So for some work related stuff i have been working with gps in an application with race cars I have to make a custom board up for this at some stage and thought I would make a windsurf style gps to test things out . It uses a ublox 7 and a samd 21 so has plendy of power to calculate things while sailing , it also has sd card for logging and an oled screen to display. Ill get a small round case machined soon with the aim to make it actually waterproof none of this bull**** ip rated stuff that you pick up with a wet and and distroy the whole thing im aiming for put it in the washing machine and not have any leaking.Still need to check its function some more but it seems good so far. The cool bit is that i can easily write code to show what i want like nm , 10 sec and so on, The only thing im not sure how to go about is alpha if any one has any tips on that.

Brilliant Simon. 45mm is just slightly bigger than a standard watch face too. Also I reckon Dylan could help you with the alpha programming (code).

yeah i was aiming for watch size but the problem was fitting sqaure thing in round holes as i figure a round housing is the only one that is going to seal properly on a budget.

I've been playing with some hardware also
- ARM M-4
- Ublox M8Q with special antenna
- Big LCD, 55x46mm

Overall size is same as GT-31

More info about the whole project
forum.windsurfing.nl/viewtopic.php?f=62&t=13890429

Bloody brilliant.

Hijack someone else's water proof box/housing.

GoPro (vast array of attachments), phone case (life proof) etc

Do you need pre orders? 1 please

PM sent re Alpha.

forum.windsurfing.nl/posting.php?mode=reply&f=62&t=13890429

google translated
translate.google.nl/translate?sl=nl&tl=en&js=y&prev=_t&hl=nl&ie=UTF-8&u=http%3A%2F%2Fforum.windsurfing.nl%2Fposting.php%3Fmode%3Dreply%26f%3D62%26t%3D13890429&edit-text=

Cases are easy, no need to hijack someone else's, just 3D print one. Once you have a 3D printer, all these sorts of insurmountable problems just go away.

If you need something done then send me some 3D models of the electronics (or just a diagram of where the mounting holes are) and I'll design and print something that should work.

I thought I would get something like this machined (unfortunately it will have to be plastic not aluminium) The idea is to have twin o ring seals so it never leaks and for the screen to be off until you stop then have it cycle through the speeds showing what you have done dependant on exact code of course. I decided to have no buttons on the outside because they seem to be the where the water gets inside things most the time. I can’t decide yet if the on off should be through a quarter twist of the case or maybe a magnet. As you can see it’s a fast gps

What are the dimensions of the battery?

I havnt bought one yet but they come in heaps of sizes.

Very elegant design Simon. Have you thought about crowd funding it? I'd happily fund something as clever as this that is specifically built for GPS sailing.

Just a thought but in lieu of magnets etc to play with settings & switch on and off, what if the unit came in a standard semi waterproof casing (think GT-31) that had access to switching and then went into a waterproof case once turned on - similar to say a GoPro. Your cylindrical design would be much simpler though than the GoPro case - I'm envisioning a screw on cap like a shoe polish tin. People could buy the waterproof casing as an extra and have various attachment straps (wrist, arm band etc) or just the standalone unit thrown in an aquapac.

What's that saying...."necessity is the mother of all inventions"

Select to expand quote

Simon100 said...
I thought I would get something like this machined (unfortunately it will have to be plastic not aluminium) The idea is to have twin o ring seals so it never leaks and for the screen to be off until you stop then have it cycle through the speeds showing what you have done dependant on exact code of course. I decided to have no buttons on the outside because they seem to be the where the water gets inside things most the time. I can’t decide yet if the on off should be through a quarter twist of the case or maybe a magnet. As you can see it’s a fast gps

Beautiful design, I'd buy one if it can be priced in that 250 /350 range

Screen needs to be on all the time when sailing with BIG numeral display to see speed while sailing!

Select to expand quote

Simon100 said...
I thought I would get something like this machined (unfortunately it will have to be plastic not aluminium) The idea is to have twin o ring seals so it never leaks and for the screen to be off until you stop then have it cycle through the speeds showing what you have done dependant on exact code of course. I decided to have no buttons on the outside because they seem to be the where the water gets inside things most the time. I can’t decide yet if the on off should be through a quarter twist of the case or maybe a magnet. As you can see it’s a fast gps

Could a touch screen be incorporated?

