Smartphone GPS Tracker Version 1.0

[dickmoran]

Derek, what is your plan for this project? Will it be completely assembled ready to fly, kit of parts or just pcb? Do you have any idea of price whichever way you go?

Thanks for the work so far,

Dick

 

[Derek]

see post #1.

 

[Reinhard]

Here it is:

View attachment 121631

On top is the finished transmitter, middle is the top of the pcb and the bottom is the bottom of the pcb.

One observation, it looks like you are using a linear regulator in a SOT-23 package. In a somewhat similar device (NL-552ETTL GPS, XBee-Pro 868, AVR and a couple of sensors), that I'm involved, we had experienced problems with a regulator overheating in warm environments when running from fully charged 2s LiPo batteries (~8.4V).

We had a rather bad design (a SOT-23 regulator and a SOT-223 regulator, both crammed into the same corner of the PCB, no thermal vias). Under most conditions the device worked fine, but on a hot sunny day, and while enclosed in a nose cone, it overheated to the point that one of the regulators shut of. If I remember correctly, we measured more than 100°C on the PCB.

I'm not sure how comparable our designs are. Yours looks like it should draw significantly less current (less parts, lower powered transceiver), so it will probably be without problems. It won't hurt stress testing it though.

Reinhard

 

[mikec]

we had experienced problems with a regulator overheating in warm environments when running from fully charged 2s LiPo batteries (~8.4V).

This is a good reason to always run off 1S batteries; the regulator being used has low enough dropout voltage to do this.

 

[Derek]

yep.

 

[Reinhard]

I agree. In our case, some 5V components prevented us from going this route. This was particularly bad, because most of our current draw happened at 3.3V (up to 800mA, just from the XBee-Pro 868 alone).

Considering that 9V batteries are quit common in rocketry, I figured it might be useful to point this out.

Reinhard

 

[Derek]

the regulators I used won't run on 9v batteries. 5.5v is the max.

 

[cerving]

You might want to think about replacing the regulator with something that will run on 9V. If you're pulling nearly 1A, you probably need something in a TO-220 package, a ST Microelectronics LD1117AV33 would do; it will handle 1.2A with a dropout of 1.15V max, and you can feed it up to 15V.

 

[Derek]

the whole system is designed to run on a 1s lipo, so there is no need to redesign it. the parts you suggested are way over sized (physically and electrically) for this application. they won't even fit on the board.

 

[sti_ffy]

Here is an idea that I haven't seen discussed yet. If you aren't squeamish about soldering, you can remove all the pins from the XBEE and the two sockets from the XBEE adapter, and then install the adapter on top of the XBEE.

In the photos below, I did this with my base unit - which consists of a SainSmart XBEE USB and an XBEE 900HP DM RPSMA module. This makes for a very compact arrangement with no potential for contact intermittence. The SainSmart board edge has been notched to clear the antenna connector and I used a temporary spacer to ensure that it cleared the I/O pins along that edge. Just to be sure, I also cut the track to the +5V pad on that edge. You don't have to put *all* of the pins back in like I did. You only need 1, 2, 3, 6, 10, 11, 15 and 20 - and 11 and 20 are only needed for rigidity.

Now the search is on for a box that will fit this and allow the usb connector to peek through one end and the RPSMA to be secured to the other with the washer and nut. Finding this will be the difficult part. :smile:




I found a case that was almost small enough. I had to hack it up a bit.

 

[dickmoran]

I flew my package yesterday and was happy with the performance of the GPS but not of my Formula54. I launched at a Rec Park near Mt Airy Maryland with my NAR Club [NARHAMS] with a G limit on engine size. I bought some CCI reloads just for this reason and was in a hurry to fly. I was going to redesign my av bay for the GPS package but with work and life getting in the way I ended up deciding to use my new untested Raven3 as a passenger and using the engine ejection to deploy the chute. You know you learn from your mistakes and yesterday was school for me. The rocket is heavy but with an H or I engine I get 2-4K. I have flown on a G a few times but used my 1st Raven 3 to fire the charges.
Things learned yesterday:
1, Never use stock motor ejection [15sec] on a 3.5lb rocket when the total flight time is 16 sec. [ A special tool is available to cut the delay time.]
2, Always pack a shovel and pick when your time is up.

