IIRC, trackback does a Hansel and Gretel "follow the breadcrumbs" thing, so if you do a figure 8, it'll walk you back through it all: it has no discretion.

It's bloody great on Dartmoor when the fog descends to 25m viz and you think "bugger this, I'm going back to the car" -- that's really what it's designed for.

[simonp]

When you start off on the tints, at the start point it has 3 options: take the 1st loop, take the second loop, or decide you've already finished.  It's pretty random as to which behaviour you'll get.

One solution is to ride the first part of the route until they diverge (first left turn, towards Henham) and then power cycle the device.  That'll sort it out as it will then be on a section of the route with no ambiguity.  A much better solution is to split the track into control->control sections which means there are no loops.  At each control you select the next track.

As for use Trackback or not, that's a personal choice.  I prefer to have a beep to remind me when to turn.  However tracks are a poor way to navigate compared to turn-by-turn routes.  If it's the same as last year's then last year's GPS route should be fine - which I may have got from Chris S.

[Chris S]

When you start off on the tints, at the start point it has 3 options: take the 1st loop, take the second loop, or decide you've already finished.  It's pretty random as to which behaviour you'll get.

One solution is to ride the first part of the route until they diverge (first left turn, towards Henham) and then power cycle the device....

However tracks are a poor way to navigate compared to turn-by-turn routes.  If it's the same as last year's then last year's GPS route should be fine - which I may have got from Chris S.

Ahem... second left ITYM.

My GPX is also flawed, because I used to repeatedly make the mistake (and it was you who alerted me to this, Simon) of closing the loops when you have a looping route. This has the rather unfortunate effect that as soon as you leave, the GPS says "Turn around - you're nearly there!". 

The simple solution is not to close the loop - stop the loop one waypoint/instruction before the HQ waypoint.

[Euan Uzami]

...
But you'd never notice, on a moving bike - a lot of the 'accuracy' angst is about getting stationary fixes and doesn't relate to following a route or recording a track.

FWIW i read in the instructions of a car GPS (but probably also applies to bike ones) that it needs 4 satellites to lock on, but then only needs to maintain contact with at least 3 while moving.
This is presumably because a (relative) set of distances to 3 different satellites will narrow your location down to two singularity points, on opposite sides of the earth. It thus needs the fourth satellite to distinguish which one of these you are at, but then only  needs to maintain 3 because it can rely on the assumption that you are not going to go right round to the other side of the earth in the time between successive polls.

[Greenbank]

This is presumably because a (relative) set of distances to 3 different satellites will narrow your location down to two singularity points, on opposite sides of the earth.

Not quite, the two points don't have to be on opposite sides of the earth.

A signal from a single satellite means the device knows how far it is from that satellite and the position of the satellite. This alone puts you anywhere that is exactly x thousand kilometers from the satellite on a really big sphere (much bigger than the earth).

Only some of this huge sphere will intersect with the Earth (which itself is roughly spherical). The intersection of two spheres is either a single point where they just touch (very unlikely) or a circle. This circle of intersection doesn't have to be a great circle (i.e. one that passes antipodal points). Imagine two footballs pressed together, the circle of intersection isn't the same diameter as either of the footballs, as you press harder the circles of intersection will get larger.

3 spheres will intersect at two points, but these, again might not be on opposite sides of the Earth. They could be relatively close together (which is commonly what happens when you're talking about huge spheres with a radius of the geostationary orbit of satellites).

4 spheres, or more, will intersect at a single point (or near enough given the relative accuracies).

The Earth isn't used as a reference sphere as it isn't a perfect spheroid. GPS accuracy is usually down to 5m or so, whilst elevation (in the UK at least) can vary by over 1000m. So it takes at least 4 satellites to get a location fix.

[frankly frankie]

I believe -
1. 4 sats are needed to synchronise the chad-valley GPS clock to the atomic clocks on the sats.
2. Once this has been done, 3 sats should be sufficient, using the radius of the Earth as a virtual 4th sat to eliminate one implausible point out of the possible 2.
3. In practice the GPS will use as many signals as it can find (the maths is terrifying) which is why the very sensitive models may be slighly less accurate in some situations.
4. A GPS in motion may have all sorts of extra tricks - extrapolating speed and direction of movement when sats are temporarily masked - integrating eg 2 sats seen on one poll with 2 others seen 1 second later.  This seems to allow travelling at cycling speed under quite dense tree cover without any apparent problems. 

I think some models (eg car satnavs) are optimised for use in motion, while others (eg surveyors tools) are optimised for stationary fixes.
The Etrex range are intended for walkers, so you would expect them to perform well at walking speeds - but in fact some models have been found to have problems at speeds under about 10kph, but they're fine at cycling speeds.

[drossall]

Try again then. I've used GPS Trackmaker Free to convert the track to a route, downsampling at the same time, and uploaded that to the GPS.

Or I may just read the route sheet

[Greenbank]

The Etrex range are intended for walkers, so you would expect them to perform well at walking speeds - but in fact some models have been found to have problems at speeds under about 10kph, but they're fine at cycling speeds.

My eTrex works fine at 1000kph too (and 31000ft ASL). It needs a window seat though.

[andygates]

And for accuracy, the Etrexes ain't bad at walking when they have good signal -



Traces overlaid on Google Earth.