Navigon, TomTom, the iGPS360 and... debugging GPSSerial

Link for iPhone and iPod Touch users here

This is a post that describes how GPSSerial was debugged to fix the heading bug that appeared in Navigon and TomTom.

As you can imagine, it is very time-consuming to get on the road and do real-life GPS testing every time the code for GPSSerial is updated. As such I decided to record the output of the iGPS360 module to a file while driving my car on a route that was known to have heading problems in Navigon.

What you see in the above video is an in-door setup where the iPod Touch is connected through an interface board to my desktop computer. The desktop computer in turn plays back the on-the-road recorded GPS trace through the interface board. This setup is a huge time-saver when it comes to testing bug-fixes in the GPSSerial source code.

In the course of the debugging process, I also updated the iPhone Toolchain on the desktop PC (running cygwin) to firmware version 3.1.x as I first thought that the bug was related to the Toolchain version. This was quite an experience in itself: not without problem; although the instructions at Saurik's page are very good, they are not complete, and neither up-to-date (yet) for the 3.1.x firmware.

After having spent a lot of time on testing and debugging the code, all the work has finally paid off: the latest GPSSerial version is finally working correctly with Navigon. The incorrect-heading bug is now fixed and the red GPS bar does no longer appear when you are not moving.

You can download and install the latest GPSSerial version by following the instructions on the Orange Gadgets site here.

The new version should hopefully be available through Cydia soon as well.

Try it out and feel free to leave your feedback in the comments.

Enjoy! PmgR

Confirmed: iGPS360 works with firmware 3.0

This is a brief post to update everyone on the iGPS360 module and 3.0 firmware.

As had already been posted on the previous post by quite a few users, the iGPS360 works with the new 3.0 Apple firmware (thank you community, for bringing this confirming news early for everyone).

I have finally had some time myself to upgrade a 2G iPhone to the 3.0 firmware and I can confirm that the iGPS360 module works fine on the new firmware.

The FAQ and product description on the Orange Gadgets website has now also been updated with this information. An updated section about shipping has also been added to the FAQ.

If you feel anything else needs to be updated or added to the website, please let me know.

I would also like to be able to confirm if any of the AppStore navigation apps works with the iGPS360 module; in particular people have been inquiring regarding G-Map, iDA, Navigon and Navionics.

If you have purchased any of these apps, consider getting in touch with me by email at pmgriphone@gmail.com so that we can officially confirm whether an application works or not.

Exactly 200 GPS modules left!

It's been a little while since the last post. Things have been very busy for a number of reasons.

As the title of this post already announces, at the time of writing this post, there are only 200 GPS modules left for sale.

I expect these remaining modules to be sold out within the next 5 weeks, so if you are considering buying a GPS module, you have to be quick!

Back when the first production run was initiated in September last year, I wanted to make 1000 modules available to you all. As a matter of fact, at that time, it was unclear what the yield of the module in production would be and as such I decided at that time to make 1100 modules instead of 1000. As nothing yields perfectly (100%) in production, I decided to put a safety margin of about 10% to cover for any yield loss.

Interestingly enough, the yield has been a lot better than that! To give you some statistics about the 1100 modules that have been produced so far:

1. Assembled PCB boards
   
   • One board had a short in a copper trace. This short was removed.
     
   • One board had a defective blue LED. The LED was replaced.
     
   • One board had an open contact on the USB connector. The connector was resoldered.
     
   • Three boards had a non-communicating GPS engine, rendering the boards in-operable (real yield loss).
     
   
   
2. Injection molded housings and lenses
   Considering the issues that had come up during the housing development, I had 1200 housings with lenses made instead of 1100.
   
   • Housing top: about 90 pieces were screened out because of imperfections.
     
   • Housing bottom: about 55 pieces were screened out because imperfections.
     
   • Lenses: about 65 pieces were screened out because of imperfections.
     
   
   
3. GPS performance (open and clear sky)
   
   • All tested modules passed (1097 pieces)!
     

