Author Topic: Trackball via Ball Transfer Units *3D Mock-up Made*  (Read 69891 times)

0 Members and 1 Guest are viewing this topic.

Offline Synnöve

  • Thread Starter
  • Posts: 26
Trackball via Ball Transfer Units *3D Mock-up Made*
« on: Thu, 12 June 2014, 03:11:19 »
Has anyone created a trackball that makes use of ball transfer bearings?

In just about every current trackball design the ball has to slide against a surface at some point in it's rotation. Stainless steel rollers have thus far been one of the better design solutions in reducing the amount of sliding against the ball, but some rotations go against one of the rollers causing an awkward pull on the ball for the duration of the direction.

http://www.omnitrack.us/mediumduty/plain-fitting-ball-unit-MG-m.php

Three of those would eliminate the sliding action in all directions, last a very long time, and produce a single low friction value for any direction of travel. They are essentially drop in and could easily be replaced in the event of a malfunction (at a low cost).

The design of this hypothetical trackball would have the following features:

-3 ball transfer units of the type that I linked to
-Aramith 57mm trackball
-3d Printed housing and buttons
-Electronics transplanted from another popular mouse with a superior sensor (a g400s or g502).
-Outer shape should be ergonomic, similar to the MSE or Marble FX. Should contain many buttons, DPI changer switches, and a place for a scroll wheel.

Thoughts?

« Last Edit: Tue, 08 July 2014, 04:48:41 by Synnöve »

Offline OverKill

  • Posts: 109
  • Location: Arizona, USA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #1 on: Thu, 12 June 2014, 08:58:45 »
I am not sure if you are proposing using these as a trackball or using these to hold the ball in the housing. either way i'll keep an eye on you ;)

Offline AKmalamute

  • HHKB Scrub
  • Posts: 837
  • Location: Western WA, USA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #2 on: Thu, 12 June 2014, 12:01:58 »
There aren't enough trackballs on the market.

Also, forefinger operated trackballs are evil. I feel sad that there are only two thumb-operated models on the market, and they're both mediocre.

Somebody said Logitech hates thumb-trackballs, and sics their lawyers on anyone who markets one. If true, the entire patent system and anything tangentially related needs to die a fiery, painful death.

HHKB-lite2, Dvorak user

Offline hanya

  • Posts: 132
  • Location: Japan
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #3 on: Thu, 12 June 2014, 12:53:04 »
Last year, I considered the same idea to use ball transfer bearings as supporter of the ball. But there is a problem about cleaning. The bearing could not opened to clean, it seems. The ball will get dirty by dust and dirt from the hands and they will be transferred to the supporter bearings.
PFU HHKB JP, Sanwa MA-TB38 trackball

Offline 0100010

  • Posts: 1127
  • Location: DFW, TX, US
  • Not Sure
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #4 on: Thu, 12 June 2014, 15:53:01 »
There aren't enough trackballs on the market.

Also, forefinger operated trackballs are evil. I feel sad that there are only two thumb-operated models on the market, and they're both mediocre.

Somebody said Logitech hates thumb-trackballs, and sics their lawyers on anyone who markets one. If true, the entire patent system and anything tangentially related needs to die a fiery, painful death.

Logitech Trackman Marble FX and Microsoft Trackball Explorer are both finger operated, and they get a lot of love.  I have a Marble FX on one KVM in my office, and a Kensington Orbit on my primary PC on the other side of my office. 

On the other hand - I can't stand thumb operated trackballs.

To each his own I suppose.  Sorry to hijack OP.
  Quoting me causes a posting error that you need to ignore.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #5 on: Thu, 12 June 2014, 16:12:44 »
How would you sense the ball’s movement, if it was encased in one of those?

Or you mean you’d rest the trackball ball on 3 of those ball transfer units? How big are they?

