where did you find the apk to read ubx?
GNSS uses NMEA
Never said I'd read ubx, I just said I'm writing it to a file (which then can later be read by GPSResults, GPSAction Replay, ...).
ublox chips and dongles send NMEA sentences by default, but can easily be configured to send ubx messages, either in addition or only. It's a better format since it is more dense (binary) and includes speed accuracy estimates. In the (near) future, that will be done by the logging app. For the first tests, I changed to ubx logging using the "Configuration" view in u-center on Windows. When you save the configuration after setting it, the dongle will remember if for a hour or a few (capacitor-backed RAM), so you can unplug the dongle from the PC and plug it into the phone.
But reading .ubx files is rather straightforward, since the format is well documented. The hardest part is knowing which messages to send and use (NAV-SOL, NAV-PVT, and NAV-DOP works; not sure you need NAV-DOP, and IIRC, there's a better message for ublox 8 chips).
This time I found a Kyocera Hydro
Just got the Kyocera Hydro, and discovered that it's not OTG enabled, so it can't be used with the dongle.The specs list "mass storage", the settings have support for various USB modes, and USB OTG checker says the phone has OTG, but no USB devices are detected by any of the 5 or so apps I tried. Bummer. So if you get a phone to play with the dongle, make sure it really has OTG support, not just "mass storage".
OTG seems to be a bit hit and miss. My phone said it had it but initially I could not get it to recognise any usb. One OTG checker said it was working and another said it wasn't!
Then it started working. Looking online it seems quite a few have OTG for a short time then loose it. I suspect might not be enough voltage to power usb or borderline perhaps.
OTG seems to be a bit hit and miss. My phone said it had it but initially I could not get it to recognise any usb. One OTG checker said it was working and another said it wasn't!
Then it started working. Looking online it seems quite a few have OTG for a short time then loose it. I suspect might not be enough voltage to power usb or borderline perhaps.
Could be for some phones. Thinking power might be an issue, I bought an OTG Y-cable, with an extra plug for a power supply. The power gets to the dongle, it starts blinking when it gets a satellite lock. Did not make any difference on two different phones, though, they did not see any USB devices. I even rooted one of them and installed some more USB tools, but to no avail. A bit frustrating since both phones can talk perfectly fine computers, and one of them has OTG support in the kernel, and has the required libs. I think OTG specs require a certain mA level of power, so if a phone has no power out on the USB connector, it has no OTG. But as you say, the power may be not enough for some devices, which could explain some of the OTG "on/off" problems. The GPS dongle has rather low power requirements, though. No problems with my one phone that has OTG so far. I even had an unpowered USB hub plugged in, and two dongles powered at the same time. Rumors have it some people think two GPS units at the same time are a good thing.
OTG is a bit hit and miss powering up the gps dongle. Sometimes mine turns on immediately other times never. Doesn't seem to be any easy way to toggle OTG on the Motorola. Also the antenna is pretty weak on the dongle, using the internal gps I see 20 sats and use 11, the dongle only sees 13 and uses 9 at the same time.
OTG is a bit hit and miss powering up the gps dongle. Sometimes mine turns on immediately other times never. Doesn't seem to be any easy way to toggle OTG on the Motorola. Also the antenna is pretty weak on the dongle, using the internal gps I see 20 sats and use 11, the dongle only sees 13 and uses 9 at the same time.
I have noticed that the phone GPS used more satellites than the dongle. I have seen even larger on my phone, like 13 used on the phone and 7-8 on the dongle. The dongle still seemed to give better accuracy (although not having SDoP for the phone data makes that harder to verify).
I wonder how much of the difference is due to the antenna, and how much due to the GPS chip.The antenna in the dongle is obviously small, but I'm not sure the one in the phones is bigger - after all, their GPS resolution is adequate for the typical use. It's also easy to find GW-60 data with low SDoP at lower satellite numbers , and higher SDoP with more satellites (e.g. 0.3 with 8 sats, and 0.8 with 11, in the same track). I guess adding a satellite close to another one with better S/N ratio may not always be beneficial.
Hi, I'm developer of WindsportTracker.
how do you want this exactly ?
I mean currently you use WindsportTracker as this:
1. start app
2. press record button
3. turn screen of and go sailing
4. turn screen on, unlook
5. stop recording.
6. open session, and look for speed details.
you will then see you best avg. over 1h.
in between you can see the elapsed time, but currently not the avg. speed since start. only max speed and max speed for last run.
so what you want in addition is:
instead of max speed and last run max speed, you want the avg. since start till 1h, after 1h it restarts automatic or you can press a restart button ?
is that what you want ?