So when it comes to the PCB, the yield has been extremely good (>99%). Only the housing has a relatively large yield loss (5%-10% overall) and I have already talked with the housing supplier to get this improved for the next run. The housings will no longer be bulk-packed and shipped by a courier (which I think has been the main factor of the imperfections). Instead each housing top and bottom will be layer packed with foam for the next run and I will go and pick up the housings myself to avoid possible damage during shipment.

Looking back over the last six months, sales have been a slower than expected, especially January was a slow month. Sales have picked back up in February, and especially March has been very busy.

As the remaining units are expected to be sold in about 5 weeks from now, I have had to put myself into high gear the last two weeks getting everything ready for the next production run. Typically, it will take another 4-6 weeks before we can expect new modules ready to ship from this second production run.

Since this is a second production run, hopefully things will go a lot smoother than was the case for the first run and hopefully no unexpected problem will show up.

In any case, I want to thank everyone for their support over the last year (yes, it's been already more than a year; the first post on this blog dates back to February 2008).

It's amazing to see that when one person in a certain country orders a module, shortly afterward a couple of more orders from the same area in that country come in! So keep on spreading the word about the module!

GPS voice navigation for the iPod and iPhone

Link for iPhone and iTouch users here

Link for iPhone and iTouch users here

I have been playing around with software related things the last couple of days.

I decided to upgrade the iPod Touch 1G to firmware 2.2.1 (the video above was taken at firmware 2.2) and I also upgraded the iPhone 2G to 2.2.1.

Unfortunately, I had some problems with both the iPhone and iPod getting stuck at the Apple logo screen after having re-installed/synced AppStore apps on them with iTunes.

So that's why I ended up upgrading the iPod to 2.2.1 (after it got stuck at 2.2) and downgrading the iPhone again from 2.2.1 to 2.2 (after it got stuck at 2.2.1). I think the Apple firmware has a fundamental flaw in the way it boots, trying to rebuild the SpringBoard application list which takes a long time and eventually times out, freezing the iPod/iPhone at the Apple logo screen. I searched online to see if there were any fixes, but none of them worked, so I ended up restoring both the iPod and iPhone.

It looks like things are working again now as I haven't gotten stuck at the Apple logo (I have removed most of the AppStore apps that I had downloaded though and did a reboot after every single package I installed with Cydia).

I also installed the latest beta version of Xwaves' xGPS software to see what had changed in the period I was gone.

There have been quite a lot of bug fixes and additions to xGPS. You can track the development roadmap here. The major new additions for version 1.2 are:

• A first beta version of voice navigation in English is now implemented and the speech engine does a pretty good job of pronouncing the directions (see the first video above). It will need some future tweaks to correctly pronounce abbreviations (like Rd, Blvd, Pkwy, N, E, S W, slashes, etc), but other than that it does a very good job. It would also be nice if the user could set the distance(s) at which voice directions are given; they are currently set at about 400m and 20m before the next turn.


• GPS coordinates are now accessible under the GPS settings menu. It would be nice if they could also be displayed in the bottom menu bar of the map view, similarly to the Google Earth application.


• A Night Mode view is now available under the User Interface settings menu. Night Mode can either be manually enabled or automatically through a timer from 8pm through 7am. The timer is currently fixed, but will be fully customizable in a future release.


• You can also have your iPhone/iPod play music at the same time when using the xGPS application. The voice directions will be played through the music (see the second video above). It would be nice to have the music fade in/out when there is a voice direction.

According to the roadmap, the next official release is scheduled for February 20, 2009. I am sure you are all looking forward to it.

GPS module working with iPod Touch 2G

With this post I am planning on regularly making posts on the blog again :-)

After a rough start of the year, today it is time for some great news!

Just last Saturday, January 31st, I got an email from Matt, an Australian customer who ordered the iGPS360 module a little before Christmas.

Matt has an iPod Touch 2G and ordered the iGPS360 module in anticipation of a jailbreak for the iPod Touch 2G, which was not available at the time he ordered.

Matt send me the above picture, showing the xGPS software running on his iPod Touch 2G. You can see the iGPS360 module in the upper right corner.