Anyway, sounds maybe promising. Try to get some parts and build a prototype? [I recommend just making a mechanical prototype first, before trying to figure out the sensors.]
« Last Edit: Thu, 12 June 2014, 16:17:33 by jacobolus »

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #6 on: Thu, 12 June 2014, 16:21:51 »
I am not sure if you are proposing using these as a trackball or using these to hold the ball in the housing. either way i'll keep an eye on you ;)

Three ball units would support a 57mm trackball.

Last year, I considered the same idea to use ball transfer bearings as supporter of the ball. But there is a problem about cleaning. The bearing could not opened to clean, it seems. The ball will get dirty by dust and dirt from the hands and they will be transferred to the supporter bearings.

The units I linked to have a drainage channel integrated, and as a result could be easily cleaned. I'll need to contact the company more however to ascertain certainty on this. That being said, the balls probably wouldn't pick up as much grime or grit as rollers or beads as they will never slide against the ball.

How would you sense the ball’s movement, if it was encased in one of those?

Or you mean you’d rest the trackball ball on 3 of those ball transfer units? How big are they?

Rest the trackball on three units with a laser or optical sensor underneath. The diameter of the main ball in the ball transfer unit is 9mm, housing is 19mm.

Concerning finger trackball vs thumb: unless the housing was rather large, or it was integrated in to a keyboard setup, I do not see how one could use the transfer units and keep the housing at a reasonable size as the units themselves are 19mm in diameter.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #7 on: Thu, 12 June 2014, 16:50:47 »
This company seems to have either steel main balls, or POM. I wonder which would work best. It might be an advantage to have a slightly grippy surface against the main ball.

Offline AKmalamute

  • HHKB Scrub
  • Posts: 837
  • Location: Western WA, USA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #8 on: Thu, 12 June 2014, 17:03:33 »
Concerning finger trackball vs thumb: [...] I do not see how one could use the transfer units

Sorry I think you lost me. Why does the size of the ball affect orientation? They're spherical, after all. How again, were you planning on using these so as to not touch the contact surface?

HHKB-lite2, Dvorak user

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #9 on: Thu, 12 June 2014, 18:08:43 »
^It's a matter of having enough space to mount the transfer bearings (units). While they are smaller than rollers (which are too bulky for a practical thumb design) they are somewhat larger than traditional "beads". As such, the housing would need to be large enough to accommodate at least one transfer bearing on the thumb side. I suppose it could be done however I'd assume the housing would be overly bulky. To re-iterate. Three of the transfer bearings/units would support one 57mm trackball. The advantage here is no sliding friction on the main trackball as the transfer bearings are omni-directional rolling balls.

Concerning the material, I'd opt for entirely stainless steel parts to reduce grime accumulation as well as for longevity.

I've just completed a few experiments using the optics of a MX518 and have concluded that the best sensor for this application would be laser as optical sensors seem very picky about the ball surface as well as it's orientation & distance to the trackball. I imagine a laser would be more flexible (not to mention the G502's laser module is very small and I could see it easily being mounted under a trackball).


Offline dorkvader

  • Posts: 6288
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #10 on: Thu, 12 June 2014, 18:22:08 »
There aren't enough trackballs on the market.

Also, forefinger operated trackballs are evil. I feel sad that there are only two thumb-operated models on the market, and they're both mediocre.

Somebody said Logitech hates thumb-trackballs, and sics their lawyers on anyone who markets one. If true, the entire patent system and anything tangentially related needs to die a fiery, painful death.

yeah, I use middle and ring fingers primarily.

I think this could work! I don't run into the issue too much, but my trackball doesn't get the third axis of rotation (twist) very well at all and of course it's not detectable by the sensor. I think it'd be great to have.

Some things I'd like to see in the trackball:
1. "stick up" a bit more from the surface. This also makes it easier to "twist"
2. having the option to actually detect twists would be great. You can use it for scrolling or whatever.
3. I think having a heavier ball would work out better here in a lower friction environment, especially with a high DPI sensor.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #11 on: Thu, 12 June 2014, 18:28:12 »
A very dense trackball on very low friction bearings would be great, if it could be done in a reasonably cheap way, without too much required maintenance of the ball or bearings. (By reasonably cheap I mean, the mechanical parts assembled DIY for, say, <$50 or at the outside <$100.)

