Quick bottom line:
| Chinavasion | Professional, responsive and web-competent. |
| Travel Honey Watch | Not much value. A bust. |
The GlobalSat GH615 watch is falling apart.
I like the watch form-factor for GPS.
- The wrist is a very handy place to carry a GPS unit.
- A watch must necessarily be nicely small.
- Synchronizing the camera’s clock with geo-tagging photos is easy. Take a picture of the watch’s GPS time. Then sync the picture in TrodTrack to the time on the watch. I do this for every hike or photo session. The camera’s clock loses a second every week or so.
What to do?
Turns out, GPS watches are rare. I’ll sadly share credit with GlobalSat for ruining the GH615’s excellent chances for success in the market. Garmin Forerunners can be had for a bit over $100 on eBay. Forerunners look kind of klunky, at best. And, new, they are overpriced.
So what to do?
A web search found Chinavasion and the unfortunately named Travel Honey watch.
Note to manufacturer: In the U.S., “honey” means either the tasty stuff that comes from bees or this. Your products don’t come from bees.
On the subject of names: Chinavasion? Guys, if my experience with you is representative, you’re on your way to the top. But, consider, what would your impression of a Japanese company named Japanvasion be?
It took a two or three weeks to get the watch by mail from Hong Kong. No problem there. ‘Bout what you’d expect.
Out of box:
The shipping box perfectly fit around the product’s box. Wow!
Software
The included iTravel software is a finished product. Its Google Maps code is better than my TrodTrack code – faster and with a couple of nice spiffs. The track point editor is a nice thing. The UI layout looks good and well thought out.
It took a product key to get the program to talk to the watch. The key was not in the box so I got one from Chinavasion by on-line chat and email. Who knows whether it’s paid for. Anyway, it worked.
But I won’t be using the iTravel software except on the laptop while traveling. I have used an open source Linux program to pull the tracks off the watch. The watch protocol is documented and if I were to use the watch, I’d probably end up writing Python code to talk to it. But I won’t be using this watch as a primary GPS.
Watch
The watch is smaller than the GH615. That’s nice.
The time-keeping part of the watch, itself, is bargain basement. “Uselessly inaccurate” might be the most accurate description. And, since it’s not a GPS-time watch, it cannot be used to sync the camera time.
The GPS is provided by a SkyTraq Venus 6 GPS chip. In this watch the GPS is clearly inferior to the SirfIII in the GH-615. It loses its way in Northwest forests often and without fail.
This is a killer.
There are other problems. For instance, I have the GPS set for 1-second samples. It occasionally switches to 5-second samples and/or no sampling. The only way to get the watch back working is to reset the settings through the PC software.
So, this Travel Honey watch was a nice experiment. I’d wanted to see how another GPS chip matched up against the SirfIII. Now I know. It could be that the weakness of this GPS is in the small, watch packaging. But why chance it? I’ll probably get a normal GPS logger that uses the SirfIII chip. And, knowing me, I’ll probably end up using the GH-615 for another couple of years.
I’m inclined to get another gadget from Chinavasion. They (and many other outlets like them) certainly open a window in to another world. … so many gadgets at cut-rate prices of probable cut-rate quality.
This other world is interesting. During the 80’s and 90’s Taiwan cranked out a lot of PC boards and such-like in white boxes for low prices. One would have expected that the quality of such devices would be low. But that was not the case. Compared to the “name” brands, they were almost always:
- Cheaper
- Simpler to install and use
- Higher quality
- More powerful
- Even with fractured English, often better (geekier) documented
My gut feeling is that these eleven-teen jillion Chinese gadgets are not like that. They give off an aura that matches the Travel Honey watch: cheap junk with a promising core. Think Japanese products from the 50’s and early 60’s.
Anyway, this evolutionary process will be fun to watch.
Maybe there is a more recent firmware update for your chip, or the battery it came with is almost dead, (that sounds as if it could cause the problem with the update speed changing) or the connection between the internal module and its antenna is disconnected, or broken. Something is not working and its quite possibly something fixable.
I would take a close look at your particular watch (if you have not returned it yet) and see if perhaps there might be some issue like an antenna problem. Its also possible that some configuration change might improve the situation.
First you should email the watch manufacturer and let them know that you are having problems, and ask if there is a setting you can change that will make your solutions more accurate in your situation.
I would also experiment a bit with the Skytraq GUI utility GPSViewer and see if changing some of the operating parameters suddenly improves performance. Those chips have a great many settings that effect performance in different ways.
And finally, if its not too difficult, I would try to open it up and take a look inside to see if anything is visibly wrong, unless doing that would make it impossible to exchange for a working unit (and an exchange is a possibility.)
If none of that works I would then email info@Skytraq telling them that you had tried the above already, and be specific about what happens.
GPSViewer can tell you the date and kernel version of your chip’s firmware – you definitely want to include that info in your email..
