Software crashes and robot shuts down

Follow

Hi GB,

It seems like your 5V cannot supply enough power and drops.
If the 5V drops, the arduino shuts down, the communication with the software shuts down, and the software probably crashes.

Make sure the power switch is ON.
If the power switch is OFF, the robot will run on your PC USB 5V, which cannot supply enough power.



It's also a good thing to test the 5V supply.
Unplug your USB cable and power the board.
Measure the voltage with a multimeter on the arduino to see if you have 5V:

If everything is normal, I would connect only one servo motor at a time on the PCB and test it with the software.

When you say «software crashes», do you mean it closes ? It freezes ? Communication is lost ?
Also, when the «robot shuts down», do the LEDs on the PCB shut down ? do the LEDs on the Arduino shut down ?
If the LEDs shut down, it's definitly a power problem.

Software closes, robot goes limp.  After 10 to 20 seconds the robot reboots and all servos quickly and simultaneously return to 90 degrees with no issues.  The LEDs remain lit.  My power supply is rated above the recommended wattage.  Voltage measured at the servo pins is 5V.  I think the J2 servo pulls to much current and flips a "breaker", but I don't know this for sure.  I have tested it without using the J2 servo and everything works fine.  If I apply to much of a moment to the J2 servo (fully extend the arm at J3 and try to rotate with J2) it will also cause the crash.

Thanks for the help. I will be away next week so I may not respond, but please offer any suggestions.

I just did a test on my robot:

First I connected to the robot with the PC software.
Then I unplug the USB cable (connection is lost).
This causes the software to shut down.

So I am pretty sure thats whats happening when your software shuts down.
And I think your communication is lost because your Arduino lacks of power when J2 is pulling too much current.

On the board, the component U3 converts the 12V coming from wall mount power supply to 5V.
It can supply 3A, which is normaly enough.
Are you also powering an external device with the 5V ?

Whats the current rating of your wall mount power supply ?
 
A good test would be to measure the voltage drop with an oscilloscope on the 5V when it happens.
Or to supply 5V to the board with a bigger power supply (5V but higher current limit) and see if it still happens.

Have you fixed something heavy on the end of arm ?
A tool too heavy would cause that problem.

Another solution could be adding a bigger capacitor (component C2).
The capacitor would supply power in high demands.

If none of that works, there might be something wrong with your servo motor.

The power supply is 3A, 36W output @ 12V.

Nothing else is being powered by the power supply.

Nothing extra on robot (end effector, tool).

What is odd is when the servos reset to 90 degrees at reboot it has no problem moving all of the servos at one time very quickly.  

I will try a few things and get back to you on how it works, but it will probably be the week after next before I can.  I will probably try to resolder some spots and I may try the capacitor idea.

Thanks for the help. I will let you know what happens.

Hi on mine the regulator temperature shoots up so fast and then will lost connection and will initialize 

Pukindat JB I think that is the problem.  I tried a few things but nothing fixed it.  Then I checked the regulator temp while repeating the motion that causes the crash.  The temp of the heat sink got up to 170 F then it crashed.  Going to try a larger heat sink.  If that does not work I may try a different regulator or a linear regulator replacement. Will let y'all know. Thanks for the help.

I added the heat sink and that seems to help.  Heat sink gets too hot to touch, but no crashing.  Probably would be a good idea to add a heat sink to the BOM.

GB, I can confirm the addition of a heat sink on the voltage regulator helps prevents crashes for me. I printed the components using more material than recommended (adding weight, and causing the servos to draw more current). I believe adding a larger capacitor would also solve the issue.