Offline AKmalamute

  • HHKB Scrub
  • Posts: 837
  • Location: Western WA, USA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #12 on: Thu, 12 June 2014, 18:35:56 »
^It's a matter of having enough space to mount the transfer bearings (units).

Not really. If you can build a forefinger trackball, you can tip it 40-odd degrees, mount it farther back, and use your thumb, allowing (IMHO) more natural forefinger-clicking. It isn't just the oddity of the motion that's bad, thumb buttons drive me up a wall.

Best solution, and I hated it, was to mark the input device as a left-handed model, and smack the primary button with the ring finger via a wrist motion.

I reiterate, forefinger trackballs are evil. If your hand tells you they're preferable there are many things about your life I might question, but whether I'm interested in your pointing device isn't one of them.

I wish you luck in your device of evil, just the same. What could be cool, though, is if you could post enough, like sensor firmware sourcecode, that were a different housing manufactured, someone like me who can't code his way out of a paper bag, could consider building a durable thumb trackball.

Its still going to get gummed up and need periodic cleaning, though. No way around that unless you add a high-pressure oil scraper ring so the part of the human-operated ball touching the ball bearings is always pristine.

Edit: I believe the 1st-gen trackman (Logitech) had ball bearings, and I remember having to stop and clean them out every couple of hours. Although Logitech might have stopped using them for cost reasons, like everything else, rather than maintenance complaint reasons.
« Last Edit: Thu, 12 June 2014, 18:52:32 by AKmalamute »

HHKB-lite2, Dvorak user

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #13 on: Thu, 12 June 2014, 19:00:24 »
Some things I'd like to see in the trackball:
1. "stick up" a bit more from the surface. This also makes it easier to "twist"
2. having the option to actually detect twists would be great. You can use it for scrolling or whatever.
3. I think having a heavier ball would work out better here in a lower friction environment, especially with a high DPI sensor.

1: Agreed. This is one reason I'm trying to find a DT225 to buy instead of a CST!
2: I agree with this, but it does present some new challenges*.
3: I imagine this trackball using a super arimath billiard ball for this reason.

A very dense trackball on very low friction bearings would be great, if it could be done in a reasonably cheap way, without too much required maintenance of the ball or bearings. (By reasonably cheap I mean, the mechanical parts assembled DIY for, say, <$50 or at the outside <$100.)
With out a large group buy this trackbal wouldn't be on the cheap side:

-30 dollars for the bearings
-17 dollars for the billiard ball
-30 to 80 dollars for a laser mouse*
-Who knows how much for a 3D printed housing

I wish you luck in your device of evil, just the same. What could be cool, though, is if you could post enough, like sensor firmware sourcecode, that were a different housing manufactured, someone like me who can't code his way out of a paper bag, could consider building a durable thumb trackball.

Its still going to get gummed up and need periodic cleaning, though. No way around that unless you add a high-pressure oil scraper ring so the part of the human-operated ball touching the ball bearings is always pristine.

No one is perfect, unfortunately!

I know zero code. Most of my knowledge is book-smarts and theory, with dabbling in graphic arts. As long as the sensor is positioned the same and the buttons are wired appropriately a different firmware shouldn't be needed for a thumb ball design.

Concerning cleaning, I emailed the bearing manufacturer for information concerning the potential for the bearings to gum up, if they are able to be cleaned, etc.


*If we were to use the guts of a standard laser mouse I do not think it'd be possible to register twisting as a scroll movement or what have you. The only way this'd be possible would be to create our own controller. To my limited knowledge, this'd involve the following (Again, prices are non-group buy):

-Purchasing this for use as the laser sensor: https://www.tindie.com/products/jkicklighter/adns-9800-optical-laser-sensor/ (30 USD)
-Purchasing a Teensy USB controller (20 USD)
-Scavenging a cheap mouse for switches & wire.
-Finding an individual(s) who would be willing to write custom code for this that'd work on linux, windows, and mac (I use Windows) either for free or for a modest fee (the latter option would require a kickstarter I'd wager).