They respond to emails, unlike many companies, their customer service is good. And their products are unique both for performance and for value.
I have a SirfStarIII based unit and a (industry leading brand PND) unit and three Skytraq based GPS dongles (two different chipsets) and the Skytraq GPSs are substantially better performers. They also are much more configurable, so one can easily mess up there unless you know what you are doing.
Wow, chrix, thanks for the suggestions!
That is loses the logging frequency setting does sound like a battery thing, now that you mention it.
I wondered whether the performance on these devices is heavily influenced by the packaging and electronics around the device. That the Honey Travel watch is a very small package seemed like a good explanation for why it had troubles. But your experience with dongles indicates that good GPS units can come in small packages.
It would be fun to get a whole gob of GPS units and take them on a hike through these Northwest woods to compare them, apples to apples.
It would be really interesting, and this is a really interesting area, Ive been finding out.
Let me give you my cut on it.. This is an extremely complicated problem. Lots of issues come into play.
If I was putting a GPS module behind or in front of a watch face to wear on a wrist, the number one issue is the antenna placement. I would face really daunting tradeoffs in placing the antenna. Also, the fact that the position of the hand is moving would make it difficult to resolve satisfactorily with antennas that do not present an unacceptably bulky profile.. (the best kind of antenna for this situation would be a special kind of helical antenna that is basically a phased loop – the only manufacturer that I know of is Sarantel.. but they are big, too big for a watch.. That leaves patch antennas..
There are plusses, though, I don’t know how much they would help..
If the antenna is tuned without a person wearing the watch.. someone putting it on will definitely detune the antenna, so it would need to be designed for somebody’s wrist being there..
However, there is a plus for a cheap GPS in that your wrist represents a temperature controlled environment and your body’s own thermoregulation provides excellent temperature stability relative to almost any other situation found in nature.
This would mean that the crystal oscillator might be more accurate than they usually are in a consumer grade device..which I would expect would help the most on a cold start..
Come to think of it, a GPS probably takes more current than the typical quartz watch.. again the skytraq gpss are very good in that respect but they still draw around 30 ma while on and acquiring data.. not including the logger part, writing that data to a flash would be an additional drain. That might mean that it would need a rechargeable battery, I doubt if a long life lithium bbattery could deliver the power required to keep it running and logging, (I use a lithium battery to keep the RTC running and allow the memory to save the last ephemeris – A warm or hot start often takes just a short time to get a 3D fix.. but a cold start always takes at least 13 minutes.. sometimes longer..
The consumer doesn’t want to deal with all this..
It is really interesting to compare the tracks recorded by several GPS loggers as you go from one environment to another. A new kind of sensor that is being exploited in some GPSs to improve their accuracy are compasses and accelerometers.. Another hot area is the use of something called AGPS in devices that have net connectivity to download the ephemeris quickly over the net, rather than downloading it very slowly from the first visible satellite.. (thats what causes the 13 minute wait) Another very cool area is differential GPS which requires two GPS units, one is the rover, the moving GPS, the other is a stationary receiver at a known position, which receives the same data and allows a super accurate positional solution to be computed. Thats what I am trying to figure out how to do today (with an open source library, RTKlib, and a Skytraq chip) which is why your blog came up in my Google search.
If you could add an external antenna connection for your GPS and then an active antenna someplace high on your body.. (maybe under a hat.. or on a shoulder or the top of a backpack.. that would give you a much nicer looking and more accurate trace…)
The Sarantel “geohelix” antennas are the best ones for situations in which you dont know which side of the GPS is going to be up. Try to get all the info you can on the specific hardware that you can, be seriously anal about it, before you do anything, get the datasheet for your chipset or module, and read it very carefully. Otherwise, you are very likely to destroy it. If you look inside the GPS watch, before you do a single thing, make sure to read up on preventing ESD destruction of your chip from static electricity. It sounds like you might be near the coast, which is good news for sensitive electronics, because on the west coast that means more humidity – but an indoor environment where AC is going can have low humidity and so, still have dangerous problems with static electricity even if the RH is over 60% outside.. If you damage your chip you will be very remorseful that you made such a stupid mistake.. but its very, very easy.
Also, some of these skytraq chips are supposed to be dried out in an oven (not your food oven!) before they are assembled onto boards if they have been in high humidity.. My friend suggests a short bake in an oven at a low temp evenbefore working directly on them with solder..also you need a soldering iron that is adjustable temperature, you want to use the minimum possible temp, and make sure it has a grounded tip and that both it and you also are grounded.
Sparkfun.com has info on all this stuff, check it out..
chrix, you’re in to the details of this way beyond what I’m doing, but your info is fascinating.