Matt used Redsn0w from the iPhone Dev Team to jailbreak his iPod Touch 2G. It is certainly not as easy a jailbreak as for the old iPod Touch and iPhone, but it certainly works as Matt's photo shows. For more information on how to jailbreak your iPod Touch 2G, please visit www.redsn0w.com.

The jailbreak is a tethered jailbreak, which means that you will need the assistance of a computer if you want/need to reboot your iPod Touch 2G for some reason.

Matt had upgraded the firmware on his iPod Touch 2G to firmware 2.2.1 before using the jailbreak. This thus also confirms that the iGPS360 module works on the new 2.2.1 firmware.

Matt has also posted his success of the jailbreak on iPodTouchFans, so if you are interested in seeing how many people are jailbreaking their iPod Touch 2G with Redsn0w and their success rate, take a look at that forum.

Matt, thanks for your feedback and great photo contribution!

Put to the test!

A brief update for today:

The four CNC machined housings arrived this week and all of them make a nice fit as you can see in the photo (I masked off the LED opening with some tape to simulate the lens effect as those lenses were to small to be CNC'ed). I have been in touch with the housing manufacturer this week as well to iron out the last details of the drawings. They will start cutting steel upcoming week! Once that is done, I will visit them to take a look at the molds before we get our first parts.

I built three additional GPS modules this week as well and the new upload tool for A-GPS data from the manufacturer seems to work a lot better now. After having uploaded the latest A-GPS data to each of the modules, I have performed open/clear sky testing with the modules and am very happy with the results! Typical time-to-first-fix (TTFF) with A-GPS enabled is ~10sec average on a cold start. Average signal-to-noise-ratio (SNR) is ~42dB (note that both the TTFF and SNR numbers are open/clear sky numbers and can vary in practice depending on surroundings and GPS satellite strength). Performance is thus very good for such a small unit! After the antenna tuning process, we might gain a little more on the SNR, but we will wait for those results first before concluding anything.

I also had to place a special purchase order (PO) for the high brightness/low current bi-color LEDs this week as they only had 200 in stock (delivery is scheduled for September 15, 2008). As I thought the green/orange color combination was a little too ordinary, I have chosen a blue/red LED which looks really cool. The blue LED will be on as power-on indicator and the red LED will flash when searching for satellites (it will be off after a GPS lock is found). So while searching for satellites, the combined blue/red LED will flash purple. Power consumption of the LED will be a little less than 1mA when the module is tracking. The rest of the module consumes ~36mA while tracking for a total of 37mA for tracking.

Power consumption, size and performance have all been traded off against each other, but it looks like the final results are pretty good in my humble opinion!

The estimated-time-of-arrival (ETA) for the modules is still set for the end of September. Let's hope there will be no further delays; thanks everyone for hanging in there!

What's (been) happening?

It has been a little while and as you can imagine things have been busy and very hectic. Here is an update on the various items:

1. Last Monday, the last revision of the PCB came back. Although I placed it as a rush order, it was late; it was supposed to have arrived the Thursday before. In the meantime I have built a new GPS module with it and tested its performance, which is up to spec now. This new module is shown in the photo (inside of the new housing prototype).


2. The biggest milestone that got done this week is the new two-piece housing design. This has pretty much kept me busy full-time since last week:
• New 3D CAD files (no logos) were completed and a new FDM prototype was made to verify that everything fits inside of it. I painted the prototype black, put the latest board inside and hooked the USB cable up. The result can be seen in the photo.
• The next thing in line was sending out the new 3D housing designs to get real prototype housings CNC machined out of solid ABS plastic (no FDM). These are needed to do proper antenna tuning of the board and the housing for optimum GPS performance. These housings are expected back upcoming week.
• Next I had to finalize the drawings for production and place the logos on them. This did not go without hick-ups and quite some time was waisted trying to get the 3D software to accept the graphics files. But eventually I managed to get everything imported.
• I am currently detailing the lens a little more to make sure that the LED light comes out properly and has a wide viewing angle. The final CAD files are due first thing Monday morning.
• A purchase order (PO) was placed for the injection mold making and a delivery of 1000 housings. The housings are to be completed by September 9, 2008.