If we chose this path, the cost for parts minus the cost for a 3D printed housing and code would probably be around 90 to 100 USD (non group-buy prices).


Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #14 on: Thu, 12 June 2014, 19:19:55 »
With out a large group buy this trackbal wouldn't be on the cheap side:

-30 dollars for the bearings
-17 dollars for the billiard ball
-30 to 80 dollars for a laser mouse*
-Who knows how much for a 3D printed housing

Well, I can cut a wood housing on a CNC router, or 3d print one myself, and I have a couple of Teensy 3.0s and a couple of laser mouse sensors already. https://www.tindie.com/products/jkicklighter/adns-9800-optical-laser-sensor/

So all I need is a trackball and some kind of bearings to hold it, and the actual mechanical design work. If you try these bearings and they work well, I’d love to hear about it and try to build one myself. My initial plan is to just make the standard little 3 static bearing type of trackball, and maybe harvest an initial ball from a Logitech or something.

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #15 on: Thu, 12 June 2014, 19:48:58 »
If you try these bearings and they work well, I’d love to hear about it and try to build one myself. My initial plan is to just make the standard little 3 static bearing type of trackball, and maybe harvest an initial ball from a Logitech or something.

Sounds like a good idea. I'll wait to hear back from the bearing company about the unit's resistance to grime/clean-ability; if the reply is positive I'll purchase some units and see how they hold up to constant exposure to human skin oils. If they pass that test then we could move on to designing a 3d-printable housing and seeing where all the parts could fit.

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #16 on: Fri, 13 June 2014, 15:52:32 »
So here is what I've discovered by chatting with Omnitrack:

"In terms of contamination, the MG8A cannot be disassembled but does have a small drain hole in the base of the unit. We would strongly advise against trying to flush the units clean in a liquid bath as the body only has a zinc plated finish which is unlikely to resist the corrosive nature of washdown situations. I expect hand oils shouldn't pose an issue to the ball units (assuming the oil would only coat the balls in a fine film, rather than dowse them) but the build up of skin debris could increase the friction in the unit over time.

It is difficult to predict how long the units would last before they become noticeably affected by contamination - it would largely depend on the duration & frequency of use as well as the cleanliness of the user's hand."

So it sounds like the MG8A might be a riskier choice, although I do think it wouldn't gum up very easily due to the low amount of contact they'll have with the trackball. There are a few other options: they do have a 22mm diameter bearing which has a felt seal to prevent debris buildup for around 15 dollars, and they also have a 19mm ball that has larger drainage holes/is more corrosion resistant, and user serviceable for around 24 dollars.

I've written again to omnitrack asking if they have a unit that would better suit our needs. I guess we'll see, based on their feedback, if this is viable or not.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #17 on: Fri, 13 June 2014, 17:00:35 »
Thanks for following up about this. Keep us posted if they have any other advice.

Offline OverKill

  • Posts: 109
  • Location: Arizona, USA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #18 on: Fri, 13 June 2014, 17:03:22 »
Has anyone created a trackball that makes use of ball transfer bearings?

In just about every current trackball design the ball has to slide against a surface at some point in it's rotation. Stainless steel rollers have thus far been one of the better design solutions in reducing the amount of sliding against the ball, but some rotations go against one of the rollers causing an awkward pull on the ball for the duration of the direction.

http://www.omnitrack.us/mediumduty/plain-fitting-ball-unit-MG-m.php

Three of those would eliminate the sliding action in all directions, last a very long time, and produce a single low friction value for any direction of travel. They are essentially drop in and could easily be replaced in the event of a malfunction (at a low cost).