But consider, that those values wouldn't be your fastest 1h runs, as they could happen, between in between 2 1h attempts.
But, you can also press the stop button, and press the start button, if you want to train the 1h in a defined time frame.
As I planned to rework the recording screen the next week anyways, it's your chance to tell me your needs ;)
I really want to make the app better, but I also need feedback what users want ;)
I also have now some uBlox Dongles. I plan to integrate the support directly in the app. So no other App or tweak should be needed.
In order to get accurate data, OTG seems to be the better way, as you get "Raw" data with SDoP values. BT is just NMEA, as far as I know.
then we can all compare the results :)
and regarding OTG on Motorola, in my tests it worked pretty well.
and about the antenna, they are mostly very small in the phones, also keep in mind, that there is also a lot of noise in the phone (cpu's etc )
many phones already have uBlox GPS chips, but if you don't have a routed once, you only get the calculated data, and not the raw data.
phones with very good GPS from my tests are so far:
Sony Z3 compact, the best
Moto G 2015
Samsung S series.
Sony Z series
Phones with bad GPS are:
sadly all Google Phones (Nexus, Pixels) don't know why as those should be reference phones.
LG phones
cheap Chinese phones
with HTC I only have some experience but they looks also good.
This is just a thought, I'm imagining my self going for an hour. Yes a start button is needed, because it may take a slow upwind journey to get to the beginning of the hr. It would be good to see the accumulated average speed through the hour. Then when the hour is up, an indication of some kind. maybe a reversal of font/background. Then as you keep sailing, an indication of whether the hr is improving or decreasing. Maybe a readout of the best hr so far. This would help you decide when to stop trying.
This is just a thought, I'm imagining my self going for an hour. Yes a start button is needed, because it may take a slow upwind journey to get to the beginning of the hr. It would be good to see the accumulated average speed through the hour. Then when the hour is up, an indication of some kind. maybe a reversal of font/background. Then as you keep sailing, an indication of whether the hr is improving or decreasing. Maybe a readout of the best hr so far. This would help you decide when to stop trying.
ah ok, so you want to track your whole session, but while doing so you want a "second" start button to start your training run ? which in this case is 1h.
and while you training you want to see if its worth continue training or just continue sailing. ?
as could it be that conditions get better, you maybe want to reset your training ?
but once finished sailing, you will hit the stop recording button, and values could will be calculated.
so in case you reseted your training because of condition change to late, it could be that the final result for 1h could be higher then the value you saw during your training run.
did I get you right ?
I guess the 2nd start button wouldn't be all that necessary, if there was a continuous average read out, with an indication to say when you've actually done an hour. Then a display of the best hour you've achieved, so if the start of the session was slow upwind, then the hour readout would increase as the low speed start was dropped off. Once the number stopped increasing then you can decide to stop.
And "training" is the wrong word here I think, I'd be after a result, not training for anything.
And of course us GPSTeamChallenge sailors would like all our 6 categories. 2s average, 5x10s average, Nautical Mile, Alpha and distance, besides the hour. Preferably without having to stop and press buttons.
Don't want much do we?
I guess the 2nd start button wouldn't be all that necessary, if there was a continuous average read out, with an indication to say when you've actually done an hour. Then a display of the best hour you've achieved, so if the start of the session was slow upwind, then the hour readout would increase as the low speed start was dropped off. Once the number stopped increasing then you can decide to stop.
And "training" is the wrong word here I think, I'd be after a result, not training for anything.
And of course us GPSTeamChallenge sailors would like all our 6 categories. 2s average, 5x10s average, Nautical Mile, Alpha and distance, besides the hour. Preferably without having to stop and press buttons.
Don't want much do we?
Just scroll through the GPSTC categories would work.
Maybe Dylan needs to get on board and work on a KA72 enabled watch which does it all including uploading at the end of the session.
Preloaded with his algorithms
ah ok, so you don't want to train, you just want to get the best result :)
This means: currently only your max speed total, and from last run is shown.
In addition you want the other values there as well, either by scrolling, or on one screen, right ?
The question would be, how big does the text have to be ?
In the usual way, when you lock the screen after pressing the record button, you only have the notification available which have a small font size, not much to influence here.
On the recording screen it's easy to provide this, but this means to get there, you need to unlock your phone first.
