Thread Tools
Jul 04, 2019, 07:01 AM
OlliW
Thread OP
I'm also seriously thinking about it
the material is clearly a good and key question, I would think it also depends on the manufacturing process

in principle, everything which is non-magnetic, not conducting, and mechanically stable enough to support the forces/torques could be a candidate

if your plastic can withstand the heat and you can manufacture it well, then it should be a great option I think

since I have no useful tools, I'm thinking about pertinax/pcb material, i.e., making an eagle design and sending it to OSHPark ... will probably require several iterations before the outline is exact enough, but on the other hand would be "easy"

one probably needs to use some somewhat larger motors than usual, on the other hand the motors being used with encoders are mostly "oversized" anyhow ...

the issue of friction would not be affected by that mod btw (I mention this because some motors I have clearly have a too high friction)
Sign up now
to remove ads between posts
Jul 04, 2019, 07:36 AM
Registered User
shuricus's Avatar
I can draw and 3D print parts from ABS plastic. It can stand heat up to 80 degree, I think.
To make PCB plates using CNC and glue them together is also excellent idea! And it's also possible but requires a bit more effort.

Do I need only make this part from plastic and add the same wires setup as at the picture?
Jul 04, 2019, 09:47 AM
OlliW
Thread OP
80 deg might be a bit tight, but for testing certainly more than OK
(the encodered version can draw quite some current and the motors can produce quite some torque, see one of my videos, and if that is done for too long the motor (wires) can get hot; should normally not happen though, and normally the motors should stay cold)

yes, just this part in plastic

if it should be the same wire setup would be a point to test. Since the resistances of the motors is usually more on the 10-Ohm-ish side, while with encoders 5-Ohm-ish should have advantages (at least in theory), one probably one might want to test also somewhat thicker wires (with the positive side effect that doing the winding would be easier). But as said, this would be really a point to test.

btw: this stator has a huge inner bore ... which to me suggests it's one of those motors with a wide hollow shaft ... in the (limited) experience I have with different motors I developed a feeling that it are exactly these motors which also have a relatively high friction ... I had actually a piece of text on this in my previous post where I was talking about friction, but killed it for brevity ... kind of funny you now show exactly such a motor ... the bottom line, it might be that the particular motor you show might be one of the high-friction ones ...

should be good enough for testing if there is a positive effect/reduced cogging though
Jul 04, 2019, 12:13 PM
Registered User
shuricus's Avatar
Great! I definitely should try then!
Only rewinding looks a little bit scare for the first time. It will take some time.

This is 2206 from hackrc gimbal. While rotating just bearings holder part by hand I can't feel much friction. I'll make video later.
Jul 04, 2019, 12:29 PM
OlliW
Thread OP
ah, it's a small motor ... I was somehow thinking it is at least 4xxx or so ... LOL

jo, winding is a thing by itself ... but with a little patience and practicing it really works out nicely ... as so often it's about the proper tools, I used a syringe with the sharp head chopped of ... worked well ... it also takes some practicing to get used to the winding scheme ... but if it would result in the best-ever gimbal motors...

how do you plan to mount the magnet? I do have these motors too but never got far since I couldn't figure out how I could mount the magnet with the tools available to me ... the diameter of the hollow shaft is quite odd valued ...
Jul 05, 2019, 04:18 AM
Registered User
shuricus's Avatar
I need to practice first with something already been broken. ))))

For small gimbals that I usually make for Gopro using encoders makes this construction to be too complicated for compact and simple DIY design. For bigger cameras T-Storm is absolutely amazing. But for Gopro I would really prefer to buy good small motors with encoder already installed, but this is more likely not going to happen as the market is too small.

I also was really interested to ask is there any special reason that NT FW version is not supporting of advanced horizon bank compensation from Ardupilot? I really miss this option so much! At the same time I'm not sure it's supported by Arduplane either.
Jul 05, 2019, 07:31 AM
OlliW
Thread OP
Quote:
Originally Posted by shuricus
I need to practice first with something already been broken. ))))
very good idea

Quote:
Originally Posted by shuricus
For small gimbals that I usually make for Gopro using encoders makes this construction to be too complicated for compact and simple DIY design. For bigger cameras T-Storm is absolutely amazing. But for Gopro I would really prefer to buy good small motors with encoder already installed, but this is more likely not going to happen as the market is too small.
yeah ... that's how things are now ...

there is what I think could be a very good candidate:
https://ae01.alicdn.com/kf/HTB1cu.VM...controller.jpg
http://brushlessgimbal.ca/images/GB2804EN-2.png
https://mad-motor.com/wp-content/upl...0/GB2804-1.jpg
and it once was easily available - but sadly not so anymore even though it appears to be still widely used in the feiyu gimbal (well, still available, but for a shocking price)

I really kick my ass that I didn't realized earlier that this motor appears to have ferrite magnet ring (like the disc motors), which implies much smaller cogging torques... and didn't at least buy one for testing

btw, even my two Mobius gimbals - i.e. smaller than gopro - are all-encoders since long ... it's undoubtedly more complicated for smaller gimbals, but with proper tools (lathe, cnc) not soo much, and the slimmer wiring adds benefit to exactly smaller gimbals, less frictions due to wire and so on.