The design of this hypothetical trackball would have the following features:

-3 ball transfer units of the type that I linked to
-Aramith 57mm trackball
-3d Printed housing and buttons
-Electronics transplanted from another popular mouse with a superior sensor (a g400s or g502).
-Outer shape should be ergonomic, similar to the MSE or Marble FX. Should contain many buttons, DPI changer switches, and a place for a scroll wheel.

Thoughts?

After a bit of research I found a few smaller version.

4MM
http://www.alwayse.com/en/products/mini_ball_transfer_units.html

3MM
http://www.holo-pack.com.tw/en/profile.jspx?no=A06001

Not sure where to buy them or anything but they are more a size that we would probably want. I also looked up a few videos on youtube from like VXB bearing and such and they all seem to run fairly frictionless so the balls being too tight in the housing might not really be an issue for us.

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #19 on: Fri, 13 June 2014, 17:46:03 »
I'll send an email to Alwayse as their stainless models might work. With the smaller bearing housing I'd wager dirt and debris would be even less likely to enter the case.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #20 on: Sat, 14 June 2014, 22:03:29 »
4MM
http://www.alwayse.com/en/products/mini_ball_transfer_units.html
11-MI-05-17 looks like an interesting one there, with an 8mm housing diameter and 8.5mm total height including the 2.5mm long M2 bolt. I don’t think the aluminum housing or 5 kg max load would be any problem for this kind of use case. I wonder how smoothly these balls spin.

I wonder what these cost, buying like 3–50 units at a time.

3MM http://www.holo-pack.com.tw/en/profile.jspx?no=A06001
Having the main ball made out of delrin instead of steel might nice too. I wonder what difference there would be practically, for trackballs of different sizes/materials.

* * *

I’d be really curious to try to find some slightly grippy material for the main trackball ball (but ideally not something that attracted dirt/grime), assuming it could be found as a fairly dense, uniform, precisely spherical ball (or it would also work to e.g. have a metal ball coated in something slightly grippy). As long as the ball transfer units spin without much friction, increased traction against fingertips would be I think a big advantage.
« Last Edit: Sat, 14 June 2014, 22:14:57 by jacobolus »

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #21 on: Sat, 14 June 2014, 22:12:51 »
[whoops]

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #22 on: Sun, 15 June 2014, 05:48:54 »
I wonder how smoothly these balls spin... I wonder what these cost, buying like 3–50 units at a time.
They don't list a friction coefficient but if I can extrapolate from another manufacturer's material, I'd say it'd be somewhat lower than Teflon, and that's including all three bearings. I'll report back pricing after I've communicated with the OEM.

Having the main ball made out of delrin instead of steel might nice too. I wonder what difference there would be practically, for trackballs of different sizes/materials.
I'd say stainless steel would be the best option for this due to longevity (and the price isn't too different it seems).

I’d be really curious to try to find some slightly grippy material for the main trackball ball (but ideally not something that attracted dirt/grime), assuming it could be found as a fairly dense, uniform, precisely spherical ball (or it would also work to e.g. have a metal ball coated in something slightly grippy). As long as the ball transfer units spin without much friction, increased traction against fingertips would be I think a big advantage.
That's a really really good idea. Since there is no sliding a semi grip surface is now a possibility. The surface needs to be able to have a moderate change from grippy to slick depending on speed... as a thought, the surface of my MX518 would be perfect actually, or the surface of my Nexus 5. They are made out of special types of plastic however, which I feel would drive the cost up quite a bit. I think sticking to an aramith ball for now might be the best option as any coatings we could apply would probably not be perfectly spherical enough.

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Hypothetical Trackball via Ball Transfer Units
« Reply #23 on: Tue, 17 June 2014, 17:27:57 »
Was able to get in touch with Alwayse in the UK about use of their transfers. They took interest in the idea and, should the shipping not be prohibitive, are willing to send several units to me for testing. They would also like us to document our process as they do not always get to see first hand the results and experiments with their products due to dealing with suppliers mostly.