Keeping the screen on, is nothing I would recommend, as you need full display light, which drains your battery. Also I need to "secure" the stop button that it don't get hit from a drop of water. Which makes it also a bit harder to stop. I know this from other apps and I hate it that it's so hard to stop ( even this are just running apps who don't really need such secure button)
Instead I would recommend to use audio.
The speakers are most of the times loud enough to hear even while sailing in stronger wind.
During winter season when you wear a helmet, or hod, waterproof BT ear plugs should work fine.
(using audio doesn't mean, the info is not available in the notification, or the screen)
what do you guys think ?
If it's 5hz or 10hz, max speed has little relevance as it's mainly a noise peak. 2s average is what we are really interested in.
It's a while since I used a phone, blue tooth earphones kept drowning, and the ones with cords kept getting tangled.
I just got fed up with all the hassle.
But from memory I preferred having the display locked on. Yes battery life was short, but if going for a long session the option of turning it off would be good.
for reading while sailing the font needs to be as big as possible, once stopped it can be normal size. So maybe if three divisions can fit on the screen, it could scroll between the 6.
If it's 5hz or 10hz, max speed has little relevance as it's mainly a noise peak. 2s average is what we are really interested in.
It's a while since I used a phone, blue tooth earphones kept drowning, and the ones with cords kept getting tangled.
I just got fed up with all the hassle.
But from memory I preferred having the display locked on. Yes battery life was short, but if going for a long session the option of turning it off would be good.
for reading while sailing the font needs to be as big as possible, once stopped it can be normal size. So maybe if three divisions can fit on the screen, it could scroll between the 6.
Maybe Gath Helmets could come up with a helmet the has BT ear pieces installed or optioned (like GoPro mount)
Then the KA72 watch could BT direct two (sic) your ears the sweet tail of PB accomplished.
If it's 5hz or 10hz, max speed has little relevance as it's mainly a noise peak. 2s average is what we are really interested in.
But from memory I preferred having the display locked on. Yes battery life was short, but if going for a long session the option of turning it off would be good.
for reading while sailing the font needs to be as big as possible, once stopped it can be normal size. So maybe if three divisions can fit on the screen, it could scroll between the 6.
+1
Being able to choose which category should be displayed would be great (GPSLogit has this already). Could be simple check boxes, with the font getting smaller depending on the number of categories. Sometimes 3 makes sense, sometimes just one is fine.
When working on the 1-hour category, and session time is less than 1 hour, showing the current average and the total time would make sense.
If it's 5hz or 10hz, max speed has little relevance as it's mainly a noise peak. 2s average is what we are really interested in.
But from memory I preferred having the display locked on. Yes battery life was short, but if going for a long session the option of turning it off would be good.
for reading while sailing the font needs to be as big as possible, once stopped it can be normal size. So maybe if three divisions can fit on the screen, it could scroll between the 6.
+1
Being able to choose which category should be displayed would be great (GPSLogit has this already). Could be simple check boxes, with the font getting smaller depending on the number of categories. Sometimes 3 makes sense, sometimes just one is fine.
When working on the 1-hour category, and session time is less than 1 hour, showing the current average and the total time would make sense.
Or just wear a watch on the other wrist with a clock, stop watch, countdown initiated..
A great big solar powered timer laminated into the sail displaying your top/chosen stat to the whole world
I also have now some uBlox Dongles. I plan to integrate the support directly in the app. So no other App or tweak should be needed.
In order to get accurate data, OTG seems to be the better way, as you get "Raw" data with SDoP values. BT is just NMEA, as far as I know.
Incorrect. From our custom Ublox based BT gps we get the full UBX ("raw') data. Otherwise it would be pointless.
If it's 5hz or 10hz, max speed has little relevance as it's mainly a noise peak. 2s average is what we are really interested in.
It's a while since I used a phone, blue tooth earphones kept drowning, and the ones with cords kept getting tangled.
I just got fed up with all the hassle.
But from memory I preferred having the display locked on. Yes battery life was short, but if going for a long session the option of turning it off would be good.
for reading while sailing the font needs to be as big as possible, once stopped it can be normal size. So maybe if three divisions can fit on the screen, it could scroll between the 6.
yes I know, but usually smartphones just have 1hz, so it's ok, to give a hint. thats what it was supposed to be.
but with the upcoming changes, you guys requested, the more relevant / correct data will be shown :)
ok, I will start with presenting the data, and after this thinking about a proper lock - unlock screen solution.
because then I can also test how to read the values best while surfing,
If it's 5hz or 10hz, max speed has little relevance as it's mainly a noise peak. 2s average is what we are really interested in.
