Have just put together the BlueROV2. Build all went well and seem to be working until I came to checking motor directions. Thruster 6 (top mounted) does not seem to respond correctly to inputs. Most of the time it does not spin at all. If I alternate directions, it will sometimes spin (but only in the one direction). When it does spin, even if the input is set to low (thruster 5 is spinning slowly), thruster 6 will spin at full throttle for appx. 0.5 secs and then stop abruptly with a screech. The motor spins fine in both directions when turning by hand, and hasn’t yet been used in the field or for long periods of time out of the water. The most worrying is if I leave the unit armed, sometimes the thruster will start spinning (which I have read is not unusual for a stationary vehicle), but it will ramp up to maximum speed, and then I am unable to disarm and have to pull the battery.
I have checked:
-mavlink shows channel 6 responding to input, but no spin
-the thruster. Disassembled and inspected; all visible parts seem fine and propellor spins freely by hand
-joystick calibration
-connections at ESC/motor terminal
-checked connections at Pixhawk
-haven’t yet plugged another thruster into channel 6 ESC/Pixhawk to test
It is possible that one of the ESCs slipped through without being re-flashed with the forward/reverse firmware. Here’s how you check:
The stock firmware (single direction of rotation, the way we get them before re-flashing) on the ESC requires a consistent PWM signal of about ~1000-1300 in order to arm the ESC. The ESC must be armed in this manner before it will spin at all, regardless of other PWM signals received. With this firmware, 1000 is low throttle and 2000 is high throttle.
The Blue Robotics ESCs require a PWM signal of 1500 to arm. 1100 PWM is full throttle in one direction, and 1900 is full throttle in the other direction.
This is what I think is happening:
When the BlueROV boots up and is disarmed, it sends a signal of 1500 to all of the ESCs. The Blue Robotics flashed ESCs are armed, and in the neutral throttle position. The rogue Stock ESC hasn’t armed yet and is at 50% throttle. When you push the throttle stick (because it is the vertical thruster exhibiting the behavior), the Blue Robotics ESC spins normally, but the Stock ESC does not, until the PWM on the Stock ESC reaches 1100~1300 and it arms. Then when you return the sticks to neutral (or disarm) the PWM on all ESCs goes to 1500, the Blue Robotics ESCs stop, and the Stock ESC spins like mad.
Can you verify this behavior using the Mavlink Inspector? You can email support@bluerobotics.com with your address to send a replacement. Sorry about this!
Thanks for that. I haven’t been able to replicate the runaway thruster this morning, but the thruster is still only spinning one direction. Some more info:
You say the thrusters are min at 1100 and max at 1900, mine are min at 1300 and max at 1700. Is this an issue? BlueROV2 was shipped in November 2016, maybe need to update firmware?
The pattern of behaviour is at throttle up, mavlink shows ch 5 to 1700 and ch 6 to 1300, but only thruster 5 spins. I can’t get thruster 6 to spin in the throttle up position at all. Even when PWM goes to 1300, the thruster does not spin. Maybe PWM not low enough to arm thruster on this stock ESC?
At throttle down, mavlink shows ch 5 to 1300 and ch 6 to 1700, both props spin but prop 6 seems to be faster(??hard to tell), and makes a screeching noise. So not sure if this is mechanical, or a combination of signal and mechanical?
In Mavlink, in RC Channel Raw, all channels are resting at 1500, but ch 5 is at 1100. Not sure what this output is, but is this a problem?
Do you still think this is an ESC issue? I’m curious why the thruster would also be screeching, and wondering whether there is also a mechanical issue, or is it possible to cause this noise by a rapid speed ramp (ie from disarmed 50% throttle to armed and 50% throttle almost immediately).
The input gain is at 50% by default, you can increase the gain to 100% to reach the 1100-1900 range. You should update your firmware soon, ArduSub 3.4 was the first stable release, and 3.5 will be out next week.
Channel 5 on the input is a ‘mode switch position’ indicator. It is a legacy thing from ArduCopter. In 3.5, this is removed.
The noise is expected when running in air. The thrusters aren’t designed to run in air for extended periods or at high throttle, they will sound much better in water!
This is an issue with the ESC, the ESC on output channel 6 doesn’t have the forward/reverse firmware. Please email support@bluerobotics.com with an address to send a replacement.