3. I also received new GPS modules with updated firmware for this new PCB design. I had problems uploading A-GPS data to them. I thought it was a firmware issue, but after talking to the GPS vendor it turns out that the A-GPS upload tool that they provided had some problems, so they provided me with an updated version which performs better but still seems a little buggy. After we get a final OK from the vendor on the module configuration and firmware, we can place a PO as well (I hope this can happen this week).


4. Regarding contract manufacturers, I have converged on a candidate for PCB manufacturing and assembly with competitive pricing. I might have to make a few minor adjustments to the bill of materials, but after that we can get the ball running.

It looks like the Orange Gadgets website has been running pretty smoothly. As of now, 177 people have signed up so far for a total of 208 modules. Feel free to do some additional advertising among your family and friends as well as online.

Bingo!

Link for iPhone and iTouch users here


To top the week off with some more exciting news, what you see above is a live demo of the xGPS software that xWaves is working on. Note that all credit for this awesome application goes to them!

This software does not require a live internet connection!

Here is how to get this running:

1. I configured one of my modules at a baudrate of 19200. That is the baudrate the xGPS program currently runs at. Hopefully they will add in an option to custom set the baudrate the program uses for the serial port in the next version. So you will need to have a GPS module with the correct baudrate settings, otherwise it won't work. Plug your GPS module into the dock.


2. Go to the Installer.app on your iPhone or iPod touch and add in http://irepo.xwaves.net under the sources. Refresh the sources and then go to the Install menu and scroll down to Xwaves and install xGPS (the version I am running is 1.0p1). Quit Installer.


3. Start up the xGPS software and hit Maps. You'll end up somewhere in France. Zoom out with the small magnifier glass and move the map to your local area, then zoom back in again with the large magnifier glass.
   


4. Now drag with two fingers a red box for the map area that you want to download and hit the Download button. This will take a while depending on how large the map is. Make sure auto-lock on your iPhone/iPod touch is set to never and that you have enough battery left. I suggest downloading a small map area the first time (< 20mb). Cancel to go back to the main screen. The map is located in /private/var/root/Library/xGPS_map.db.
   


5. Then hit Enable GPS and voila, your GPS should be working! A pin will drop on your current location.


6. Take your phone, get into your car and have fun!
   

As this is still a very early version of the software, it is a little bit buggy when it comes to zooming in and out (double tap with one finger is zoom-in, double tap with two fingers is zoom-out). Also, the voice navigation is not working yet, although I did get it to display the purple-line directions on the map occasionally and the program spoke the first navigation direction out loud.

It looks like things are coming along very well in having a complete GPS navigation solution available.

So make sure up to sign up for one of the iGPS360 modules. The first production run will start very soon and will be a quantity of 1000 units. Make sure you are among the first 1000 to sign up!

Micro USB cable samples have arrived!

I have received three samples of the USB-A to micro USB-B cable. The cables are 6ft long which should be plenty to hook up to your module from wherever you want. I have already tested all three cables with my modules and they are performing fine, including syncing of the iPhone from iTunes.

I will give you a brief progress update so you know where we are at:

• Housing: A housing manufacturer has been found. I have visited him twice and toured the facility and liked what I saw. They have also been very helpful in getting the housing prototypes done. I will be updating the housing design next and after that get a revised quote. Mold making will take 3-4 weeks ARO and then the first parts will be delivered one week after mold approval.
  
  
• PCB board assembly: I have received about half of the quotes that I requested. The remaining ones are expected by the end of this week. Typical lead time for 1000 units is about 4 weeks (this includes PCB fabrication and assembly).


• PCB board re-spin:Just today I sent out another PCB prototype board design that will address the somewhat lower performance on the last set of prototypes. This board is expected back mid next week. By the end of next week, I should have build another three prototypes with these new boards and have tested them for performance.


• USB cables: USB cable samples have been received and tested. An order will be placed tomorrow. Lead time is 6-8 weeks.


• Other: Some other things are running in the background in parallel... I'll keep you guessing at what that may be for now...
  