But from memory I preferred having the display locked on. Yes battery life was short, but if going for a long session the option of turning it off would be good.
for reading while sailing the font needs to be as big as possible, once stopped it can be normal size. So maybe if three divisions can fit on the screen, it could scroll between the 6.
+1
Being able to choose which category should be displayed would be great (GPSLogit has this already). Could be simple check boxes, with the font getting smaller depending on the number of categories. Sometimes 3 makes sense, sometimes just one is fine.
When working on the 1-hour category, and session time is less than 1 hour, showing the current average and the total time would make sense.
Or just wear a watch on the other wrist with a clock, stop watch, countdown initiated..
A great big solar powered timer laminated into the sail displaying your top/chosen stat to the whole world
supporting the watches is already planned for next year ;)
I also have now some uBlox Dongles. I plan to integrate the support directly in the app. So no other App or tweak should be needed.
In order to get accurate data, OTG seems to be the better way, as you get "Raw" data with SDoP values. BT is just NMEA, as far as I know.
Incorrect. From our custom Ublox based BT gps we get the full UBX ("raw') data. Otherwise it would be pointless.
ah , ok, which uBlox BT dongle did you use ?
you said custom, so it's not public available ?
I just got some insights into dropped points from the USB dongle. My first guess was to blame Android, but given that even cheap phones have quad-core gigahertz CPUs and rather fast RAM, that make little sense. Instead, the culprit seems to be a combination of default settings and electronics. The Vk-172 dongle internally uses a serial-to-USB converter, and the default baud rate is 9600 baud. At rates higher than about 2.5 Hz (depending on what messages are selected), this is too slow, and leads to lost messages. In my tests, I lost about 10% of the messages at 10 Hz, and 3% at 5 Hz. Simply setting the rate for the UART port to 115200 Baud fixed this; I have a stationary test running that has not lost a single point in almost one hour.
This is a bit counter-intuitive, since the u-blox manuals state that the chip supports USB directly. USB does not have a baud rate, so it cannot be set in the config commands, and every USB version should have plenty of throughput. However, there are a few obscure references to serial-to-USB conversion in dongles that can be found on drone forums. Mostly, I just got lucky when playing around in u-center, and ended up noticing much fewer dropped points after setting the baud rate higher. I verified that the UART baud rate is indeed the cause of the problems yesterday and today, using different messages and tracking the loss of messages directly on the phone.
I'll do another driving test today, but my windsurfing tests are on hold until the end of January - temperature have dropped to below freezing at home. If anyone wants to do their own tests, send me a pm. At this point, some Android programming may be required, but perhaps that will change in a couple of weeks.
I just got some insights into dropped points from the USB dongle. My first guess was to blame Android, but given that even cheap phones have quad-core gigahertz CPUs and rather fast RAM, that make little sense. Instead, the culprit seems to be a combination of default settings and electronics. The Vk-172 dongle internally uses a serial-to-USB converter, and the default baud rate is 9600 baud. At rates higher than about 2.5 Hz (depending on what messages are selected), this is too slow, and leads to lost messages. In my tests, I lost about 10% of the messages at 10 Hz, and 3% at 5 Hz. Simply setting the rate for the UART port to 115200 Baud fixed this; I have a stationary test running that has not lost a single point in almost one hour.
This is a bit counter-intuitive, since the u-blox manuals state that the chip supports USB directly. USB does not have a baud rate, so it cannot be set in the config commands, and every USB version should have plenty of throughput. However, there are a few obscure references to serial-to-USB conversion in dongles that can be found on drone forums. Mostly, I just got lucky when playing around in u-center, and ended up noticing much fewer dropped points after setting the baud rate higher. I verified that the UART baud rate is indeed the cause of the problems yesterday and today, using different messages and tracking the loss of messages directly on the phone.
I'll do another driving test today, but my windsurfing tests are on hold until the end of January - temperature have dropped to below freezing at home. If anyone wants to do their own tests, send me a pm. At this point, some Android programming may be required, but perhaps that will change in a couple of weeks.
Doing my head in!
Here's something easier - the comparison of the USB dongle at 10 Hz (left side) and the GW-60 at 5 Hz (right side), top 5 10-second speeds from a test drive. The top 5 numbers are all very close, and the ranges overlap as they should. Missing points were not an issue after configuring the chip properly. Numbers for the phone's GPS as recorded with GPSLogit showed larger deviations (around 0.5 knots). Once again, encouraging results. Proper analysis will take more time...
I also have now some uBlox Dongles. I plan to integrate the support directly in the app. So no other App or tweak should be needed.
