Author Topic: Friction ball drive - for 'gear' reduction  (Read 331 times)

Offline sorveltaja

  • Sr. Member
  • ****
  • Posts: 489
Friction ball drive - for 'gear' reduction
« on: September 13, 2020, 05:41:45 PM »
This is an attempt to find out, if it's possible to make a friction ball reducing drive, using printed parts and bearing balls.

The need for such thing emerged, when I was testing several vocoder settings, and noticed, that some of them have quite narrow 'sweet spot' -ranges.

I thought of buying multiturn potentiometers, but as they cost at least five times more, than ordinary pots, it was time to look for another options.

Concept is simple; to make the reduction system as small as possible, by using a 'planetary gear set' -configuration, where the centerpiece is 'sun', rotating balls are 'planets', and outer ring is, well, outer ring.

At first, I did some tests using printed gears to achieve the reduction. It kind of worked, but was rather clunky and rough in practice. So, on to the next option.

Idea for using the bearing balls, as 'planet gears', came from the pics found on the net, and also from a video on Youtube(thanks to Joules for pointing that out).

To start drawing, this was used as a reference:



A simplified cross-section, for calculating the reduction ratio. In this case, it's 17.66/6.34 +1 = ~3.78(assuming the formula is right one for that). The ball size is 8mm:



Variation of the above concept:



Bit of development:



And finally, the working model:



Being static(non rotating) part, the outer ring extends all the way to the bottom, where the M7 thread is, for the potentiometer.

I would have never expect that much friction between two smooth metal surfaces. Suitable tension is easy to find by testing with different size drills/metal rods.

Diameter of the drill in the above pic is 2.3mm, and balls 10mm. Outer diameter, and height are both about 25mm.

Calculated reduction ratio is 10.26, while in practice it's almost 11.

Some observations of the current concept so far:

+ smooth action, once both ball mating surfaces are smooth
+ no need for separate tension spring
+ instead of being conical, the middle part, 'sun', is a straight steel rod
+ printed outer ring is flexible enough to provide friction between moving parts, or then it could be machined out of hard rubber, like that used in hockey pucks(I think I'm going to try that one also).
+ can be attached straigth to the potentiometers thread

- exact reduction ratios might not be that easily achieved, but could be possible by tweaking the parts.
- not sure, how well the printed outer ring can stand the continuous tension
- outer ring ball mating surfaces, if printed, needs to be smoothened, by machining or by some kind of sanding fixture

Further things to do/add/test:

Pointer, that shows the actual position of the pot's wiper.
To secure, that the center metal rod stays in place. Perhaps by machining a tiny v-groove to the ball mating point. But we'll see.
To make scaled down versions, to see how small it can get, to be practical.

 


Offline WeldingRod

  • Sr. Member
  • ****
  • Posts: 300
Re: Friction ball drive - for 'gear' reduction
« Reply #1 on: September 13, 2020, 06:29:31 PM »
Plastic, with its low elastic modulus, is your friend on the outer race!  The race is the spring!

Sent from my SM-G892A using Tapatalk


Offline sorveltaja

  • Sr. Member
  • ****
  • Posts: 489
Re: Friction ball drive - for 'gear' reduction
« Reply #2 on: September 14, 2020, 11:48:11 AM »
Yeah, didn't even think about it until I poked a drill bit in the center.

The smaller version works just like the bigger one:



It uses 4mm bearing balls, and 1mm drill. Outer diameter is 15mm and height is 17mm. There is no room for a grub screw, so the potentiometer shaft needs to be filed to have a flat surface to suit:



What comes to outer ring material, pla could be good enough, if there aren't heat waves or such high temperatures, that turn it into cheese.

Offline pycoed

  • Jr. Member
  • **
  • Posts: 30
Re: Friction ball drive - for 'gear' reduction
« Reply #3 on: September 15, 2020, 04:48:24 AM »
Are there pictures attached to this? If so I can't see them

Offline PekkaNF

  • Hero Member
  • *****
  • Posts: 2439
  • Country: fi
Re: Friction ball drive - for 'gear' reduction
« Reply #4 on: September 15, 2020, 07:58:43 AM »
I can see the pictures. They are on external server. Maybe overzealous firewall or settings? Or big bad company enforced policy?

Offline sorveltaja

  • Sr. Member
  • ****
  • Posts: 489
Re: Friction ball drive - for 'gear' reduction
« Reply #5 on: September 15, 2020, 08:18:03 AM »
I use Firefox, and tried with Opera also, and with both, I can see them. They are hosted in my local internet provider's server.

Offline pycoed

  • Jr. Member
  • **
  • Posts: 30
Re: Friction ball drive - for 'gear' reduction
« Reply #6 on: September 15, 2020, 02:58:18 PM »
 :scratch: :scratch: :scratch:
I can see them now! :scratch:
I'm on my laptop, but they weren't showing on my desktop this morning. Both running Ubuntu 20.04 & Chromium. Time for some investigations, I think...

Offline sorveltaja

  • Sr. Member
  • ****
  • Posts: 489
Re: Friction ball drive - for 'gear' reduction
« Reply #7 on: September 18, 2020, 09:46:44 PM »
As the concept works with printed parts, how about machining the outer race using other material:



I chose the hockey puck -rubber, because it's somewhat flexible, easy to machine, and cheap. When inserting the bearing balls, it needs to budge just a bit.

On the other hand, the 'carrier', that transmits the ball's rotation to the potentiometer shaft, will still be a printed one. Besides, it's not a part of the 'tension network'(consisting center metal rod, balls, and the outer race), and should retain its shape in use.

The balls seem to need at least some kind of groove in the outer race, to keep them in their rotating orbit. As I don't have too many boring tools, I'm thinking of using a 4mm ball end milling bit to machine the groove for the 8mm balls:   



Most(if not all) of the pots, that I've used in recent projects, have M7 x 1 threading, so that size tap is also needed, if one chooses to machine the outer race.

In the end, the center metal rod, 'sun', might not need any kind of groove in it, as the tension itself is enough to keep it in the place. In fact, when testing with printed parts, I had to pry it out.