Home        Store        Learn        Blog

Custom ROV disconnect on full throttle

Hello everyone,

Before starting, I will explain you the goal of my project and how it is set up.

We are students and for our final project we are currently developing an ROV capable of identifying waste in underwater environment, the original idea is to use an underwater robot to perform this task and incorporate our identification program by machine learning. However, we are students and don’t have much money to buy a BlueRov2, so we opted to build a custom ROV according to our needs.

We followed all the explanations on Ardusub, the different assemblies. Today we have a ROV (simple-rov-3) that connects well to QGroundControl, I get the position of the gyro and my external GPS (which is not always displayed by the way…), when I test the motors in the settings section: engine the 3 motors work perfectly, but if I launch a “fly” “Armed” when I push the throttle to the limit, the connection is interrupted, but the robot is still working simply no more link between the robot and the computer parent.

The robot contains 3 brushless motors of 3100KV, a battery of 12v 17Ah, connected by ethernet to the pc via a raspberryPi which transmits the info from the PX4.

So now I’m stuck, I ordered a DC Converter to increase to 14.4V the whole thing hoping that this is the problem.

Do you have any idea what could cause this disconnection that does not turn off the ROV?

Is the battery sufficient?

PS: the battery is not on the ROV but on a floating solar platform with a charge regulator for a certain autonomy.

Thank you for your answers.


Hi @ekivox ,

It sounds like your system is experiencing excessive voltage drop under load, either due to a battery or tether than cannot handle the current draw requirements of you system. Make sure you battery can supply the current required, and you can use our voltage drop calculator to determine if your tether is up to the task. You will likely need to to increase the thickness of your tether power wires, depending on it’s length.


1 Like

Hi @adam,

Thanks for your support, i’ll change wires for a bigger diameter and try my DC Converter to check if the voltage can be the cause.

1.5mm can works perfectly ? (shorturl.at/doEK7 amazon link)

When i try to test each engine one by one, the rov keep connection up but crash only on FLY (never crashed in test) dont know why…

Thank you !

Hi @ekivox,

Not sure what you’re referring to here, and your link doesn’t seem to work/go anywhere.

3 thrusters can use 3 times as much power as 1 thruster. When you’re testing one at a time then you make sure each thruster is working correctly, but it’s only testing whether the power supply is capable of running one thruster. When you then start running the ROV normally it uses multiple thrusters at once, and that can be too much power draw for the power supply to handle.

If you’re interested, I do a relatively detailed breakdown of power usage and conversion efficiencies for a surface-based power supply in this post. You can apply similar ideas to the cables, converters, and equipment that you’re using to get a sense of how much capacity your setup provides, and/or go in the opposite direction to determine how much capacity you need to provide to support full thrust of all your thrusters while still running the rest of your electronics :slight_smile:

Hi @EliotBR,

Sorry i wasnt clear with my example of testing engine from TestMode, i’ve done a test with a single Engine : test mode is ok, but Fly disconnect (same with, 1, 2 or 3 engines) don’t know why. But with a multimeter, i’ve seen the voltage goes under 9v…

I’ll read your topic about surface-based power supply thanks a lot.

Sorry for the link : (New Link To Amazon) or this one : 16 AWG wire

In the test mode you use a slider which will have a bit of a ramp while going up to faster speeds. It’s possible that in ‘fly’ mode there’s a large current spike when a thruster is turned on to high speed immediately, which could cause a brief drop in voltage. If several thrusters are used at once then that effect is worse, and there can also be a quite large constant current from driving multiple thrusters (or even driving one thruster quite hard).

Right, you just meant a cable with 1.5mm^2 conductor area. That’s a parameter that you can put into the voltage drop calculator that Adam linked you to earlier, along with the voltage that you’re transferring through that wire, and the required current the ROV’s electronics will draw (motors + lights + 5V converter capacity, etc)

Hi everyone!

Good news for me today! Last night i was thinking about montage, wiring and every component one by one, and i spotted one thing weird. i placed my companion-computer in derivation of the main branch (serie of ESC + engine), i removed it and give power directly from my solar converter USB (2a) and everything work perfectly !!

For the GPS i made a bad configuration with GPS_TYPE… i was trying uBlox or NMEA but AUTO works perfectly. Can i know what AUTO value is when it defined the best to use?

Thanks for each answer, you guys are really helpful!

Last question for me : how can i reduce latences between my computer with usb controller and the SUB ? Radio is really better? We are working on WIFI with a RaspberryHotspot and a very good antenna (400mb/s) and sometimes, no latences and another are 1 or 2 real secondes… Any tips ?

Thanks a lot!