After undertaken extensive testing on the Pytrack/Fipy combination we have found that the voltage required has to be around 4.5 volts or above to be stable. This then makes battery selection very difficult when we require a small unit with a long battery life.
For small batteries with a high Ah rating the voltage will be 3.6V. There are almost no small batteries within the 4.5 to 6 Volt range that have a long life.
Pycom give very little information regarding the voltage requirements on their data sheet for the Pytrack, and it appears rather strange that the Pytrack/Fipy combination can not run on a standard 3.6 Volt battery.
We are not able to us a voltage regulator as the continuous current consumption would be around 1mA where we are trying to keep the current down to about 20µA when the Pytrack is in sleep mode.
As we intend to run this in a remote location for long periods of time it has become a problem.
If anyone has a good solution for this it would be very much appreciated.
Thanks Clinton. We did have a look at the Power Boost 1000 Charger but at the best this would double the power consumption (20µA) during the sleep mode of the Pytrack.
We have seen information that suggests the Pysense has the ability to run on 3.6 Volts so we are now testing this. The Pysense does not have an on board GPS which we were thinking of using later in the project.
Okay, what setup are you using in order to power and interact with the devices? Also, what’re the exact issues that are occurring with your particular devices?
I am using a 3.7V LiPo battery to power the devices. It is connected directly to the pytrack’s battery connector (JST).
I am using the pytrack’s deep sleep functionality, so it cuts power to the gpy board completely and then gives it power back after the set time.
I also have a temperature sensor connected to the devices over I2C. I am using a sparkfun board with a TMP117 sensor (https://www.sparkfun.com/products/15805). This is how it is hooked up:
3V3 sensors pin on pytrack/gpy -> V+ on the TMP117
GND on pytrack/gpy -> GND
P9 -> INTERUPT
P21 -> SCL
P22 -> SDA
The device wakes up periodically (1-5 minutes) or when interrupted using a pin.
Then reads the temperature from the TMP117, reads GPS data (time, coordinates, etc), and transmits it over to a server using LTE.
Then it goes back to sleep.
However, it stops waking up after some time. Sometimes it could be about 5 hours, other times it would stop booting after 10 minutes. I haven’t noticed a pattern, but I am running a few experiments right now to try and find it. I suspect it might becoming less stable when the battery is getting discharged more and more. But I haven’t done enough to actually confirm this hypothesis.
No worries, we’re here to help! It sounds like all of the wiring and devices are correctly connected, it may be worth increasing your battery capacity to see whether you can achieve the same results. Your best bet in this case would be to add an external watchdog timer, to compare whether the same incidents occur after a set period in time. I’ve linked a suitable timer below that should be able to provide a suitable output to the required interrupt pin in order to refresh the sketch. The deep sleep functionality shouldn’t have a limitation on the time for which it is able to perform that action, but it sounds like there may be some kind of issue with an under-voltage creating some unusual logic. Please let us know how you go!
There are low quiescent current, Low Drop Out (LDO), voltage regulators available you could use as an alternative to replace the on board regulator - which might solve your issues:
From memory the regulator on the Pycom boards has a quiescent current of around 20mA.
After doing some more troubleshooting, I suspect that the reason it stops working is the connected temperature sensor over I2C.
When the TMP117 sensor is disconnected the pytrack/gpy keeps waking up without an issue. Even if I only connect the TMP117 to the power and ground, without connecting the SDA, SCL, and INT wires.
