Hello,

I currently have a USV equipped with a port and starboard thruster. We are working on transitioning from a skid steer system, to dual rudders and we are having issues with our thrusters turning off unexpectedly.

Here is a basic layout of our electrical schematic of the propulsion system:

So the pixhawk outputs a PWM (no ground necessary) to both ESCs and both ESCs are also sharing the same power source with separate fuses. Note the pixhawk is also powered from the same power source but with a regulator in between. 

The Problem:

If the fuse blows (or positive is disconnected) going into the port ESC, both thrusters will stop working. The same thing happens if the ground is disconnected rather than the positive (or if both are disconnected). One would think that only the port ESC would stop working rather than both... right?

Our Theory:


If the pwm going to the port ESC is disconnected, while + and/or - is also disconnected, the starboard thruster starts working again as it should. This leads me to believe that the PWM signal which should be going from the pixhawk to the starboard ESC is somehow being "drained" into the port ESC. Perhaps the giant capacitor on the non plugged in ESC is absorbing all of the signal?

My Currently Pursued Solutions:


If our theory of why this is happening is correct, then using an NPN transistor as a switch could work. If the ESC loses power, the transistor closes its gate, and the pwm line is disconnected, and thus the other thruster would work.

I am also looking into using an arduino to read the pwm from the pixhawk and output that same signal through two of its own pins, and thus isolate the PWM between the ESCs.

Ideally, I would like to be able to just output the main thrust commands through 2 of the pixhawk rc pins, but I have not figured out how to do that.

If anyone has any ideas as to the cause of this issue, or would like to verify our theory, please let me know. Thanks for any help.

Adam

Views: 381

Replies to This Discussion

Adam,

i think, the ESC with the broken Fuse will pull the PWM SIgnal somehow to ground. An easy way to isolate the two Signals is a "wired OR" built with diodes. Because the PWM Signals are not that fast, this may work.
I have uploaded a pic with the schematics. As Diodes you can use standard ones (i.e. 1N4148).  
Hope this helps

Harald

Hi Harald,

Thanks for your help. Could you explain how your wiring diagram works please? Or maybe point me in the right direction? I am more of a mechanical engineer rather than electrical so I dont understand a lot about circuitry. Thanks again.

Harald Molle said:

Adam,

i think, the ESC with the broken Fuse will pull the PWM SIgnal somehow to ground. An easy way to isolate the two Signals is a "wired OR" built with diodes. Because the PWM Signals are not that fast, this may work.
I have uploaded a pic with the schematics. As Diodes you can use standard ones (i.e. 1N4148).  
Hope this helps

Harald

OK, 

First, you will need 

- two diodes (e.g. 1N4148, wired types, no SMD)

- two resistors 1K, wired 

connect the diodes on the side marked with a bar together (the cathodes).

Connect  this side of the assembly to the PWM Signal that comes from the pixhawk.

On the opposite side of the diodes (the anodes) connect a resistor on each leg.

Connect the opposite sides of the resistors to the +5V leg of the control wires of each ESC (+5VESC1, +5VESC2) (normally red or orange) 
Connect the anodes of each diode to the corresponding PWM input of the ESC.

Thats it.

What happens:

The AP issues the PWM (PPM) Signal which is low for approx 20ms.

Then it is pulled high for a time between 1ms and 2ms (pulsewidth time)

When the signal is high, on the anode side of the diodes, the signal is pulled also high by the Power Supply from the

ESC (should normally be around 5V).

If the signal goes LOW (e.g. ground), current flows over the diodes, pulling the voltage on the anode side of the diodes
to approx 0.7V, which is also LOW.

If one of the ESCs fail,  the SIgnal on the anode side of the failing ESC will be pulled to LOW, but the Diode

prevents, that the LOW propagates backwards.

Great. Thank you very much Harald. 

This solution worked really well and I learned a lot in the process.

Adam

Nice to hear that it works!

Adam Day said:

Great. Thank you very much Harald. 

This solution worked really well and I learned a lot in the process.

Adam

Hello,

I'm facing the same problem, i.e. trying to integrate four (4) thruster to my boat, to improve redundancy, and I want to send the same output PWM signal pair-wise to two thrusters (skid steer setting): both LEFT ESC should receive the same input and both RIGHT ESC should receive the same input.  

It works of course fine with Y cables from the PIXHAWK to the RIGHT and LEFT ESCs, but when one of the ESC is not powered, even the second switches off.

I wonder if it is possible to adjust the settings of the PIXHAWK software (ardurover) so to be able to couple the 4 ESC directly into the PIXHAWK, and have them fed pair-wise with the same controller output. In principle it's about duplicating the LEFT/RIGHT output to a second pair of pins.

Compared with the solution based on the diodes this option would be more simple, not requiring the additional circuit.

Do anyone have any experience on that ?

Thanks in advance

Hi Roberto,

This was my original idea for solving the issue but unfortunately I did not actually find a way to do it. I tried messing around with the setting for the ServoX_function params in mission planner but had no success. For example:

http://ardupilot.org/rover/docs/parameters.html#servo10-function-se...

Setting the value of this parameter to 33 seems like it would map the thruster 1 pwm signal to the servo10 on the pixhawk. So I thought that maybe if you use the value 33 for servo 10 and 11 for example, you would duplicate the pwm like you said.

Unfortunately this did not work.

Please let me know if you find a way to accomplish this.

Thank you,

Adam Day



Roberto said:

Hello,

I'm facing the same problem, i.e. trying to integrate four (4) thruster to my boat, to improve redundancy, and I want to send the same output PWM signal pair-wise to two thrusters (skid steer setting): both LEFT ESC should receive the same input and both RIGHT ESC should receive the same input.  

It works of course fine with Y cables from the PIXHAWK to the RIGHT and LEFT ESCs, but when one of the ESC is not powered, even the second switches off.

I wonder if it is possible to adjust the settings of the PIXHAWK software (ardurover) so to be able to couple the 4 ESC directly into the PIXHAWK, and have them fed pair-wise with the same controller output. In principle it's about duplicating the LEFT/RIGHT output to a second pair of pins.

Compared with the solution based on the diodes this option would be more simple, not requiring the additional circuit.

Do anyone have any experience on that ?

Thanks in advance

Adam, thanks for your answer.

I'll do some more attempts.

It would seem very silly to me if it could not be done by tuning the software... it is just a simngle line of code to be added at the end of the thread, just copying teh PWM value to an additional assigned servo output channel.

If I found a way I'll let you know

Regards

Roberto



Roberto said:

Adam, thanks for your answer.

I'll do some more attempts.

It would seem very silly to me if it could not be done by tuning the software... it is just a simngle line of code to be added at the end of the thread, just copying teh PWM value to an additional assigned servo output channel.

If I found a way I'll let you know

Regards

Roberto

Hi,  we have faced the same problem with our USV. We have attempted to solve the problem by adopting possible optic separation, but without success.  We also believe that there should be a software solution, but we haven't fount it yet. Any contribution in this direction will be welcome !!

RSS

© 2017   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service