The Good and the Bad...

The photos show two rough prototype housings made on a fused deposition modeling machine (FDM). As the resolution of these machines is about 5-10mils (127-254um), the surface of the housing is not very smooth, but it is good enough to do form fitting of the board and dock connector. Also, the housings look white as the machine uses this color plastic to do the FDM. The ultimate housing will be black.

The next photo show you how the module will look on both the iPhone and iPod touch.



Now for the good news:

• All the clearances for board to housing were designed correctly and the boards fit nicely within the housing.


• I have also received the A-GPS modules in the meantime and built a new module with it and put it to the test. The way it works is as follows: you first download 7 or 14-day valid data to the module (this can be done through software), then use the module as normal. I have put this module to the test and compared it with a module without A-GPS. For the A-GPS module, the time-to-first-fix (TTFF) in open sky after a cold start was reduced from 43sec to 29sec (average over 5 measurements) with times as short as 16sec.


• A large volume of dock connectors for the modules have arrived!


• NDAs have been put in place with quite a few PCB contract manufactures and quotes are expected back soon.

Now for the bad news:

• It turns out, there is no way to get the plastic lens into the LED opening of the housing and have it sit straight and tight. So it's a good thing to have built these prototypes to do form-fitting before getting a housing injection molded for real. We will have to come up with something to solve the lens problem: the housing will be split in two half so that the lens can be inserted from the inside and the wall can be thickened locally for a straight and tight fit.


• Recent testing has revealed that the latest prototypes lack a little in performance compared to the earlier prototypes. A re-spin of the board can remedy this and will bring you the best performance possible (note that one board design is currently still in the pipeline and is expected back later this week for more performance testing). This will delay things a little, but not too much (about a week).

As you can see, this project is keeping me busy full-time, but all aspects of the project are progressing very well in parallel.

Software compatibility with RoadMap

Link for iPhone and iTouch users here.


Before we discuss the video, let me first mention that we hit 2000 votes yesterday!

This is again great news and I will update you on the manufacturing progress in a separate post (I have been very busy taking care of all the logistics behind the scenes and acquiring quotes for the manufacturing now that NDAs have been signed).

The video above shows a demonstration of the open source software RoadMap. This open source software was ported to the iPhone by Morten Bek. Of course it is a work in progress and not everything is perfect yet and neither are all features ported, but at least it shows the concept and future potential: 3D real time location mapping (with local maps on your iPhone/iTouch; no need to connect to the internet) and hopefully in the (near) future full voice-assisted navigation support.

I was very impressed by what Morten has accomplished so far!

It seems there are two branches of the RoadMap software. The one that Morten is working on and the one that Ehud Shabtai is working on. The second branch is a little further ahead maybe in terms of features, see this video, but unfortunately the language and Wiki for that branch is in Hebrew and it only comes with the map of Israel. Hopefully the two branches will be able to join efforts and converge into one great iPhone/iTouch application.

For Morten's branch, I downloaded all the maps for California from here and I can now locate myself on my iPhone without the need for the Maps.app or internet connection.

The video shows the main aspects of the RoadMap application:

• Arrow icon at left top: this switches between a fixed map view or a rotating map view. In the last case, your driving direction is always pointed upward and the map will rotate along with you.


• Satellite icon: its color varies, green for GPS lock and red for no GPS signal. Pressing the icon will move the center of the map to your current GPS location.


• Satellite screen: provides a view of the signal to noise ratio (SNR) for each satellite, longitude, latitude, elevation, speed, etc. This screen is accessible through the left button on the menu bar in the bottom (hard to see in the video)


• 2D/3D icon: pressing this icon will switch between a 2D view of the map and a more navigation like 3D view of the map (see end of video)


• Zoom in/out icons: pressing these buttons will change the scale of the map accordingly.


• There are some more buttons in the bottom bar, but not all of them are working. With one of them, you can email your currently logged GPS route to your friends.
  

Unfortunately, my video recording ended abruptly as the camera ran out of memory, so the 3D view of the map (which I think is really neat) is not that long. But I hope you get an idea of what it looks like.