They also had some advise concerning implementation:

-They gave friction of the transfers as a percentage of the weight load at 1 to 1.5 percent, which equals 0.5 grams per transfer unit given a 170 gram billiard ball.
-They advised to not use a felt sealer as not only would it be difficult to install in a transfer that small but it would also increase friction.
-The units should be angled so that they are facing the center of the trackball as if they simply face straight up there is a potential for the balls in the transfers to jump.

They mentioned they'd like to see a mock-up of how we plan to implement the transfers at some point, which means a design should be started. Anyone have CAD skills?

IMHO, the housing should be quite similar to a Trackman Marble FX,  but with a 57mm ball (thus a larger housing), more buttons, detachable cable, DPI switches. The opening on the left hand side for the thumb to grip the trackball would be perfect for creating a twist-to-scroll operation (though I'd suggest expanding the opening a bit length-wise). I believe this design will produce the best ergonomics for a large trackball.

If transfers turn out to work well in trackballs would y'all be interested in starting group buys/kickstarters to gather parts and fund production of housing and software design?

Offline CPTBadAss

  • Woke up like this
  • Posts: 14363
    • Tactile Zine
Re: Hypothetical Trackball via Ball Transfer Units *Reply from Alwayse*
« Reply #24 on: Tue, 17 June 2014, 17:41:24 »
I would be interested in a GB. But a starting point shouldn't be CAD. Hand sketches first :D.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothetical Trackball via Ball Transfer Units *Reply from Alwayse*
« Reply #25 on: Wed, 18 June 2014, 02:47:29 »
If you can figure out what their pricing is like for small numbers of units, I’d also like to play with some, and will potentially be willing to help with design, code, etc. (Or if they sent you like 10 units and you want to pass 3 along :-)

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothetical Trackball via Ball Transfer Units *Reply from Alwayse*
« Reply #26 on: Wed, 18 June 2014, 03:16:07 »
By the way, the guys who make those Aramith balls might have better suggestions about trackball balls:
http://www.preciball.com/index.php/en/links.html

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Hypothetical Trackball via Ball Transfer Units *Reply from Alwayse*
« Reply #27 on: Wed, 18 June 2014, 18:18:08 »
If they decide to send me six or more, I'll be sure to relay three of them to you for experimentation.

If this project gets to place where we can successfully design and put together this trackball on a reasonable budget, I'll contact the ball makers for advise.

Offline dorkvader

  • Posts: 6288
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Hypothetical Trackball via Ball Transfer Units *Reply from Alwayse*
« Reply #28 on: Fri, 20 June 2014, 15:40:40 »
By the way, the guys who make those Aramith balls might have better suggestions about trackball balls:
http://www.preciball.com/index.php/en/links.html

I'm talking to a chinese manufacturer to see if I can get a small run of ceramic balls made.

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Hypothetical Trackball via Ball Transfer Units *Reply from Alwayse*
« Reply #29 on: Sat, 21 June 2014, 02:18:16 »
A ceramic ball?! Won't that be prohibitively expensive? Are you planning to have the size match that of a standard billiard ball?

Alwayse has not reported any developments to me concerning sending sample units for testing. I'm wondering if their silence implies they are no longer interested in the situation. I'll report back if anything changes.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Hypothetical Trackball via Ball Transfer Units *Reply from Alwayse*
« Reply #30 on: Sat, 21 June 2014, 04:01:57 »
If they don’t reply, there’s always the ebay route. E.g. http://www.ebay.com/itm/331217858644

Offline damorgue

  • Posts: 1176
  • Location: Sweden
    • Personal portfolio
Re: Hypothetical Trackball via Ball Transfer Units *Reply from Alwayse*
« Reply #31 on: Sat, 21 June 2014, 07:49:31 »
Remember to keep both the shell dimensions and thickness small to limit material and build size.
Acetone can make parts printed in ABS really smooth.

Good luck, I will be following the development of this.

Also, why does everyone seem to favor freakishly heavy balls? Past a certain point, I don't think they become smoother, just more cumbersome.