In order to get accurate data, OTG seems to be the better way, as you get "Raw" data with SDoP values. BT is just NMEA, as far as I know.
Incorrect. From our custom Ublox based BT gps we get the full UBX ("raw') data. Otherwise it would be pointless.
ah , ok, which uBlox BT dongle did you use ?
you said custom, so it's not public available ?
Correct. They is not one commercially available now sice the Winter G-Rays stopped production. These one were home made as explained in other threads.
Full ubx data up to 18Hz.
I just got some insights into dropped points from the USB dongle. My first guess was to blame Android, but given that even cheap phones have quad-core gigahertz CPUs and rather fast RAM, that make little sense. Instead, the culprit seems to be a combination of default settings and electronics. The Vk-172 dongle internally uses a serial-to-USB converter, and the default baud rate is 9600 baud. At rates higher than about 2.5 Hz (depending on what messages are selected), this is too slow, and leads to lost messages. In my tests, I lost about 10% of the messages at 10 Hz, and 3% at 5 Hz. Simply setting the rate for the UART port to 115200 Baud fixed this; I have a stationary test running that has not lost a single point in almost one hour.
This is a bit counter-intuitive, since the u-blox manuals state that the chip supports USB directly. USB does not have a baud rate, so it cannot be set in the config commands, and every USB version should have plenty of throughput. However, there are a few obscure references to serial-to-USB conversion in dongles that can be found on drone forums. Mostly, I just got lucky when playing around in u-center, and ended up noticing much fewer dropped points after setting the baud rate higher. I verified that the UART baud rate is indeed the cause of the problems yesterday and today, using different messages and tracking the loss of messages directly on the phone.
I'll do another driving test today, but my windsurfing tests are on hold until the end of January - temperature have dropped to below freezing at home. If anyone wants to do their own tests, send me a pm. At this point, some Android programming may be required, but perhaps that will change in a couple of weeks.
We found we had to configure the BT Baud rate to 115200 with the BT GPS as well, for the same reason.
ah ok, so you don't want to train, you just want to get the best result :)
This means: currently only your max speed total, and from last run is shown.
In addition you want the other values there as well, either by scrolling, or on one screen, right ?
The question would be, how big does the text have to be ?
In the usual way, when you lock the screen after pressing the record button, you only have the notification available which have a small font size, not much to influence here.
On the recording screen it's easy to provide this, but this means to get there, you need to unlock your phone first.
Keeping the screen on, is nothing I would recommend, as you need full display light, which drains your battery. Also I need to "secure" the stop button that it don't get hit from a drop of water. Which makes it also a bit harder to stop. I know this from other apps and I hate it that it's so hard to stop ( even this are just running apps who don't really need such secure button)
Instead I would recommend to use audio.
The speakers are most of the times loud enough to hear even while sailing in stronger wind.
During winter season when you wear a helmet, or hod, waterproof BT ear plugs should work fine.
(using audio doesn't mean, the info is not available in the notification, or the screen)
what do you guys think ?
Yes, just a Start/Reset button for the 1hr would be good, with a large font readout so you can track average speed during the 1hr. At the end it would perhaps split into two readouts, one with best 1hr since last Start/Reset and the other continuing to show the average speed since the last Start/Reset. That way, if the conditions improve, you can continue to try and improve on your best 1hr.
A brief update is at boardsurfr.blogspot.com/2018/01/dongle-update.html with the perhaps most interesting part that the Motorola/Lenovo E4 phone supports OTG. I got it for $40 here in the US, but it will probably we 2 or 3x that much in other parts of the world. Did the first short driving tests with 2 dongles at 1 Hz, and they seemed somewhat more accurate than the GW-60. Hoping for some windsurfing tests soon...
A brief update is at boardsurfr.blogspot.com/2018/01/dongle-update.html with the perhaps most interesting part that the Motorola/Lenovo E4 phone supports OTG. I got it for $40 here in the US, but it will probably we 2 or 3x that much in other parts of the world. Did the first short driving tests with 2 dongles at 1 Hz, and they seemed somewhat more accurate than the GW-60. Hoping for some windsurfing tests soon...
Very interesting.
Just an aside.. my VK172 no longer works in my Chinese otg phone. Measuring the power output of the otg Port was mostly zero ,sometimes 1.7v , sometimes 2.7v and occasionally 5v. Reading online I found a site that used to sell VK172 but no longer did. Said about 100 out of 1000 were detective. He also said the VK172 was very susceptible to voltage fluctuations and would stop working if v went to < 4.8v.