Frequently asked questions are about powering an APM2.5 or why the same does not work or why it has failed.
Here is a simplified schematic showing the major components of the 5V/Vcc circuit on the APM2.5
Target Audience: Readers who do not have Eagle or are confused about using it.
You Need: Basic electronic schematic reading ability
Note that I have separated the PWM Output section into the upper left corner with a heavy purple line that is bridged by the jumper JP1 on the APM2.5. This is to emphasize that diode D1 passes current through JP1 (and drops voltage). The remaining energy is then called 'Vcc'. Everything outside the purple corner is Vcc.
This is why the USB will power the APM2.5 when connected to a computer/USB hub. Note the fuse on the APM2.5 side of the USB connector AND another fuse on the output of D1. Another change from APM2.0 is the D4 6.0V Zener diode. If an over voltage is applied to the board, D4 will crowbar what it can to protect devices.
As with APM2.0, any 5V source connected to any of the indicated Vcc/5V points outside the purple corner (with GND of course) will power the APM2.5.
Also of note is the 'Mystery Port'. This port takes 5VDC before D1 and the fuse and is called 'JP_VCC'. Other clues to the mystery are the lines to 'ADC 12' and 'ADC 13'. A mystery indeed!
Any errors in this drawing from editing are mine.
I hope this can be used as reference for solving problems.
27-FEB-2014 Eagle files are now located HERE
I never quite understood why I cannot power an APM 2.6 from both the servo side and the power module side
The power module is directly connected to the 5V Servo rail with JP1 installed.
The diode D1 stops servos from being powered from the USB port.
The diode D4 shorts to ground with a voltage over 6V, the main load being dissipated with the 500mA (now 1.1A) solid state fuse F2 (as current goes up, its resistance increases, having a larger voltage drop across it)
Aaah, the power module isn't protected from powering the servo rail with the diode when JP1 is installed?
Ok, not sure why it would be designed that way, but now that makes sense, why you'd have to remove JP1 if you use the power module.
There's one big thing I think we're all missing here. Imagine I"m using one battery to power the plane, and a BEC on the output rail. Even if I have JP1 open, the servo return current, can still go back to the battery via the ESC leads OR the APM ground traces. THIS IS A GROUND LOOP! I think there should be a way to disconnect the ground between the output rail and the rest of the APM board, in the same manner as JP1 disconnects the supply voltage.
I have a question that's bugging me.
In some places including arducopter site (compassmot specifically) it's saying we CAN connect both BATTERY and USB at same time
In a case where I use a power module and no JP1 installed is this true or not ?
Thank you for your question.
If everything is wired correctly - we have seen many cases where this is not true - having battery (to power module) and USB should not harm anything. It is discouraged because the ESC are then powered and if you are working via USB to the APM, testing system parts, we certainly do not want the props to spin up.
The USB port is fused. I wish there was a power removal jumper for the +5 but, Pixhawk is the future and this entire thread may one day be retired when the hardware is no longer supported.
I have this weird problem with apm 2.6 power supply related!
These tests, Rx Connected to apm input rail, 3.2 FW, Quad config.
1) 5v analogue ubec connected to apm output rail, JP1 in place, measured input rail 4.4v
All functions work, arms, motors run for about 5 secs, varies in speed and then cuts, one blip of the
throttle, then returned to low and won't run again without rebooting!
2) as (1) above, but with USB also plugged in, works fine, input rail 4.67v
3) 5.01v analogue ubec connected to apm input rail, arms but immediately resets. Blinking led
4) As (3) above, but with USB also plugged in, works fine, input rail 5.05v
5) APM power module connected, won't arm, input rail 4.8v
6) APM power module connected, but with USB also plugged in, works fine, input rail 4.83v
If you follow the above tests, it doesn't seem voltage related, but appears that the usb supplies a part of
the circuit that the other supplies don't, Anyone care to comment?
The esc's are not the cause, have used them on many previous UAV's. This is about not arming, and occasionally arming then immediately resetting (not armed), There are no pre arm check problems showing, when connecting the usb, I'm not opening MP, just using the laptop USB v supply...
A UBEC will need more input voltage than the output.
example: +12VDC ---> UBEC ---> +5VDC
If using APM power module, JP1 should not be connected.
USB power should only be used for log file transfers and other PC/APM related work.
Having the PM and USB supply on at the same time could lead to a ground loop if you wiring is incorrect.
Could you provide a sketch of your wiring?
Thanks for the reply Doug,
With the PM supply only, JP1 not connected, mis behaves, I'd be inclined to say that something is receiving marginal voltage! Meanwhile voltage at the input pins is 4.8v
Now add usb supply (I guess connecting this is ok because USB v is not far off input pin v) Now works reliably! Meanwhile voltage at the input pins is 4.83v. Ie only 0.03 higher than without usb connected.
Supplying only input pins with full 5v = mis behaves, just don't make sense!
Its as if usb +v is supplying something differently!, will scope it tomorrow, see what difference there is, will also supply the apm with an ultra smooth supply, see if that works!
Meanwhile still scratching my head! It's not a duff PM either, I have 4 here!
Ground loop could come into play when USB is connected but we are only talking about 140mA or so!
The logs can/could also be analyzed for dropout/brownout conditions.
Though not a direct 5V sample, it is/was done by an algorithm used in Arduino code groups and implemented in APM Copter...at least before the last release. I've been running fixed wings for months now and dust is collecting on my rotary wing birds so.. not sure what is current regarding logs.
Have had chance since to test again with apm 2.6, apm 2.5, different esc's and it seems the cause is a bug in the 3.2 code! Loaded v2.9 worked fine, loaded v3.1 fine! 3.2 same old problem...
so currently running on v3.1
If I get time I will look into the code changes between the two!
I have no clue why the older versions would impact Vcc.
Have you checked the other forum for information?
I don't think that it is voltage related anymore!
The apm is disarming after being armed for about 10 secs, some times unable to re-arm, have to do a reboot!
But if plugged into usb then works fine every time, thinking now its something to do with usb signal lines rather than pure usb voltage.
NUTS? Yes I agree!