problems with GPS-PA6B module

[koumpakias]

Hello,

I have been trying to get information out of a GPS-PA6B module.
I set the baud rate to 4800 bps, but I receive "garbage" on the serial port.
The "garbage" seems to be repeating itself at the rate that the module is
expected to send data (about once every second).
I have tried using other baud rates at the serial port setting with no success.

Also, the 3D fix pin keeps pulsing, indicating that no fix is achieved.

Has anyone else had this problem?
How did you solve it?

Thank you!

[kaddrison]

How have you connected it - to a pc? its TTL level out which is inverted to serial level - do you have a MAX232 type IC convertor?

[koumpakias]

I have tried in three different ways:
via a uUSB-MB5,
via a uUSB-CE5,
and via a MAX3232CPE.

I got the same response in all three cases.

I must point out that I did get correct responses (proper $GP... text) in one
occasion, for about a quarter. Then, the module reverted to sending "garbage"
again.

[ESPsupport]

This module (and all GPS modules) require VERY stable power to work properly. If you can measure the ripple ensure it is below 50mv, as per the datasheet.



In 'Open sky' (i.e. outdoors a fair distance from any buildings) it should lock withing 35 seconds or so as per the datasheet.



I'm thinking that your ripple is so high that it is affecting the readability of the serial data.



Do you have a circuit of how you have it connected?

[koumpakias]

I have used a Devboard G1. I used the 3.3V supply found at the top of the
breadboard and then I routed the module's rx/tx pins to the USB rx/tx pins.

Originally, I used the USB supply, but then I also tried an external power supply
and a 9V battery connected to the battery connection pads of the board.

The power supply is noisy in all cases. Do you suggest that I should use a
dedicated 3.3V regulator for the GPS module?

[ESPsupport]

Hmm, I'd try the Devboard.



Make the power wires as short as possible and put a couple of hundred mf and a 1nf across it as close as possible to the PA6B.



No where near perfect, but worth a try. Note that the 'sample' circuit uses a ferite bead..

[koumpakias]

I have rebuilt the circuit on a veroboard. Placed an LM2936Z-3.3 regulator to power the gps module and added a coil at the positive supply input. Noise on the supply connections of the module is about 10mV. I got the same results: garbage on the serial port and blinking LED, indicating a no fix.
I then changed the module with another one and I got the same behaviour.
Are there any suggestions on what might be going wrong?
Thank you.

[ESPsupport]

IS it possible your PA6Bs are somehow utilizing a different baud rate?



Try 2400, 9600 and a few other common rates

[koumpakias]

I have tried 1200, 2400, 4800, 9600, 19200, 38400, 57600, 115200.
It is still the same situation.

[ESPsupport]

Can you post some of the rubbish you get at 4800?

[koumpakias]

I have recorded the incoming byte stream into the attached message. Although it is a binary file, I had to rename it into txt in order to attach it.
Attached files gpstrial.txt (2.4 KB)

[ESPsupport]

So, is each 'bunch' of gibberish coming at 1 second intervals?



I presume it is your terminal program that is adding the CRLF pair every 103 bytes?



Anyway, can you try 7200 baud, please? (i.e.1.5 times the speed of that log)

[koumpakias]

This is what comes out at 7200.
Attached files gpstrial_7200.txt (4.5 KB)

[koumpakias]

I forgot to add that the packets are sent every second.
Indeed, the CRLF is added during the saving of the data
onto the file.
The initial log was taken at 4800.
To keep up with your request to sample the stream at
five times the original speed, I set up the terminal program
at the odd speed of 24000. I attach the log of that stream
too.


Attached files gpstrial_24000.txt (2.4 KB)

[ESPsupport]

Sorry I meant 1 and a 1/2 times, we use the . as the decimal point.



Funny how 24000 and 7200 look the same, is it possible the same baud rate ended up getting used for both?



I think I made a mistake earlier Can you do the test at 3200 baud? (now there's an odd rate)