Servo Holding Brake

 
#1

Might be biting off more then I can chew, but going to try and design a servo holding brake. Initial plan is the try and keep it 15mm thick by 25mm in dia.
When not powered will hold the servo in position.
Hope to 3D print most of it.
Ideas?
Comments?

User-inserted image

#2

My God. Please make this happen. I've been working on this for two years. I just had an arguement with a technician from Hitec who said he doesn't even care about the idea of holding brakes. They are compulsory. For now, I'm going to simply include torsion springs to fight gravity.

#3

By the way, I already know how to provide power to this release mechanism at the EXACT moment the servo motor gets a PWM signal. You wont need a separate PWM channel - you can run a power wire from the DC motor inside the servo. Whenever the motor is sent power to rotate, this will send power to anything else connected to it. The microprocessor inside the servo has done all the work for you Winky

Read some of my old posts. Solenoids are always too weak and bulky, by the way. I've tested hundreds of them. At least now your signal and power questions are all solved.

#4

The idea of a permanent magnet with positive polarity downwards against a coil of insulated copper around an iron bar with positive polarity facing upwards when power is applied seems viable if you can set up a nice grippy stopper. So by default, the permanent magnet is pulling the grip covered coil towards it, preventing rotation. When power is applied, the electromagnet is repelled. A spring can help with clamping also.

Looking at your diagram, your 'end plate' or something near there could be made of a permanent magnet.

#5

I will start on the CAD work tonight and hope to start some 3D printing this weekend. Ordered 28awg magnet wire this afternoon to wind the coil.

#6

The permanent magnet in the end plate also works like a magsafe connector in my designs to connect servos together.

#7

When power is applied, the armature is pulled down allowing the rotor to turn freely. No magnet needed. When power is removed, the springs push the up against the rotor pads and locks it in place.

#8

Can't wait to see it working.

#9

Hi @rz90208

Your idea is great !

If I may throw my thoughts in, If you can get a guaranteed full movement of the armature disc release, instead of a friction disc use a serrated disc. (3d printed? )(iron or metallic filament I think is available?) The teeth will hold better than friction allowing a lighter spring tension.

I have used this type of brake (industrial size) and found you need to release the brake first, even for a fraction of a second. Otherwise the motor tries to drive thru the brake causing high motor current.

I wish you luck with this idea. There is a need and it would be great to see it built.

Ron

PS. If you want, contact me at rtnron@att.net. I have an idea how to print
a semi-metallic disc with regular filament.

#10

Thanks Ron,
I have some carbon fiber filament I was thinking of using for the rotor but after some more thought, I decided not to because it might make the splined shaft collar not slide smoothly. So I now plan to make the end plate of the carbon fiber and the rotor of ABS, so I think the serrations would weaken it to much being that I only plan to make it 1.5mm thick (not counting the hub). For the armature I am hoping to find a steel washer that I can drill out and use. Not sure what to use for the friction material, I considered some type of sand paper but am concerned that it will wear out to quickly. Maybe putting the serrations in the armature (thinking as I type). First hurdle will be the drawing of the splines in Sketchup.

RichardZ