Hi everyone. Iv been looking at the thruster configuration for the new chasing m2 ROV. Its really interesting. I’m sure Iv seen that type of design in this forum before but can’t seem to find the vehicle it was on.
The design looks really stable and pretty fun to fly.
My question is. With 8 thrusters vectored in the configuration of the chasing m2, Would you get the same amount of thrust in the fwd, reverse, lateral movements as you would with the standard 4 vectored thruster configuration?
I’ve seen that design a few times too from Boxfish Research and RJE Oceanbotics so its not a new design.
Boxfish Research:
RJE Oceanbotics SRV-8:
M2:
Its an interesting frame design for sure! I think the only downside is that it isn’t inherently stable, so the gyro and control system really has to be spot on. The M2 configuration appears to be more similar to the Boxfish frame with the thruster arrangement. The thrusters are opposite on the SRV-8, I don’t think it matters.
As for the thrust question, one of us would have to do some vectored thrust calculations, but I think it would be equivalent to the standard 4-thruster vectored configuration.
I’d be interested to see if anyone makes up a custom frame design based on ArduSub!
Hi Kevin,
Thanks for the reply. Im sure some one will try it with a Blue Rov at some point. will be interesting to see how it handles.
Hi all,
I was wondering if any progress was made to integrate this type of thruster vectors.
It seems a good way to maximize propulsion especially for tunnel inspections.
Cheers,
E.
Hi @etienne,
It’s not one of the configurations that’s available by default, so would require building ArduSub with a custom frame configuration and setting the relevant motion components of each thruster. That shouldn’t be overly difficult, especially because these configurations seem to have all the thrusters with the same relative rotations, so the factors would be the same for each thruster, with some minus signs flipped.
Hi @EliotBR,
Thanks for this. This is quite simpler than I thought.
I was hoping someone had done this already. If not, I will share my compiled ArduSub build and detailed instructions when I have done it.
Would BR consider adding this as a standard setup?
The mention 90deg refers to the orientation of the pixhawk right?
Cheers,
E.
Hi @etienne!
90deg here means that the horizontal thrusters are pointing either forward/backwards or left/right, opposite to our regular configuration where each thruster is pointing 45º off from forward.
Note that the first 4 thrusters (the horizontal ones) only control either forward or lateral motion, never both together.
We have an open issue for that. we just need someone to fill the table. If you share it here, we can get it into next Ardusub release.
1 Like
Hi Willian,
Its been a while. 
I will post my results when I get them. I am actually planning on limiting my release to 0.75 because its what works best with my power supply.
I have a client going for this in January so will work on it then but I also have another one that may want it earlier.
Cheers,
E.
That’s right.
so for the “fully vectored” frame, each thruster will contribute to each of the 6 degrees of freedom
1 Like
It may make more sense to allow the full range but set the limits for a given vehicle using parameters (as covered here)
This is preventing me from doing this.
The option you want is the RC3_MIN etc point (a couple of points below that), which there’s a video for how to do at the bottom of the comment.
Cool, I will look into it.
Here is first draft of the frame.
I still need to add trays on the top and bottom for newton, dvl, ping1d, cables, etc.
Bracket for pin360, Waterlinked beacon and more lights and ballast.
Buoyancy is around the 2x 4inch housing.
Back housing is for second BR2 camera, battery/power module, additional electronics.
Aluminum block in the center for bulkheads.
Any comments, suggestions?
Once I got this done, I’ll post step files for those interested.
Cheers,
E.
2 Likes
Looking slick!
I’m curious how the outer tubes are connected to the frame, as well as how the electronics enclosures are connected to the central aluminium block (would they slot onto opposite sides of it or something?)
I like the symmetry and smoothness - looks like it would have decent drag characteristics, while also being quite hardy to move around as necessary
This center block will be fixed to the frame on either side.
Center block has o-ring carrier.
Via threaded rods and threaded caps on the outer posts. that will secure everything hand tight.
If its not enough I will add a couple more near the center.
I might be have the buoyancy one solid block that just slides over the tube.
2 Likes
Hi everyone,
I have been designing a proof of concept ROV that will have motors in a configuration similar to Chasing/Boxfish with all motors vectored. See image below.
If it works well, would you consider adding this configuration to QGC so end users do not have to go through the manual setup/firmware?
Hi @gwa-gwa, I’ve moved your post here because it’s a more relevant discussion.
It would need to be added to ArduSub first (it’s not possible to select a frame that doesn’t exist in the firmware) - there’s some previous discussion above:
Once it’s in ArduSub I don’t see a reason why it couldn’t/shouldn’t be added to QGC as well
Any issues with: SUB_FRAME_FULLY_VECTORED_6DOF as the name of the config? I will make an ardupilot pull request.