Warning.....

That time has come again..... Apple's 2.0 firmware upgrade will be out shortly, so I would like to issue a warning....

To make it clear to everyone, if you intend on using the GPS module on your iPhone or iTouch, make sure NOT TO UPGRADE your iPhone and iTouch firmware for the time being. The same goes for iTunes. Leave it alone till it is confirmed that it is safe to upgrade.

In order to communicate with the GPS module, a jailbroken iPhone or iTouch is required. Upgrading to Apple's official 2.0 firmware will put your iPhone/iTouch back into a sealed environment again until the iPhone Dev Teams comes up with another jailbreak.

So if you would like to keep your options open, play it safe. Leave your current firmware alone and wait till the unlock development teams give you the green light.

Simplified housing design

This week has been a very busy week. What you see above is a simplified housing design for which it will be cheaper to make an injection mold than for the first design. This is important as we are talking about relatively small number of volumes for injection molding (a couple of thousand) and the mold-making cost will need to be kept as low as possible so as not to drive up the price of the module. An alternative to traditional injection molding is rapid injection molding. I will need to look at both options to find out which one gives the cheapest price per unit.

For this latest housing design, the USB connector has been moved to the bottom side of the module. I have updated the PCB layout accordingly. The size of the housing is 26mm x 22.2mm x 9mm.

I have started working pretty much full-time on this project since July 1st. I have drafted up a project schedule for getting the module into production. Quite a few things need to be synchronized in terms of time-line. Here is what currently has been accomplished or is still in the works:

• NDA's have been put in place with various contract manufacturers (CM) for both the PCB assembly and the housing.
  
• The PCB has been designed and is waiting for feedback from the CM in order to determine manufacturability (1-2 weeks; we will be going into the second week). The PCB design might have to be tweaked after we get feedback from the CM. This process is called design for manufacturability (DFM).
  
• After the DFM is completed, a few final prototypes need to be built and sent to the CM (2 weeks).
  
• A few prototype housings also need to be built/machined and sent to the CM (3-4 weeks).
  
• GPS antennas need to be built (1 week setup + 3 week lead-time).
  
• GPS chip needs to be manufactured (4 weeks).
  
• Bill of materials (BOM) needs to be sourced. Some parts have a lead-time (up to 3 weeks).
• Housings need to be produced by injection molding at CM (no time-line available yet from the CM for this).
  
• After all this is completed, usually an engineering pilot build is done for design verification and after that a production pilot build.
• Test procedures need to be implemented and test fixtures designed and put in place (4 weeks). This includes both testing of the PCB as well as a final test of the performance of each GPS module.

As you can see, quite a lot of things need to be managed and synchronized. But in order to bring you the best working GPS product out there for your iPhone and iTouch, no corners can be cut.

I will keep you posted as things progress :).

First design of the housing

Top view:

Bottom view:

I have completed a first draft design of the housing. Here are a few comment on the housing:

1. Take a close look at the Apple design of your dock connector. I have managed to take that connector apart without destroying it. It's a pretty sleek design, consisting only of two parts: a sleeve that goes around the connector and a cap that slides over the connector and inside of the sleeve. I have made the housing for the GPS module in a similar rounded/flat shape as the Apple sleeve. Apple's sleeve is 26.1mm x 9mm x 5.6mm (width x length x thickness). The GPS module housing is 25.4mm x 22.2mm x 8mm. So it is a little thicker as the dock connector and about 2.5 times as long.


2. You can see the micro USB connector on the left side of the module (top view).This plug can either accept micro USB-A or micro USB-B cables. I have found a trust-worthy supplier for the micro USB-B cable and the cable will be included with the GPS module to make life easy :).


3. On the top view you see a small green light. This is the LED indicator that displays the status of the GPS module: flashing orange means the module is searching for a satellite lock. Once a lock is found, the indicator will light steady green. My plan is to make this indicator aperture similar to Apple's MagSafe power adapter that is used on the MacBook.

In the meantime, I am playing with some more ideas to make the housing design even simpler...

