Simple logger with Bluetooth connectivity

Some news :
- fourth LCD type is a loss, one out of six failed like types two and three.
- fifth LCD type is much more resilient but less sunlight-readable. Still is but not at all-angles like type 1.
- I'll decide wednesday based on first trials.

- It's now COVID lag effect. I made stocks at the start and it had no impact so far. But my last orders are harder. Broken inventories, shipping restrictions, and legal cloudiness means my last few orders are 7-15 days past arrival date. This concerns solvents, resins, glues, memories, batteries, shipping goods, textile, accessories. And I can't laser cut textile/packing foam for the foreseeable future. Got a bit of stock left of everything but replenishment are late.

- All in all it's fine and moving forward. I'll be done with Small 5-packs this week and LCD is making its come back.

- Doc for Small is here, second block : www.motion-gps.com/motion/documentation.html
- Please firmware update it before use, it takes 5 minutes.
- Map is disabled now, will be added next month as it requires your device to connect to internet and we need more trial time.

This popped up in my facebook feed... Obviously not waterproof, but CBC milled on demand.
mainlandelectronics.com.au/arduino-mega-case?utm_source=23844458463680422

Looks great, love the size!

There's a dumb 3am-coding bug in all new devices : a full battery shows up as 89% (thx Russell & Kamil). I'll publish the firmware during the weekend as I have many improvements to verify.

LCD type 5 are fine so far.

OAO are supported directly by GPSResults, GPSSpeedReader, GPSAR but you might need to update them first.
And KA72 and GPSSpeedSurfing accept its direct upload.

Awesome news, hopefully the Tassie ones will b here shortly

Was under the assumption that the loggers came about as a solution to the halt in production of the full Motions because of the screen failure issues- people asking for 'screen-less Motions' as an interim measure... so the loggers have exactly the same internals, no screen, smaller battery, bigger antenna.
If I am correct in the above, hopefully formal approval of the small Motion will be expedient.

I have our shipment and have started testing. It is more a matter of working out if there are any particular instructions or settings that we need to specify. We are extremely confident that the data itself will be excellent. We should know in a day or so but any help from the community is most welcome. If you have any comments or side by side files from side by side testing with any approved GPS, please send them in.

Looks like a great "pack", nice work Julien
Sailquick, how does it feel strapped to your arm? I currently use a Paqua + GT strapped to the shoulder strap of my vest. Not ideal I know, but strapped on my arm it constantly slides to the underside, which is probably worse as well as being bloody annoying! The solid Motion strap looks better to me. I might cut some slots in the back of my helmet too.

So far the data from the Motion Logger is pretty much as expected. There seems to be an issue with a couple of minor things.
Here are my preliminary thoughts so far:

The first is a charateristic of this Ublox chip to keep reporting speeds after a crash when the satellites drop to zero. The software filters should pick this up and discard those points but it appears that not all software is doing this at the moment.(GPS-results seems to be handling it correctly.) This is just a matter of the software analysis authors adjusting their software if required and users checking their results carefully. It is more likely to affect longer divisions like NM, but could affect others if there are sudden crashes. This was most obvious in some foiling sample data.

The second is an apparent tendancy for missed points (200ms) occaisionally when the Motion is using over 20 satellites. Julien is adjusting the max number used in the calculation to 18 best sats. He thinks this will solve that issue. Watch out for a firmware update to fix this.

While the logger has a larger antenna, it is still vulnerable to losing reception from poor wearing location. For example, today I did a test on a motorbike with it in my jacket pocket and there are a LOT of low sat numbers and therefore filtered points. This is particularly relevant to some sailors who would like to carry it in their bouancy or crash vest pocket. As with all the other GPS devices, this is a very bad idea for two reasons: 1. the body shields and blocks the view of a large number of satellites meaning the unit may not see enough for a fix (minimum of 5)
2. The material of the garment, particularly if it is wet, can very effectivly block or severely reduce the satellite signal reaching the antenna.
For best results it will need to be worn somewhere on the body where it has best, unrestricted sky view. The upper arm is excellent with the usual caveat that it must not be allowed to slip around under the arm. The wrist is a very poor option for the same reason as the GW-60. Underhand grip will severly degade the satellite signal and therefor lead to more errors, missed points and degraded accuracy. The Top of the helmet is ideal but not always practical for everyone.

More info to come soon.

Here's a longer explanation of the quality filter. Most devices (all I've known) are "best effort". They try to produce a solution with whatever they have, be it 5 poor links. This leads to disputable records where data is on the line between valid and invalid, eventually because the device knew something was wrong but its filter inertia delayed the alarm bells. I'd like to move away from this.

If the scale of link quality is 0-100, Motion Logger gets :
- 94 in speed sailing,
- 88 in a car,
- 60 in a dense forest, indoor or in a jacket,
- 30 when you waterstart.

Now, 60 is enough to produce an okayish solution. Is it interesting? No. I much prefer cutoff at 75 and have indisputable records. There's no reason your run would start at 90 and end under 75 other than a freak event like a fall. Doing it like this means the fall is detected and the run invalidated straight away, no slow fade of accuracy.

It wouldn't make a great forest-trekking device because the filter would throw away the okayish points. But I don't expect anyone to use it for anything else than speedsurfing. If it's your plan, hit me up.

Excellent post Julien. Thanks. :-)

Just awesome... apart from being able to download files and post, through KA72, to GPSTC, think I have worked out that I can update the firmware via my phone also.

Here is an interesting observation I made while looking at two Motion 10Hz Logger tracks made at the same time.

There has long been speculation that the small sawtooth variations we see in 5Hz and 10Hz unfilterd, non 'smoothed' speed graphs may be some sort of random error.

This seemed like a plausible theory for quite a while until we noticed that the pattern from identical devices worn at the same time often show a pretty good correlation in the sawtooth. That suggested they were either real variations in speed (at a very small level) or that whatever error caused them was happening at almost the same instant. We thought the latter was probably less likely as random errors would not be expected to correlate.

Today, while studying some recent comparisons I did on the motorbike, I noticed a pattern in the graph which I realised was the very slight pauses as I changed gears. It's a constant mesh type gearbox, so the changes are only a fraction of a second, but they are clearly visible in the speed graph at 10Hz. In this case, the two Loggers were not both together in my helmet. One was in the helmet and the other on my arm. One (the blue line) seems to have picked up the gear change/speed change, more obviously than the other. I did not note which GPS was in the helmet, but I suspect it was the one that picked it up more obviously.

To me, this is just more evidence to suggest that the small sawtooth we see is actually, or mostly, real changes in speed, and not so likely some random error.

It's also interesting to note that, if the Blue line one was in my helmet as I strongly suspect, it is seeing a couple more satellites (with strong signal) most of the time. This would be consistent with it having a slightly more unobstructed sky view compared with the upper arm.

I guess I am now going to have to do another test with both loggers in the helmet. It would be nice if I could do some sailing tests, but we are in the Autumn Doldrums down here.