I Have inserted the starting pass at a much needed section on Vibration Control in the ArduCopter Wiki and would very much like feedback for corrections or additions from you all.
This is the close coupled mount that I use on my little F330 Flamewheel.
To achieve this plus and minus one tenth G vibration dampening.
In any case, please review the Vibration Control Wiki section under Assembly and give me some (constructive) feedback.
Thank You,
Gary McCray
Comments
Hi Steven,
Moon Gel or Kyosho Zeal (which is very similar) have shown very good results in this application.
Having the APM and case mounted on a separate plate is a good way of building mass and quite likely the size pads you have cut will work well.
Using double sided foam tape between and on both sides of them introduces an interesting and likely worthwhile secondary damping capability that would likely have a different frequency resonance characteristic than the gel.
Probably a good thing and their different frequencies would probably augment each other in damping.
I would caution you not to use too thick or too soft of a foam tape.
You really don't want the APM/Case/Base plate assembly to move noticeably in response to normal airframe motion.
That said, I think your system is likely to work quite well.
Although anything on the RAW ACCEL log file scale under 5 (~1/2 G) is acceptable, you are likely to acheive and should try for 1 or 2 (1/10 to 2/10 G).
Can anyone pass their opinion on this proposal for vibration damping please.
I am building an Octo which is completely carbon fibre based.
The APM case will be mounted on a separate stack plate which will be fixed to a lower plate using soft gel anti-vibration bushes.
I have read somewhere that when using moon-gel for this purpose, its a good idea to use 2 pieces stacked on top of each other. Therefore I have cut 8 pieces into 13mm 'pads'. In order to fix the two pads together, I am considering using self-amalgamating foam tape which will also be used to fix the pads to both the APM plate and APM case.
Anyone have any views on this proposal before I get the APM plate CNCd.
Many thanks,
Steve
Yep - there has been a few reports of massive altitude jumps from people lately which has been put down to excessive accelZ vibes, but if the board can detect that vibration is excessive it should probably either revert to the previous behavior (if possible, I know lack of memory is a problem on the APMs) or abort the flight mode that requires it to be vibration free, putting the user back to Stabilize mode.
I also think any mode that requires GPS lock shouldn't engage if GPS lock wasn't established before takeoff. It's a lot easier to answer people's queries of "my copter won't go into auto mode, what gives?" than "my copter behaved in [insert myriad of strange behaviors here] manner, what gives?"
I think the software would benefit greatly from better checking of its inputs before actioning complicated outputs.
Interesting thought Andrew, Certainly in Loiter Mode or alt hold, vibration could overwhelm the Inertial compensation.
You could switch back to stabilize and that would no longer be a problem.
But it might also be a good idea for the flight control board to auto detect excessive vibration and to increase filtering or even disable inertial navigation automatically (or adjustably in any case).
The vibration frequency is normally considerably higher than normal flight maneuvering frequencies so this condition should normally be detectable.
The other thing that has been worrying me with all this focus on vibration control is that we could lower the likelihood of a hex/octa maintaining control in a partial prop loss situation, which turns any multi-rotor into a shaker-bot.
I hope the search for anti-vibe perfection and the benefits it brings doesn't have us forget about this emergency case, it's really important that when we're flying these larger and heavier rigs that they don't just fall out of the sky in common failure cases such as this.
Yes Gary, it has to be addressed formally, sooner or later. I'm sure it's being looked at.
Thing is, it almost certainly can't be something as simple/functional as what I euphemistically called "grommets" above. Due to the APM's low mass, though one of the custom isolation places could likely makes some very soft captive screw grommets that might work well enough, and not be expensive in 3DR-type quantities. Rather it's more likely that a slight redesign of the "stack" portion of the copters will provide a better future path (IMO it would...), and with this greatly increased mass to isolate, all sorts of inexpensive commercial options become available. This will probably take longer to implement... Whatever they do, 3DR doesn't give out much advance info of products, just obtuse hints lol.
Do 90% of the job for 90% of the people: I'd say that's a great success. The other people (most of us here...) will tinker anyway, from that good starting place.
I agree Craig,
The logical thing is that now that vibration control is actually necessary, 3DR ought to build a sufficient example of it into the standard "enclosure".
If they don't, it's going to be a continual case by case headache.
In no small part, this section and the section I put in the Wiki was to assist them in becoming aware of the importance of solving this problem.
If they don't do it themselves, it will in the end be their customer service department theat keeps on hearing about it.
I won't be satisfied until there are some kind of "grommets" that can be screwed down. A totally repeatable method/process that also anchors the APM, IOW something that 3DR could include in a kit and say "use these" and pretty much no other explanation and directions are required. (Besides balancing props/motors that is not unique to this app.) This is totally doable. Easier by a larger entity who can get engineering samples, or more likely a custom-made product (lots of places do that). I can do "random" too, and sometimes have no option, but it offends my engineering sensibilities. No offense to those it doesn't offend. :)
Josh on your Discovery frame you could run a rubber band through each of the frame Velcro slots and around the outside of the top plate, looped back through itself and connect 2 (or 4) of them in the middle over your flight control board using a plastic circlip or binder ring.
Just a thought and easy to remove.
Rubber bands should be chosen to keep board from moving while not applying much more tension than that needed to do so.
-
1
-
2
-
3
-
4
-
5
of 12 Next