Where are the ports for charging & down loading or use WiFi & that smart charging on a plate type thing.

I can make different version of code that show different things or make it selectable with a configuration file on the sd card

I think we are missing something here. I recon there is a better way.
Technology had advanced to where there are other better options.

The 'all in one' windsurfing GPS logger with screen and 'Genie' was the old way.

Now we can do it better, and far cheaper.

Gps-Logit Android app already exists to do all the calculations for instant feedback with a large colour display and massive processing power. It can send the data to GPS-Results at the end of the session for immediate gratification results and then even upload it to the website of your choice.

The only problem is that the gps and antenna in phones are useless crap for our purposes.

Enter a Bluetooth GPS dongle with UBlox .ubx 10-18hz output. Gps-Logit can read and accept this (not yet in the production model but I have a test version that can do it now which should be released very soon) and display and calculate everything using the very best data available!

Advantages:

You can use any Android phone, even the cheap ones, and you can place it anywhere you like where you can see it or hear it best. Put it on your wrist or boom if you like, as location for best GPS reception is no longer a problem.

Wear the GPS BT dongle in some other position on your body where it gets ideal reception, like the top of your head/helmet. It is small and can be very waterproof so it can go anywhere where it gets good sky-view/reception.

The dongle can use a very high quality UBlox gps with all the very latest accuracy and error data at 10-18hz but still be quite inexpensive as it doesn't need all that extra screen, processing power, larger power supply to run them and switches etc.

It is easy and cheap to add another dongle for two GPS verifying data if you want to claim a record and they can be so small you can wear them together also in the best position.

This all seems like a no brainer to me and uses what is virtually existing technology instead of trying to build something complicated with a very limited and specialised market (windsurfing). High quality dongles are in demand for all sorts of other applications and phones are dirt cheap (eg. Pixie3 = $29!)

Unfortunately, with the Wintec G-Rays and G-Rays11 now out of production, there are no UBlox based, stand alone BT dongles on the market now that I can find.

Back to where I started: It would be relatively very cheap and easy to build these compared with a specialised all in one windsurfing GPS.

Now a look into the slightly more distant future (a year or two?):

There is already an Android phone app that can process high quality GPS data and do RTK position data to cm level accuracy. No internal phone GPS are currently capable of supplying the quality of GPS data that RTK requires. But! There are now inexpensive Loggers that can supply that data like the Bad Elf GNSS SURVEYOR ($600). That one is only claiming 10-50cm accuracy in RTK, but there are scientific research papers published that have demonstrated cm accuracy using similar GPS engines with consumer grade 25cm patch antenna and even smaller!

It would not be a huge step to design a BT dongle around an inexpensive RTK standard GPS chip and consumer grade patch antenna, that could supply the quality of data for RTK to an Android (or Apple) phone. I suspect that is well possible right now!

To put that in perspective: The WSSRC approved the Trimble Survey GPS for the Macquarie Innovation team in the early 2000's for their world records (since used for every WSSRC record attempt expect the Kite and Windsurfing ones). Their requirement is 10cm or better accuracy. When it was approved, those systems cost in excess of $60,000 and they claimed 8cm accuracy. Now, every large tractor, header and cotton picker has these things built in and do better than 2cm accuracy! Those are still too expensive and too large for us ($5,000-10,000?) but it is now possible build that accuracy, and better into a consumer device, and at a price that makes it available to windsurfing enthusiasts. No one is going to do that just for us windsurfers, but there are big markets in other areas for such accuracy. They will not be all in one devices with a 'Genie' for windsurfing, so we need to change out mindset to embrace the possibilities offered to us as I have outlined above.

Exciting times!

My self im more in favor of a water proof reliabe option , not so much chasing the last 0.01 knot acuracey and also i would be concerned about the reliablity or the connection between the 2 devices. there are usb ubox receivers that could be plugged into a phone quite easily.

I think the big misconception a lot of the readers here suffer is that I am banging on about 'chasing the last 0.01 knot accuracy'.

