I finished assembling the first revision of the tracking device.  Unfortunately there was a slight problem with the footprint for the BMA180 so the accelerometer portion of the board is non functional.  I've also come to the realization that I'm no longer going to use 402 SMD components for future designs. As much as I love the challenge of trying to place these tiny little parts on a board I think I'm going to move to 603 and up from now on.

So how did the rest of the tests go? Fairly well I'd have to say.  Initially I was a little concerned about possible interference on the GPS because of the close proximity to the GSM module and high frequency buck converter for the power supply, but the impact seems to be minimal if any.

 

The charging circuit was designed around an MCP73833 single cell Lipo charger configured to 500mA charge rate supplied through a a mini USB port. There are also three on board LED's to show the charging state.  

 

For the 3.3 volt power supply I decided to go with a Fairchild FAN5362 3MHz buck regulator.  It's capable of sourcing 500 mA which is plenty of current for powering any other 3.3 volt external devices that would be connected to the board. Another benefit to using the buck regulator over a standard LDO is that the buck has under voltage lockout protection (UVLO).  Basically the regulator will shutdown if the battery voltage falls below 2.5 volts.  It's a nice feature to have when you're powering something with a lipo battery.

 

As I mentioned earlier I had a problem with the accelerometer footprint so I wasn't able to test much of the sleep mode options with the board.  I wanted to see how low I could get the current draw in sleep mode as compared to the average 50 mA it draws while fully powered up. I guess I'll have to wait for the next revision. For the next revision I'm planning on switching from a BMA180 to a BMA250.  It's a comparable accelerometer with a little lower resolution but features built in algorithms for orientation detection. 

 

Another option I'm adding on the next revision is a connection for an external battery backup for the GPS. Currently the battery backup pin is wired directly to the Lipo battery which, as long as the battery is connected, does provide backup power.  Unfortunately if I disconnect the battery the GPS resets and I have to wait for the ephemeris data to download again which takes about 12 minutes.

 

I'm sending out for the next rev board this week so stay tuned...

 

Comments   

# Bob 2011-12-19 00:42
Wayne,

I agree w/ you about 0402 parts - I can do them but 0603 is my sweet spot too. W/ 0402 I find I build completely under the microscope @ 10x which is fine for 1 offs, but if I want to make a few, too time consuming, 0603 is no problem w/out it and I just have a final look w/ the microscope before reflow for some possible little alignments. I would like to do 0201 just to say 'I did it', but time, well, where does it go?

I would like to hear your reviews of the BMA250. I fly multirotors and currently use the BMA180 but have a bob, some BMA250s and a stencil here I just haven't made yet due to time constraints. I have talked to Bosch US and was informed that they are also going to be bringing out 12 & 14bit units soon that are pin compatible w/ the BMA250, which I think is pin compatible w/ the BMA222? (will need to look @ DS for that). Did not get an exact timeframe, but it was not that far down the road (hopefully not an Invensense MPU6050 distro.... :-x ).

Have a good one,
Bob
+1 # Wayne Truchsess 2011-12-19 13:24
Hey Bob,
I just started playing around with the BMA250 yesterday. I'm mostly playing with the orientation detection features to see the best way to tie it into the tracker. Looks like an I2C device manufacturer finely wised up to bus lockups also, they implemented a configurable watchdog timer to reset the bus if the accelerometer causes a bus lockup.
# jegabose 2012-01-26 14:13
Wayne, any updates of rev-2?
# Wayne Truchsess 2012-01-26 14:50
Quoting jegabose:
Wayne, any updates of rev-2?
Yup. Looking good so far. Had to cut a couple of traces for the accelerometer and add a jumper wire but other than that everything's looking good. I haven't had a chance to play around with some of the more advanced features of the GPS yet but I'm hoping to do that today. I need to get a better idea of power consumption with the GPS. How did your testing go with the first rev? Any particular problems jump out at you? Feedback (good, bad or indifferent) is always welcome.
# jegabose 2012-01-26 15:54
Cool, I am looking forward to grab one ASAP. Well, I should say that I am very much pleased with the performance of rev-1. I tried at least 6 other GSM/GPS shields with SIM500, SIM900, Telit GE86xx, etc. Nothing comes close with your implementation. Very low power consumption, turn on/off is straight forward, decent TCP/IP stack. I would strongly recommend to add ability to turn on/off GPS too, I understand that the next GPS engine has very low power mode, but still I think its a nice idea for future version. One of my requirement is to keep the unit live for atleast 365 days without charging, looks like I can push upto 200 days with a 4400 mah lipoly reporting 1 TCP/UDP message a day with deep sleep mode. It looks like truly a industrial range circuitry, so far I tried with 0F through 155F and it worked great. Great piece of product, keep up the good work.
# Wayne Truchsess 2012-01-26 16:36
Quoting jegabose:
Cool, I am looking forward to grab one ASAP. Well, I should say that I am very much pleased with the performance of rev-1. I tried at least 6 other GSM/GPS shields with SIM500, SIM900, Telit GE86xx, etc. Nothing comes close with your implementation. Very low power consumption, turn on/off is straight forward, decent TCP/IP stack. I would strongly recommend to add ability to turn on/off GPS too, I understand that the next GPS engine has very low power mode, but still I think its a nice idea for future version. One of my requirement is to keep the unit live for atleast 365 days without charging, looks like I can push upto 200 days with a 4400 mah lipoly reporting 1 TCP/UDP message a day with deep sleep mode. It looks like truly a industrial range circuitry, so far I tried with 0F through 155F and it worked great. Great piece of product, keep up the good work.

I appreciate the feedback...than ks. I have a question, was your 200 calculation based off deep sleep on the Atmega and GSM only? Were you using the current consumption of the GPS? I ask because I just finished a test using the standby mode on the GPS and current consumption goes from 20-25mA (while tracking) down to 200 microamps in standby. It's done through software and wake up is done byte simply send a byte to the GPS.
# jegabose 2012-01-26 18:03
Quoting Wayne Truchsess:
Quoting jegabose:
Cool, I am looking forward to grab one ASAP. Well, I should say that I am very much pleased with the performance of rev-1. I tried at least 6 other GSM/GPS shields with SIM500, SIM900, Telit GE86xx, etc. Nothing comes close with your implementation. Very low power consumption, turn on/off is straight forward, decent TCP/IP stack. I would strongly recommend to add ability to turn on/off GPS too, I understand that the next GPS engine has very low power mode, but still I think its a nice idea for future version. One of my requirement is to keep the unit live for atleast 365 days without charging, looks like I can push upto 200 days with a 4400 mah lipoly reporting 1 TCP/UDP message a day with deep sleep mode. It looks like truly a industrial range circuitry, so far I tried with 0F through 155F and it worked great. Great piece of product, keep up the good work.

I appreciate the feedback...thanks. I have a question, was your 200 calculation based off deep sleep on the Atmega and GSM only? Were you using the current consumption of the GPS? I ask because I just finished a test using the standby mode on the GPS and current consumption goes from 20-25mA (while tracking) down to 200 microamps in standby. It's done through software and wake up is done byte simply send a byte to the GPS.


Yes, you are right, the calculation is based on Atmega and SIM900 only.
# Sulphar 2013-06-19 07:21
How to burn the bootloader into the Atmega328 Chip on the Tracker, if anyone know plz tell me

Login

Go to top