I e.g. like funth1ngs': https://www.rcgroups.com/forums/show...postcount=1053

Quote:
Originally Posted by shuricus
I also was really interested to ask is there any special reason that NT FW version is not supporting of advanced horizon bank compensation from Ardupilot?
there is actually no real technical reason, could be done, but it's somewhat more complicated, and would also complicate the code and code maintenance quite a bit, and I also admit that NT isn't as high on my preference list anymore as encoders are, and this are the reasons for why I didn't add it, but just did it for TSTorM32

as said, even my two Mobius gimbals are all-encoders ...

Quote:
Originally Posted by shuricus
I really miss this option so much! At the same time I'm not sure it's supported by Arduplane either.
it is not in ArduPlane, but it should be really easy to carry it from ArduCopter to ArduPlane, but I never did myself since I do not have a plane, so it would have needed some testers. The situation has also become more complicated since ArduPilot's move to Chibios. I have not yet followed that move (mainly because of having to get along with the new build system which came with that move, and which still isn't really Windows-user friendly).

I TOTALLY AGREE, especially for plane it would be really useful ... I believe I have mentioned that repeatedly, and expressed my surprise that especially plane users are not asking more frequently and forcely for this feature ... (I mean, not pushing me but ArduPilot to add it )

btw, I also argue that especially plane users should want to use encoders and not NT. The encodered gimbal can fight much better against winds, and also withstands the rattling during take-offs & landings much better, which as I understand it often is a problem. Moreover, in some not too distance future there will be even a mode to put the gimbal into a kind of servo mode, which should also be interesting for plane flyers, since it allows to "freeze" the gimbal during e.g. takeoffs&landings, making it all much more reliable. So, ....
Jul 11, 2019, 01:14 PM
Registered User
shuricus's Avatar
Thank you for so many valuable info and for explanation! I really would love to switch to T-storm immediately. Unfortunately it's impossible for me to do it now with current gimbal design. funth1ngs's design is very cool, but it requires CNC which is not easy to find.

I hope I will find a solution for new design soon, and it will be really cool to have arduplane support, this will be the main reason for me to upgrade to T version. I will ask about this as frequently as you wish ))), just tell me where is the correct place to do this?
Jul 11, 2019, 01:39 PM
OlliW
Thread OP
yeah, that's always the challenge with TSTorM32, to find ways to build this dammed gimbal ... (I, e.g., still haven't mastered this)

the correct place to ask? Not here. Anywhere where there are ArduPilot devs, e.g. ArduPilot discourse or ardupilot github issue.
as frequently as I wish? It's not my cup of tea
if arduplane support is the main reason for switching to TSTorM32, I predict you will have still much time ...

I say that with a bleeding heart, but have to admit, STorM32 NT is cheaper and easier
Jul 16, 2019, 03:05 PM
Registered User
RocketMouse's Avatar
Quote:
Originally Posted by OlliW
@RocketMouse:
YOU want to test this!!
The servo mode is so-to-say replacing the xQN/xQM modes, in a much better fashion.
Hi Olli!
Please advise how to make motor to rotate continuous? Whatever settings I choose for RC Pitch Mode, motor only rotates to some degree and stops.
I use servo tester at the moment.
Thank you!
Jul 16, 2019, 03:27 PM
OlliW
Thread OP
use the yaw axis and set to relative turnaround
Jul 16, 2019, 03:50 PM
Registered User
RocketMouse's Avatar
Sorry, I thought they all have same functionality in servo mode. And how to adjust speed?
Jul 16, 2019, 04:07 PM
OlliW
Thread OP
this would have been a natural thought, I however aimed at reusing the parts which were already there, also to stay consistent.Thus yaw.
You should be able to adjust speed as usual by setting the speed limit.
Jul 16, 2019, 04:31 PM
Registered User
RocketMouse's Avatar
Could it be so that with xQN/xQM rotation speed was much higher? I might be wrong but I got this impression somehow, that's why I got confused - it looks like this new big 5208-200T motor rotates much slower than a small one I used before. Visually I simply didn't notice a difference when been changing Speed Limit for not too much. But it does work.
Jul 16, 2019, 04:37 PM
Registered User
RocketMouse's Avatar
It also looks like mosfets got too hot with this motor and 12V power supply, when Vmax is set to 150. But motor doesn't feel too strong. I also have a feeling that something already got burned ))))
Can the motor rotate if one of three mosfets is dead? It shouldn't I think. Just in case.


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion STorM32-NT BGC DIY Modules davidbitton Multirotor Drone Electronics 86 Dec 15, 2017 04:24 PM