Although that is important to have, the biggest issue is not accuracy level, but accuracy verifiability. i.e.. Always knowing how accurate every result is!

It is no good having 0.01 accuracy 90% of the time and 'who knows what' the other 10%. One 'accurate' reading does not mean another is accurate or trustworthy.

I would actually rather have a less accurate device, (even if it is only 0.1kts accurate), that is 100% verifiable! (with SDOP).

If you don't want 'accurate' or 'verifiable', or 'verifiable accuracy', we already have GPS-Logit and cheap android phones and expensive, but highly waterproof GPS watches! Don't bother going any further.

If you want to have any kind of ranking and competition that is fair and meaningful, you must have a verifiable level of accuracy, otherwise you might as well just make up some numbers.

If you know of any ublox based receivers that do actually work now with an Android phone I would love to know about them. I have not been able to find any at all.

My understanding is that RTK is inherently verified as the calculation is accurate if it is resolved, and just fails if it is not. It is resolved through differential correction from a geostationary base station.

Using the H2O audio armband enclosure, the Android phone is far more waterproof and reliable than anything else I have used, ever! A BT dongle is easy to make absolutely 'bombproof' because it needs no external controls or screen. BT is apparently good to a range of about 10 meters. Far more than we require and, of course, requires no wires for a connection.

Select to expand quote

Simon100 said..
My self im more in favor of a water proof reliabe option , not so much chasing the last 0.01 knot acuracey and also i would be concerned about the reliablity or the connection between the 2 devices. there are usb ubox receivers that could be plugged into a phone quite easily.

Not many phones have OTG and even ones that do have heaps of problems connecting judging by the forums I have read. I have a usb ublox gps (A$14) but it does not connect to Android only windows. I have seen ublox gps mouse receivers with stereo jack connector (for Dash cam I think) but I cant find any that can be plugged into a phone easily.. I'm with Andrew on this . Make small BT gps receiver and let the phone do the work.

Select to expand quote

BSN101 said..

Simon100 said...
I thought I would get something like this machined (unfortunately it will have to be plastic not aluminium) The idea is to have twin o ring seals so it never leaks and for the screen to be off until you stop then have it cycle through the speeds showing what you have done dependant on exact code of course. I decided to have no buttons on the outside because they seem to be the where the water gets inside things most the time. I can’t decide yet if the on off should be through a quarter twist of the case or maybe a magnet. As you can see it’s a fast gps

Could a touch screen be incorporated?

Where are the ports for charging & down loading or use WiFi & that smart charging on a plate type thing.

you would have to unscrew the back to charge and remove sd card , its the only relaible way to seal it with it costing alot more . charging would also be through a usb in the back when its open. It could have far more advanced features but we have to keep an eye on the price with something like this.

Simon100, keep at it we are all watching. 800 reads in 2 days.

There is going to be an evolution of your device, keep thinking outside the box.

Dongle sounds light & unobtrusive although Id like to see a screen that i can easily scroll through my results AND READ (Canmore too small)

Well done guys

Select to expand quote

sailquik said..
I think we are missing something here. I recon there is a better way.
Technology had advanced to where there are other better options.

The 'all in one' windsurfing GPS logger with screen and 'Genie' was the old way.

Now we can do it better, and far cheaper.

Gps-Logit Android app already exists to do all the calculations for instant feedback with a large colour display and massive processing power. It can send the data to GPS-Results at the end of the session for immediate gratification results and then even upload it to the website of your choice.

The only problem is that the gps and antenna in phones are useless crap for our purposes.

Enter a Bluetooth GPS dongle with UBlox .ubx 10-18hz output. Gps-Logit can read and accept this (not yet in the production model but I have a test version that can do it now which should be released very soon) and display and calculate everything using the very best data available!

Advantages:

You can use any Android phone, even the cheap ones, and you can place it anywhere you like where you can see it or hear it best. Put it on your wrist or boom if you like, as location for best GPS reception is no longer a problem.

Wear the GPS BT dongle in some other position on your body where it gets ideal reception, like the top of your head/helmet. It is small and can be very waterproof so it can go anywhere where it gets good sky-view/reception.