Board and chipset update

This is a brief note to give you an update on the board assembly that I started last week.

It turned out that the problem of the GPS chip not sending data was not in the GPS chip itself but caused by my own mistake (:-o). I was trying to get the module together quickly to test the new board, and in the hurry, I didn't solder one of the components onto the board, which left part of the GPS chip disconnected. After a very helpful phone conversation with the chip vendor, we traced the problem down in 5 minutes.

I guess sometimes it is indeed good to take a brief break from everything and let things come to peace; after I soldered the missing component onto the board, everything worked flawlessly :)

And I have another update for you: the next chip-sets that I ordered will be A-GPS enabled!

GPS and A-GPS

As a few questions were posted earlier regarding A-GPS (assisted GPS). I will try to outline the differences between regular GPS and A-GPS.

A GPS module is dependent on the GPS satellite signals to find a fix. The weaker this signal, the longer it will take to get a fix. If too weak, no fix will be found at all.

• When a GPS module is first turned on, it will usually have no satellite position data (ephemeris). This is what is called a cold start of the module (assuming the module does have a valid almanac stored). It is then necessary to place the module in open sky to have it acquire satellite data. This can take anywhere from a minute to several minutes, sometimes even longer, depending on the satellite signal strength, surroundings and the sensitivity of the GPS module. This sensitivity of the GPS module typically ranges from -140dBm to -146dBm for a cold start (the more negative the number, the more sensitive the module is).
  
• Once a GPS lock has been obtained, the GPS module will store this recent satellite data (ephemeris, almanac and time) in its memory and the module is said to be in tracking mode. When in tracking mode, the sensitivity increases as the module now has satellite data available. Typical tracking sensitivities are from -155dBm to -160dBm.
• If power is cycled off on the GPS module, then fairly immediately back on, the ephemeris data is still "fresh" (but the time might not be), and a GPS lock will be very quick since the GPS module does not have to collect new ephemeris data. This is called a warm start and the time to fix is typically less than half a minute.
• If the GPS module also still has an accurate time, the fix will be even faster (usually on the order of a second) and this is called a hot start.
  

In general, if ephemeris data is not available or outdated, obtaining a GPS lock can take some time. This is where A-GPS can help.

An A-GPS receiver can help getting ephemeris, almanac and time data from an assistance server. The assisting data can either be downloaded every time the GPS module is turned on (e.g. through the GSM network or WiFi) or in advance (e.g. once every 14 days at home from the internet). In the first case, ephemeris data, almanac and time are downloaded. In the latter case, differential almanac correction data is downloaded.

In order to be able to use A-GPS, we need three things:

1. A chip that is A-GPS compatible, i.e. the chipset allows for download of assisting data to the GPS module,
2. An assistance server from where the assisting data can be downloaded. Assistance servers are usually provided by GPS chipset manufacturers and need to be mirrored for client usage. (Apple is probably mirroring Broadcom's servers for use on the 3G iPhone).
   
3. A way to upload the A-GPS data to the GPS chipset. This can usually be done through a software download daemon which both acquires the data from the assistance server and sends it to the GPS chip at certain time intervals.

So A-GPS can make your GPS module find a faster lock, typically increasing the cold start sensitivity to about the level of a hot start.

It needs to be mentioned however that if a satellite signal is simply too weak (e.g. <-160dBm in parking structures or big concrete buildings), you won't see any improvement at all in the performance of your GPS module, even if you have an A-GPS enabled chipset as the GPS receiver won't be able to receive any signal even though you tell it were the satellites are.

Regarding the GPS chipset that I am using, the chipset can be enabled for A-GPS use and assistant data could be downloaded to the module through the serial port if the data can be obtained from an assistance server.

The Return of the Boards (part 3)

As you can see in the photo, the two boards that were sent out a while back have returned!

Also, before we discuss the boards, we hit 1500 votes this week as well (and that despite all the distracting iPhone 3G buzz) !

