gps.time() rolls back to zero after mqtt.publish()

[sebi5361]

I have noticed that gps.time() rolls back to its original value 2525040000 after executing mqtt.publish(topic, message). I believe the GPS resets completely in that situation for an unknown reason.

[pulkin]

Maybe power?

[sebi5361]

I am not that sure it comes from my power source. Could you try that code?

import cellular
import gps
from umqtt.simple import MQTTClient
from time import sleep

cellular.gprs("tm", "", "") # Adjust with your APN

gps.off()
gps.on()
gps.time() # 2525040000
sleep(4)
gps.time() # 2525040004

mqtt = MQTTClient("A9G", "test.mosquitto.org")
mqtt.connect()
gps.time() # >2525040004
sleep(4)
gps.time() # Rolls back to 2525040000
mqtt.disconnect()

[sebi5361]

Did you have time to try this simple code and confirm the bug?

[sebi5361]

I confirm that my GPS module resets (gps.get_satellites() == (0, 0) and gps.time() == 2525040000 and gps.get_location() == gps.get_last_location()) after issuing mqtt.connect() or mqtt.publish(...).

...thus making it impossible to send the actual GPS location by MQTT at reasonable rate, as one has to wait for the GPS module to start over each time.

I think this is not related to mqtt.connect() and mqtt.publish(...) only: once my GPS got reset after issuing from umqtt.simple import MQTTClient.

[sebi5361]

Dear Pulkin, Did you experience the same issue too? Are you planing on solving this bug soon? Thanks,

[pulkin]

Yes, I will investigate it but it is not at high priority right now. In the meantime I suggest you to spy the GPS module by reading from its UART2 which should be possible at this point. The only problem I can think of right now is a power issue: mqtt is a pure-python module which is not even aware of this port capable doing GPS.

[sebi5361]

Indeed I understand there is no reason why MQTT would be responsible for the GPS to reset. I will spy UART2 to see if I can find additional information, and test with a lab bench power supply to see if the problem persist. In the meantime if you could just try on your board if you encounter such a bug, it would help me to decide if I need to buy an additional board~~

[sebi5361]

With a lab bench power supply my GPS still resets... I really wonder if it is for my board only or for all the boards?? Do you know the baud rate and the settings of UART2 to spy it?

[pulkin]

It should be the default 115200 (or maybe 9600). You first open the port and then turn on the gps.

[sebi5361]

The correct baud rate is 9600.

I have tried using VUSB with 5V and VBAT with 4V, with I>2A but it didn't change anything: I am getting gibberish on UART2 TX regularly. Nothing special is emitted on the RX line when the gibberish occurs as you expected.

I think gibberish occurs with every GPRS emissions. Maybe it has to do with interferences. I hope not all the boards are like mine...

Here is an example of the NMEA output I get (I have added blank lines and a return line after the gibberish for better lisibility):

...
$GNGGA,000009.029,4337.2350,N,00350.5428,E,0,0,,103.1,M,49.7,M,,*5C
$GPGSA,A,1,,,,,,,,,,,,,,,*1E
$BDGSA,A,1,,,,,,,,,,,,,,,*0F
$GPGSV,1,1,00*79
$BDGSV,1,1,00*68
$GNRMC,000009.029,V,4337.2350,N,00350.5428,E,0.000,0.00,060180,,,N*5F
$GNVTG,0.00,T,,M,0.000,N,0.000,K,N*2C

$GNGGA,000010.029,4337.2350,N,00350.5428,E,0,0,,103.1,M,49.7,M,,*54
$GPGSA,A,1,,,,,,,,,,,,,,,*1E
$BDGSA,A,1,,,,,,,,,,,,,,,*0F
$GPGSV,1,1,00*79
$BDGSV,1,1,00*68
$GNRMC,00001ðÿþÿÿþÿÿ

$GNGGA,000000.029,4337.2350,N,00350.5428,E,0,0,,103.1,M,49.7,M,,*55
$GPGSA,A,1,,,,,,,,,,,,,,,*1E
$BDGSA,A,1,,,,,,,,,,,,,,,*0F
$GPGSV,1,1,00*79
$BDGSV,1,1,00*68
$GNRMC,000000.029,V,4337.2350,N,00350.5428,E,0.000,0.00,060180,,,N*56
$GNVTG,0.00,T,,M,0.000,N,0.000,K,N*2C

$GNGGA,000001.029,4337.2350,N,00350.5428,E,0,0,,103.1,M,49.7,M,,*54
$GPGSA,A,1,,,,,,,,,,,,,,,*1E
$BDGSA,A,1,,,,,,,,,,,,,,,*0F
$GPGSV,1,1,00*79
$BDGSV,1,1,00*68
$GNRMC,000001.029,V,4337.2350,N,00350.5428,E,0.000,0.00,060180,,,N*57
$GNVTG,0.00,T,,M,0.000,N,0.000,K,N*2C
...

If you confirm this issue doesn't occur on your board I will buy a new one.

[sebi5361]

Good news for me: by placing the GPS antenna as far as possible from the GPRS antenna, the GPS module doesn't reset. Maybe my GPS antenna has a weak electromagnetic protection. Hopefully other boards won't encounter this issue.

[pulkin]

I thought about this. For me shit mostly happens when I shortcut GSM antenna and one of the nearby pins.