We watched the launch and followed the flight as it arched over and headed straight into the ground but about 1500' away. My heart sunk as I thought I had lost my rocket, Raven3, GPS and my HD GearCam video camera.
Good news. The GPS tracker worked very well. I had it record the downloaded info just in case but the display froze when the signal stopped. The Lat/Lon was loaded into the smartphone and off went sti_ffy, his dog and me. We headed in the general direction but as sti_ffy read off the distance to go I see drainage ponds, small creek, concrete drains and other less than safe landing areas. As we passed these areas we ended up going up around a large area on a hill with sticker bushes in 2 groups and ended up at the top of the hill. I fired up my smartphone and walked back down through the bushes and with 90’ to go I spotted my rocket. It was sticking up in the ground with about ½ buried. Make a long long story short; once I bought a shovel and pick I was able to get my rocket out. As soon as I pulled out the av bay the Raven3 beeped the altitude. It was alive! Got home and cut the side of the av bay out and found the GPS had shifted and the battery cable broke. Put a new cable on and it is working! All I lost was a few batteries and my av bay case. The nose cone has some cool looking scars from the rocks it displaced on its 1.5’ journey toward the center of the earth. The 90’ off set was probably because the unit sends out it’s NEMA string once a second. So I guess it traveled 90’ in that last second. The video camera was still recording and I will upload to YouTube once I edit it down a little.

Dick
NAR 6306L1
Tri 14074L1

 

[CarVac]

You do not need the tool to drill delays; a regular drill bit works. For CTI, you remove 1/16" per second, if I remember correctly. Consult the instructions to confirm the 1/16" number before actually doing it though.

 

[bobkrech]

You do not need the tool to drill delays; a regular drill bit works. For CTI, you remove 1/16" per second, if I remember correctly. Consult the instructions to confirm the 1/16" number before actually doing it though.

You need a CTI delay reducing tool. This is the only certified way you can reduce the delay, and that's what the instruction say.

Bob

 

[cerving]

10-4, the suggestion was based on a previous comment that the XBee used about 800 ma. 1S LiPo's are the bomb if you can use them, light and cheap.

the whole system is designed to run on a 1s lipo, so there is no need to redesign it. the parts you suggested are way over sized (physically and electrically) for this application. they won't even fit on the board.

 

[Derek]

10-4, the suggestion was based on a previous comment that the XBee used about 800 ma.

that isn't the xbee module I'm using. those are only legal for use in europe.

my us spec xbee draws 215 ma max.

 

[Derek]

Dick,

Glad to hear you got your rocket back mostly intact!

were you using my transmitter board or something you built?

 

[dickmoran]

I used yours Derek. The only damage I see is the ground flange on the bottom of the UP501 shield lifted. I think the trace came up but I will super glue it back down. The GPS system was mounted to the av bay rear bulkhead using the antenna connector for the XBee. I had already made a big hole for another antenna I gave but ended up using a direct connect rubber ducky 900mhz antenna from sti_ffy so I made a plate to go on the inside of the av bay with a smaller hole and glued it to the bulkhead. This ended up coming loose during the below ground portion of the flight. The UP501 came in contact with the green terminal strip on the back of the Raven3, bending up the terminal strip at a 45 degree angle and dislodging the UP501 ground shield tab. All in all I am VERY happy with the results of my flight. I need to find some Google earth way to import my NEMA data so I can see the actual flight on sat data.

Your design worked great and would have kept sending signals except for the loss of power.

Thanks,
Dick

https://www.youtube.com/watch?v=qe7eG_vYhvA

 

[MarkH]

Derek,

If you make another run of transmitter boards, I'd like to buy one or two. I've flown the Pro XSC probably about 5-6 times, but used the Sparkfun regulator board. Your board is much more efficient packaging wise. Shown here is the transmitter and battery screwed to a FG sled, with the GPS receiver mounted on the other side. The masking tape is there to keep the Xbee from coming out of the socket. Mike C has a similar set up that had a pretty good write up on Rocketry Planet.