Regarding the boards, I started assembling one board last Thursday night but ran into problems testing the board. It took me two nights to debug the problem. The boards seem to be oke; the GPS chip I am using is the same as the one on the prototype II boards and it functions there without problems. I suspect the problem is in the GPS chip itself: although all output voltages seem oke, no data is transmitted. I had a spare unit which I tested off the board and this one has the same problem. I will have to contact the manufacturer to see what the problem is and to get functional replacement units.

Well, I guess things can't always go smoothly. I am sure we will recover quickly from this after I get new functional chips. I'll keep you updated once I find out more from the manufacturer. This shows that you always need to double check your suppliers, especially when you are using their latest technology. It also shows that it is a must that every single unit that gets shipped is quality controlled. In fact, I will guarantee that every GPS module that I will ship, will have been inspected and is 100% functional. Or as we say in dutch "De klant is Koning" (literally translated as "The customer is King" :).

In the meantime I have also been working on housing related things this week: I have pretty much completed the 3D modeling of the most important parts on the board (which will be needed to make sure that the housing doesn't interfere anywhere on the inside with the board/components). Also, I have already outlined the housing itself. Detailing of the inside will be next.

iPhone 3G buzz....

After having read some of the discussions on the previous post, I think it might be good to speak some sobering words and try to see things in perspective.

• The new iPhone is extremely cheap, even cheaper than an 8GB iTouch..... BUT you will have to sign a two year contract with AT&T (see here) for most likely a minimum of $39.99 a month.... or sign and then break the contract and pay the termination fees (probably around $200). So that puts the price tag right back at $399. My guess is that AT&T is subsidizing Apple those $200 for every phone sold.
• AT&T is still the exclusive U.S. carrier...
  
• Rumors are going around that you won't be able to buy an iPhone without signing a contract, either in the Apple Store or the AT&T store (see here)
• Regarding the iPhone 3G itself, how many of the current iPhone users would actually upgrade to the new model? Most likely those users who are already with AT&T and who are willing to spend the additional $199. I would guess that for most of the hacked phone users (who don't wan't to be with AT&T to start with), dishing out $199 + cancellation fee would be a little over the top.
• Steve mentioned that around 6 million of the original iPhones were sold. It was estimated that around 750,000-1 million phones were hacked at the end of Jan. 2008 (see here) and maybe it is even up to 20-25% now. How many of these phone users would want a plug-in GPS module? Even at less than 1%, it would be worth to make the effort.
  
• Note that even today after the announcement, iPhone users have been voting that they still would like a GPS module.
  
• Lastly, don't forget the iTouch users. The iTouch doesn't have GPS (at least for the time being), so who will take care of them?
  

So to make the story short, I will keep on doing what I have been: working hard to come out with a GPS module for iPhone and iTouch users at an acceptable price and as soon as possible.

PmgR

Compatible? YES!

Link for iPhone and iTouch users here.


Just recently, a GPS plug-in for the Maps.app was released by xWaves. The video above shows how my GPS module also works with that software.

The source code for the software was released as well and can be found here and checked out with regular svn. All credit for this software goes to the people at xWaves.

Below is some more information on how the software works:

• The software currently talks to the GPS module directly through the serial port. The standard baudrate is 19200 which I had to adjust to match my module. It will be best if the software is updated in the future to automatically check what baudrate the GPS module is running at. Another option is to add in functionality to work with a gps daemon.


• The xGPS program is a daemon that runs in the background and is started through a plist file in the Launchdaemons directory. This deamon talks to the Maps.app and the serial port and tells the Maps.app to update its location with the data obtained from the serial port.


• The software currently needs to connect to the internet to update the map as it calls the Google website to update its location. It would be nice if someone figures out how to circumvent this and have the Maps.app update its location directly from its database without having to connect to the internet. In that way you could upload your own maps to your iPhone/iTouch and use those while on the road.
  

Then finally one comment on the video. I stated: "...the GPS module is charged through the USB..." What I actually meant to state is that the iPhone is charged through the USB. If you look closely, you can see the charging sign in the video.

Then lastly, before signing off for tonight, I also wanted to let you know that I am currently working hard on the housing design and getting manufacturing for the boards set up.