The dongle can use a very high quality UBlox gps with all the very latest accuracy and error data at 10-18hz but still be quite inexpensive as it doesn't need all that extra screen, processing power, larger power supply to run them and switches etc.

It is easy and cheap to add another dongle for two GPS verifying data if you want to claim a record and they can be so small you can wear them together also in the best position.

This all seems like a no brainer to me and uses what is virtually existing technology instead of trying to build something complicated with a very limited and specialised market (windsurfing). High quality dongles are in demand for all sorts of other applications and phones are dirt cheap (eg. Pixie3 = $29!)

Unfortunately, with the Wintec G-Rays and G-Rays11 now out of production, there are no UBlox based, stand alone BT dongles on the market now that I can find.

Back to where I started: It would be relatively very cheap and easy to build these compared with a specialised all in one windsurfing GPS.

Now a look into the slightly more distant future (a year or two?):

There is already an Android phone app that can process high quality GPS data and do RTK position data to cm level accuracy. No internal phone GPS are currently capable of supplying the quality of GPS data that RTK requires. But! There are now inexpensive Loggers that can supply that data like the Bad Elf GNSS SURVEYOR ($600). That one is only claiming 10-50cm accuracy in RTK, but there are scientific research papers published that have demonstrated cm accuracy using similar GPS engines with consumer grade 25cm patch antenna and even smaller!

It would not be a huge step to design a BT dongle around an inexpensive RTK standard GPS chip and consumer grade patch antenna, that could supply the quality of data for RTK to an Android (or Apple) phone. I suspect that is well possible right now!

To put that in perspective: The WSSRC approved the Trimble Survey GPS for the Macquarie Innovation team in the early 2000's for their world records (since used for every WSSRC record attempt expect the Kite and Windsurfing ones). Their requirement is 10cm or better accuracy. When it was approved, those systems cost in excess of $60,000 and they claimed 8cm accuracy. Now, every large tractor, header and cotton picker has these things built in and do better than 2cm accuracy! Those are still too expensive and too large for us ($5,000-10,000?) but it is now possible build that accuracy, and better into a consumer device, and at a price that makes it available to windsurfing enthusiasts. No one is going to do that just for us windsurfers, but there are big markets in other areas for such accuracy. They will not be all in one devices with a 'Genie' for windsurfing, so we need to change out mindset to embrace the possibilities offered to us as I have outlined above.

Exciting times!

Firstly the 600usd bad elf gps receiver is just a gps receiver and all rtk is done by an external device, all rtk devices need information from a ground station in one way or another which isn’t practical for our application and is going to cost money , its advantages come in with positional accuracy although its an interesting technology its not of any realistic use to us.

Error calculations are done on all modern gps chips they also have filters and are pretty smart, there may be the consequence of losing an instantaneous 0.1 of a knot through these but its not going to be significant over the 20 readings in the 2 second. The most obvious and fault proof error checking is when the rate of acceleration is calculated in the software on upload (although probably very similar to the internal filtering) when we were measuring gps at only 1 hz there was far less potential to catch this but at 10 hz its very easy as for example we know that if someone is going 30 knots then 0.1 of a second later they are going 35 knots then after another 0.1 sec 30 again we can say that was an error just discard that 35 and call it 30 and that’s worst case.
All gps regardless of which technique system have some degree of estimation averaging and educated guessing.

Its definitely possible to transmit to a phone using gps and making such device is easy but in effect you have the same device as an all in one unit except you ditch the screen and add Bluetooth the down sides are that you now have to deal with a Bluetooth connection and the stability of a cheap phone and have 2 devices to charge and 2 units not to get wet. There are also potential issues with the Bluetooth signal interfering with the gps signal ( I’m no expert on that stuff though) if it was an issue it would probably only be small but considering you want the best its still a step back of course testing and design could solve this but its less that ideal considering the time/ budget constraints.