I've flown it to about 10,000 feet with good results other than the GPS (LS20031/ Mediatek 3329) receiver losing lock easily and not regaining for a minute or so.

Great work btw.

 

[mikec]

Mike C has a similar set up that had a pretty good write up on Rocketry Planet.

Thanks for the mention. Derek's PCB is definitely the way to go; attached is a picture of the most compact GPS I was able to assemble without using a custom PCB. It works well but I do have some residual concerns about the GPS module's performance; the one Derek chose seems like the best of the Mediatek options. I've been looking at some modules that are smaller, but they often require a ground plane that ends up using just as much, if not more, board area than the UP501.

 

[Derek]

Derek,

If you make another run of transmitter boards, I'd like to buy one or two. I've flown the Pro XSC probably about 5-6 times, but used the Sparkfun regulator board. Your board is much more efficient packaging wise. Shown here is the transmitter and battery screwed to a FG sled, with the GPS receiver mounted on the other side. The masking tape is there to keep the Xbee from coming out of the socket. Mike C has a similar set up that had a pretty good write up on Rocketry Planet.

I've flown it to about 10,000 feet with good results other than the GPS (LS20031/ Mediatek 3329) receiver losing lock easily and not regaining for a minute or so.

Great work btw.

very nice Mark!

I have one transmitter pcb left. PM if you are interested.

 

[Derek]

Dick,

That was a clever way of mounting it. I wouldn't have thought about that.

Looks like you did a nice job of building it. Great work!

 

[accpool]

Derek, any luck on that list?
Lance

 

[Derek]

Hi Lance,

I just saw your email. it got lost in my inbox. my fault. sorry!

Derek

 

[accpool]

your paycheck will be withheld. :dark:

 

[Derek]

fair enough!

 

[accpool]

just kidding you complete pay is in the mail.

 

[sti_ffy]

Dick, That was a clever way of mounting it. I wouldn't have thought about that.

Yes, mounting via the RPSMA connector is a viable solution, even moreso if you construct the tracker within the footprint of the XBEE itself.

This is my third protoype, and I think the final PCB arrangement would be through-hole for installing on top of the XBEE (solder to the tails of the pins) with a thin mica or similar insulator between them. One would then cut off the XBEE pins or not - based on personal preference. I removed them entirely, and only reinstalled those I wanted to provide rigidity, power and signals.

The GPS module can be spaced off the board slightly using either a spacer or something like servo tape, to prevent shorting the XBEE pins. It would be connected to the ciircuit board with SMT pads/header on the top side of the PCB.

In my prototype shown below, I used an XBEE breakout board with a very thin mica spacer soldered on top of the XBEE module. I removed the LEDs and such and notched that edge of the board to clear the RPSMA connector. The GPS unit is sitting right on top of the regulator, and is soldered to the Arduino pins - which were cut away from any connection to the XBEE pins, and then wired as needed. I placed a pin in the XBEE position 14 (which can be grounded) and soldered this to the GPS module can to secure that side of the module. In the production PCB the regulator would be positioned below the GPS unit but still within the footprint of the XBEE or possibly just beyond it.

This is as compact as I can make this without using a custom PCB, and it uses the RPSMA mounting method. Servo tape can be used between the metal shield on the underside of the XBEE and the adjacent mounting surface to provide additional rigidity and absorb vibration. It works very well indeed. I'd like for the production version to use a battery similar to what is used for the Raven altimeter. It may be possible to keep the entire package very nearly within the XBEE footprint, and it could just be bubble-wrapped instead of mounted if preferred.

 

[accpool]

very nice.

 

[Derek]

looking good, sti_ffy!

removing those pins or soldering other pins to them could be a pain. I had an idea I still may pursue where I cut off most of the right side pins on the xbee module beacuse those aren't really needed. this should free up enough room to mount the gps on the other side of a mating pcb. that is a lot easier (and should be stronger) than trying to piggyback one module on to another. the result will be a little fatter than your solution, though. those are the tradeoffs...

 

[Derek]

yeah! looks great. I see you went with the standard led colors.