3DR power brick and PX4 - leads to in-flight reboots?

[Andrew Tridgell]

Hi All,

I've connected a 3DR power brick to a PX4, and I have it working for
current sensing to some extent, but I noticed a potential problem that
is probably worth making clear in our doc pages.

Note that the 3DR power brick was never designed to work with the PX4,
but I had one handy, and I think it is likely that others in the
community may also try to adapt it if switching from an APM2.5 to a PX4.

According to the specs, the 3DR current sensor is a 90A part over a 5V
range. The current sensing pin on the PX4 has a range of at most
3.3V. So I wondered what would happen if you put more than 3.3V on that
pin. I found that if I put 5V on that pin the PX4 FMU rebooted.

So this means if you have an aircraft that draws more than 60A on
takeoff, then you may well find the FMU reboots when you takeoff if you
use the 3DR current sensor.

The other thing I would like someone to check is the schematic for the
3DR power brick. It is I probably just don't understand the schematic,
but I can't see anything in it that limits it to 5V. What if you measure
120A, does it show 5V out, or does it go above 5V ? If it goes above 5V,
how would the APM2.5 handle this?

Cheers, Tridge

[Ritchie Wilson]

I don't have the schematic in front of me but I'd assume a divider for the maximum current. So going over will either be higher than 5v or cause it to stop working dependent on method

Regards
Ritchie.

> --
> You received this message because you are subscribed to the Google Groups "drones-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to drones-discus...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Andrew Tridgell]

Hi Ritchie,

> I don't have the schematic in front of me but I'd assume a divider for
> the maximum current. So going over will either be higher than 5v or
> cause it to stop working dependent on method

the schematic is here:

http://stuff.storediydrones.com/3DR-APM25-Power-Module-V10.zip

perhaps you or someone else who understand schematics can look at it?

Cheers, Tridge

[Bill Bonney]

The voltage divider is 10:1 so at 60V input you get 6V out (in this case the regulator would be fried at 18Vmax)

The current is measured by amplifying the voltage across a very small resistor 0.25mOhm. The amplifier is powered by the Vbattery so the max out is Vbatt, so for 4S that would be 16.8V and 3s 12.6V but to get these voltages the current is going to be huge!

@120A you would get about ~7V

Not sure what the AVR chip will do through.

Bill

[Olivier ADLER]

I see no overvoltage protection on the schematics.

If the AVR can crash or reboot with more than 5V this could be a problem specially fo Hexa during agressive fly.

If a motor or ESC become shorted during a few milliseconds before final death (the time needed to burn the wires or PCB tracks) the APM could reboot or crash as well.


Nevertheless, the AVR do have clamping diodes on all inputs, except /Reset. For this protection to work the input current must be limited to about 1 mA. I can't see any serie resistor on the schematics so i suppose that the current is limited only by the op amp output capacity (the op amp measuring the shunt voltage. This one has certainly a current capacity of 10 mA or more.

Obviously this is something to be tested to be sure, and issue a warning to the user base if the test do show a possible crash or reboot.



Olivier

[Arthur Benemann]

The INA169 has a current output, that is converted to voltage on the load resistor. At an output voltage of 5V the current flowing trough this resistor is 45uA (5V/110koHm), which a lot less than ESD clamp diodes in most chips can handle.

An easy way to solve it is to add external clamp diodes to the power supply positive rail.