Just thought id share how a gps device like this works ( the simple way) as a lot of people think it’s some complex device.
All the hard work in this one has been done by the ublox engineers and the people before them all my assembly of parts really does is on start-up send the configuration to the gps which then just sends back the messages it was asked for . if no settings are set it will just send the default messages. The message are just read by the mcu(little computer) which then copies that to the sd card and interprets the message to give some numbers on the screen. basically the unit is 100% the ublox chip and the only possible influence I have on the design is antenna placement and interference.

Simon, the more you play around with this sort of stuff the more you (and we, via the forum ) learn. Through natural evolution the best solution will win out and that solution may be different for different people or even the same person whose needs change over time.

BSN 101, I too liked the Canmore and easily swapped over when the Navi started playing up. Wanted to get the 101 (2.25x bigger screen,better shape (1cm bigger L&B, 2/3 as thick) plus skytraq chip which could be configured to 5Hz) but could find none for sale on web so got the 102+. Very accurate, great reception but no knots and a few probs with memory so when Manfred sorted out the Logit shut down prob I had no probs swapping to a $24 Pixi.

I did see the Canmore 101 available the other day for $49 but I think I will stick with the phone for now as it has an even bigger screen in knots.

I'm looking for a magnifying bubble to place on top of my Canmore so I can read the bloody thing. I struggle even with my glasses on!

Select to expand quote

BSN101 said..
I'm looking for a magnifying bubble to place on top of my Canmore so I can read the bloody thing. I struggle even with my glasses on!

As Yoyo says, wear the canmore on your head and just use it for posting data.
Use a $29 phone on your arm, for realtime feedback, both audio and visual.

Mike, as of this arvo, the Halls Head KMart has 5-6 Pixis at $24. I believe this is an Australia wide special up until end of month. It makes the $49 Canmore 101 I saw on Amazon seem expensive.

Select to expand quote

decrepit said...

BSN101 said..
I'm looking for a magnifying bubble to place on top of my Canmore so I can read the bloody thing. I struggle even with my glasses on!

As Yoyo says, wear the canmore on your head and just use it for posting data.
Use a $29 phone on your arm, for realtime feedback, both audio and visual.

I'm not overly worried about knowing while I'm on the water. It would be nice to read it when I've stopped and sipping a delicious.

Select to expand quote

Simon100 said..
I thought I would get something like this machined (unfortunately it will have to be plastic not aluminium) The idea is to have twin o ring seals so it never leaks and for the screen to be off until you stop then have it cycle through the speeds showing what you have done dependant on exact code of course. I decided to have no buttons on the outside because they seem to be the where the water gets inside things most the time. I can’t decide yet if the on off should be through a quarter twist of the case or maybe a magnet. As you can see it’s a fast gps

Looks brilliant. I like the idea to automatically turn on logging by twisting it closed, and to turn it off by twisting it (partly) open. Does not get much simpler. Using a phone and a bluetooth connection to a GPS dongle is all nice and good, but you end up with twice as many devices that can cause problems, plus a bluetooth connection that can be iffy. And until GPSLogit has an option to directly upload to GPSTC, getting the data off the phone adds extra complications.

There's plenty of examples out there that ublox-based units can give the required accuracy and error estimates. Unless you run into interference or antenna issues, the data should be great. Easy to test (but harder to fix if you run into problems).

Assuming the tests go well, it would be fantastic if you could get the unit produced. The GW-52 still leaves plenty of room for improvement. Crowd funding would be great to get this out, you'd sure get my money. But while the display and the controller are fine for windsurfing, they may limit expanding into much larger markets like running or skying, where you want to see more data. That would mean a market limited to a few thousand units spread out over several years, at best. So keep your day job!

Perhaps you can try open sourcing the entire thing, publishing the source code and all manufacturing diagrams and information.

Nice work Simon, I love the simplicity of your design and a fully waterproof design (don't compromise on that)
KISS is good!

Select to expand quote

sailquik said..


If you want to have any kind of ranking and competition that is fair and meaningful, you must have a verifiable level of accuracy, otherwise you might as well just make up some numbers.

And therein lies the problem, the current rules aren't enforced so the competition becomes meaningless. A verifiable level of accuracy is great in theory but if you aren't going to follow up on the basic rules what's the point.