News:

Attention: For security reasons,please choose a user name *different* from your login name.
Also make sure to choose a secure password and change it regularly.

Main Menu

Weird baud rate

Started by mscaff, November 12, 2010, 12:16:54 AM

Previous topic - Next topic

mscaff

Hi,

I just received a few modules for evaluation.
I soldered one of them in my board and just started to watch it's output in minicom.
It's was all ok, when, after a power cycle, i was just receiving garbage.
After a few tests, i discovered with an oscilloscope that the original 38400 baud rate now is 24400bps, which is a completely weird baud rate. Even turning the module off, and waiting for a long time didn't help...
Does someone seen this before?
How can I change the baudrate of this module? Is that change supposed to be volatile or not?
If it is a command that i have to send to change the baud rate, how can i guess which is my initial baud rate?
My second module is working fine at 38400 so far.

Thanks, Mauricio

Horsa

Quote from: mscaff on November 12, 2010, 12:16:54 AM
I just received a few modules for evaluation.

Please tell us what modules and what firmware version(s). One way of getting the firmware version is through the MiniGPS tool.

Quote from: mscaff on November 12, 2010, 12:16:54 AM
Even turning the module off, and waiting for a long time didn't help...

Did you disconnect physically every pins? VBackUp also?

mscaff

Hi,

The module is a GTPA010,
I do not have the firmware version right now, I'll try to get it in the weekend.

And, the module was physically disconnected from the board and after that I even tried to short circuit the modules power supply pins to discharge any internal capacitors.

Yet, no luck.

Thanks, Mauricio
 

Horsa

Quote from: mscaff on November 12, 2010, 05:46:35 PM
The module is a GTPA010,
GTPA010 is the name of the integrated patch antenna, not of the module:
  http://forum.trenz-electronic.de/index.php/topic,51.msg71.html#msg71
Please have a look at the bottom side, not at the top one.

mscaff

Under the module are those strings:

G.top
9803-PA6B-V01
2009/06
MV1194V-0-138-0

Thanks.

Horsa

Please now tell us what firmware version(s) have all the modules with the same code on the back as the one with the weird baud rate.

Horsa

#6
Quote from: mscaff on November 12, 2010, 05:46:35 PM
I do not have the firmware version right now, I'll try to get it in the weekend.

Dear Mauricios,
as you did not provide us firmware version(s), I will try to provide some general answers to this issue.


  • the 24400 baud rate does not belong to the MediaTek baud rate specification
  • please disconnect the VBackUp pin and power off / wait / power on the GPS module (by the way, you already did this)
  • connected the VBackUp pin, send the MediaTek command
      $PMTK251, 38400*27
    to GPS module and power off / wait / power on the GPS module
  • send the MediaTek packet
      605 PMTK_Q_RELEASE
    to module to get the firmware version (by the way, this will not solve your issue, but will help your technical support representative)

If nothing of the above items helps and you are not happy with this behavior, please contact your GlobalTop supplier.
Without any further information, I consider this post closed.

mscaff

Sorry about the long time to reply to this topic.

On Startup the module sends the following:

$PGMOD,2*4F
$PMTK011,MTKGPS*08
$PMTK010,001*2E

And as reply to the PMTK_Q_RELEASE command:

$PMTK705,AXN_1.30,2227,,1.0*0E

Thanks, again, Mauricio



mscaff

I forgot to tell,

Sending the $PMTK251,38400*27 command does nothing...
The module still @ 24400 bps :(


Horsa

Dear Mauricios,

we reviewed your issue carefully. It can be basically two things:
- firmware issue
- hardware issue

Firmware issue: we have ready exactly the same firmware version as you have. We propose you to try flashing your unit using this exact same firmware. If you experience troubles getting the firmware to flash, try powering on the GPS module just before pressing the "download" button in the firmware flash program.

Hardware issue: our initial assessment is that the TCXO or the crystal inside the module has probably malfunctioned, leading to the incorrect timing of the UART port; this may be due to either a faulty module or possible problems with the power source


Recommended support strategy for your case:

  • email us to get the recovery firmware and the flashing tool
  • flash your module with the recovery tool
  • test you rmodule again
  • if the problem occurs again, we assume it is a hardware problem and please contact us to open an RMA

mscaff

#10
Hi,

How can I get the firmware and download tool? (my email is mscaff...).

It's hard to believe this is a HW problem, since (i'm guessing here) it's very unlikely that even with such a huge deviation in the oscillator frequency the module still can receive satellite information.

Thanks, Mauricio

jolcese

Hello,
I have the exact same problem: Worked OK couple of times at 38400 and now it's running at 24400....
Did you solve this issue?

Thanks
Jose

Horsa

Hi Jose,
as of now, we do not know whether Mauricio could solve by upgrading the firmware.

In the meanwhile,

Regards

jolcese

Hi Horsa,
I didn't buy the module from your shop so I assume you don't care about the serial.. :(

Thanks anyway
Jose

Horsa

Hi Jose,

thank you for telling us that you did not buy the module in our shop, even though this is not the answer to our technical support question.
Your assumption about we not caring about the serial number of your module experiencing a weird baud rate is wrong. The reason behind that is quality control and fault tracking. Your answers to our two questions can help us helping you and the community.

Regards,
- Horsa

jolcese

Thanks Horsa. I'll send the SN information.
Jose

Horsa

#16
Posting the following information would also be very useful for further investigations:

  • a screen shot of the module running at 24400 bps
  • data log (using mini GPS tool) of the module running at 24400 bps
  • Which method was used to power cycle the module? Was it through enable pin or by cutting the power to VCC?
  • schematics to the main-board connected to GlobalTop GPS module
  • a picture of the main-board
  • check that the Vcc ripple is controlled under 50 mVpp

- Horsa

mscaff

Well,

Almost 2 year later.
I found those GPS modules inside a drawer and decided to play with then again, since i'll need a gps in another project.
Well....
Still (even after 2 years disconnected) running at 24400bps.
So, I decided that it is a faulty module. replacing with the other module that i have (same batch, never used).
It is too at 24400bps. I just turned on my hardware and it started to receive nmea messages at 24400bps.
So, I don't think it's hardware.
Being a firmware problem, does anyone managed to solve it? I've tried to send messages to the module but it seems to ignore all of them no matter which baud rate.

Thanks, Mauricio

mscaff

I found a solution for the problem.
If you power up those modules with the 3D-fix output connected to a LED attached to VCC, instead of GND (I wanted the LED On when there is a Fix) the module will start running at 24400bps.
No idea why this happens, but it does. After changing the LED from VCC to GND the module started to output messages @ 38400bps. Now I'll try to send packets to configure it.
Hope it can help other people too.
Mauricio


Horsa

#19
Thank you for your perseverance. We eventually got it.

This issue is described in "GPS Module Application Notes".
http://www.trenz-electronic.de/download/d0/GlobalTop_Technology/d1/application_notes.html

The 3D-Fix pin can affect the internal crystal oscillator (not the TCXO). If the 3D-Fix pin is connected directly to Vcc like you did, the MediaTek chip sets the internal crystal oscillator frequency to 20 MHz instead of the regular 32.768 MHz. This is why your baud rate changed to 24400 b/s instead of your expected 38400 b/s:

  • 24400 / 20000  ~=  1,2
  • 38400 / 32768  ~=  1,1