Yes, near Seattle, static electricity isn’t that big a deal, usually. It can be somewhat wet around here. 🙂
Both kinds of watches I’ve used have the antenna in a sort of smoothly integrated part of the bezel between watch and band. When I’m in the forest or when the device is first turned on, I tend to position the antenna facing up and clear to the sky. A watch form-factor is very, very handy. But the downside when hiking is the watch swings to and fro at waist level … at best. Maybe hooking the the watch around the grab-loop at the top of the pack might work better. Maybe not. A head could be a pretty solid barrier.
Your mention of the effects of things as small as the watch hands (the Honey Travel has hands, the Globalsat does not), jibes well with a gut feeling I’ve gotten about larger environmental things affecting startup, etc. Car windows. Some cars seem to have windows almost impervious to GPS signals at startup time. And, I’ve wondered whether pushing power-on with the device already positioned as it will be for the first minute helps get the thing started.
My GPS track processing is at a high level. The watches log the lat-lon-alt every second. One thing I’ve noticed is that two watches – even two similar Globalsats – put out very, very similar tracks in space. But they differ in time. Car tracks are notable. One watch will have me going through a freeway on-ramp a full second or two before the other. They both get the side of the road I’m on and everything. So the spacial information is dead on. But the times don’t match. I’d think that the filtering they are doing internally would affect the spacial info more than the temporal info.
Roger that Sparkfun. Occasionally hit the site, though I’ve never quite had enough incentive to break out the 35 year old soldering iron.
Yes, if you put your watch (or any GPS) near the top of a backpack, you’ll get a visibly smoother GPS track. You’ll see how it will be noticeably more accurate. I would think that having a GPS in a watch would be sort of like having it in a car windshield, the sky view it is going to aquire its fix from is one in which its horizon is skewed to that side.. typically the left. On the top of a backpack, though its higher, I typically carry my “hiking” GPS (a PND that has a SIRFstar III chip in it, in my backpack in an outer, upper pocket when I’m recording a track while hiking. There I get pretty good accuracy, much better than if I keep the GPS in a pocket. Getting it even just a few inches away from your body helps a lot.
Gee, your talking about hiking in a real forest makes me miss the West Coast. I am in the NYC area right now, but Ive lived in Northern California in the past, for many years, and Ive really enjoyed that fresh sea air. Here right now its hot and humid.. lots of t-storms.. Not my favorite weather..you can give me coastal fog with its 50 degree summer days, or the West Coast inland’s dry heat over this anyday..
There is a GPS-enabled satellite tracker/SMS messaging unit (forget its name) that they market for heavy duty hikers and campers who want their families to be able to se where they are (and that they are okay) and need a way to get a “help” message out anywhere in the world, and they strongly recommend carrying it as you described, near the top of your backpack where it can get a sky view. When carried like that it can get a good sky view to get a fix and a good shot back up at the satellite so it can send its data to mother ship (for use on its web mapping app), Carried like that it can track people reliably even way down in the narrowest slot canyons.. That crazy guy who got stuck in one and had to hack his arm off with a dull Leatherman lookalike would have been able to get help easily with one if he followed their advice.
His story is pretty awesome, I think..
Thats strange that the times are off on the two different GPSs. I think that for a GPS to function it needs to (internally) keep time with a level of precision that rivals the very, very best timepieces available to mankind.. The GPS/GLONASS/Galileo satellites.. (the three different US/Russia/EU GNSS systems) all contain atomic clocks that are accurate to something like one part in 10^12 – and are synched to literally the best clocks in the world.. to the nanosecond.. So that time offset must have been inserted after the location solution was done.. something in your logging toolchain is adding it.. otherwise the satellite fix would not have been done, and the logs would just be plain wrong..
If you look at the NMEA logging data as its coming in, typically when you turn a GPS on from a cold start, it will start up with some arbitrary date before it was manufactured.. As soon as it gets its first satellite it will know the time accurate to one second, and the approximate date of the year, but I think it can be a few seconds after that before it knows the year.. once it has a lock on three satellites it can give you a 2D location fix (sans altitude) and once it has four sats it will compute your altitude and start giving you that nanosecond accurate time..it should also provide somewhere a digital signal that pulses once a second which is usually configurable to be available to external instrumentation as a calibration signal. That so called 1PPS is an extremely useful signal because the combination of the NMEA text string and that pulse is the gold standard for microsecond-precise timekeeping. Using it you can do all sorts of cool stuff like knowing where lightning strikes..and how fast signals travel in cables and over the net- You may start finding out how many microseconds of time are added by various devices, or even a length of cable…its useful to see where the bottlenecks – or problems caused by so called “bufferbloat” are.
Precise timing applications, which GPS now makes possible, are like Prometheus giving fire to mankind.. its a force multiplier for science.
This looks like an interesting web site on antennas.. I found it on a search and had bookmarked it to stick in my earlier post but I forgot to add it..
http://www.antenna-theory.com/intro/history.php
(actually the whole site, which I am going to explore with this NEC antenna design program I have but never have gotten very deeply into, for lack of a resource like this to make sense of it..)
I have to check it out later on..