Offline dorkvader

  • Posts: 6288
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Hypothetical Trackball via Ball Transfer Units *Reply from Alwayse*
« Reply #32 on: Sun, 22 June 2014, 17:14:19 »
A ceramic ball?! Won't that be prohibitively expensive? Are you planning to have the size match that of a standard billiard ball?

Alwayse has not reported any developments to me concerning sending sample units for testing. I'm wondering if their silence implies they are no longer interested in the situation. I'll report back if anything changes.
most places don't make them in 2.125", but you can get 50mm (2") pretty easily. One place goes up to 70mm!! O_O

As to the expense, I've politely asked for some samples.

Offline Synnöve

  • Thread Starter
  • Posts: 26
^Well if you're getting samples be sure to get a few as I could certainly use one for testing! : D

On that note, I have good news! Alwayse has confirmed that they are sending me 6 miniature transfer units for testing. Jacobolus, once I receive them I can make arrangements to post you three of the units to facilitate your design work.

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #34 on: Thu, 26 June 2014, 15:24:21 »
I just received the bearings today, and after some very rudimentary tests with sub-optimal placement/angling of the units, I have to say they are incredibly smooth! By far the smoothest trackball I've used with little resistance in any direction, so much so that little finger pressure is required for the trackball to spin several revolutions before stopping. Finally getting my hands on these things has lifted my hopes for the feasibility of this project.

There is one, very important caveat though: they are noisy, think skateboard noisy. Now it's important to keep in mind that they are not in a housing and the ball I'm using isn't very dense, but I think it will be similar to the Model M: loud but probably the best switch in terms of feel and durability.

I'm going to purchase some shaping clay to experiment with housing shapes and optimal ball placement. Will post updates when they happen.
« Last Edit: Thu, 26 June 2014, 15:38:33 by Synnöve »

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #35 on: Thu, 26 June 2014, 15:31:39 »
Huh. Why so loud? It seems like noise would loosely correlate to friction.

Which size/model are the ones you got? And do you know the bulk price on them?

(Also, if you’re still willing to send me 3, I’m happy to paypal you the shipping cost; shoot me a PM with your paypal email addy.)

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #36 on: Thu, 26 June 2014, 15:48:37 »
The balls do have a bit of friction (else the trackball would spin forever) so that's to be expected. The noise, I imagine, probably comes from all the little support bearings inside the units, as well as it having a metal casing which would conduct vibration/sound quite easily to the ball. As mentioned, I do not have them mounted in a sturdy way (they are simply pushed in to cardboard at the moment) and I'm using a lightweight trackball, which probably contribute somewhat to the noise factor.

Alwayse 11-MI-05-17. I'm not sure what the prices are directly from Alwayse, or if they have a minimum quantity threshold, but I did find this distributer: http://uk.rs-online.com/web/p/ball-transfer-units/7431408/ .

I'd be glad to send you three for testing.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #37 on: Thu, 26 June 2014, 19:49:11 »
Here are some distributors of Alwayse stuff (according to the Alwayse site) with branches in California. I’ll try poking around their sites / calling them sometime:
http://www.applied.com
http://www.bdi-usa.com
http://www.bearingengineering.com
https://ec.kamandirect.com/us/index.jsp
https://www.motionindustries.com/motion3/jsp/mii/productCatalogSearch.jsp
http://www.powerindustries.com

« Last Edit: Thu, 26 June 2014, 19:54:31 by jacobolus »

Offline hanya

  • Posts: 132
  • Location: Japan
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #38 on: Fri, 27 June 2014, 03:20:25 »
I can buy size compatible one with Alwayse 11-MI-05-17 from IGUCHI KIKO in Japan, model IS-05SNM [1] at $13 each in many shops.

If you want to detect twisting the ball, it seems two sensors are required like the following. So, larger ball transfer unit can not be placed like it. Another way to detect the twisting is to place a sensor under the ball and to place another sensor to the side of the ball. But I do not want to put the sensor under the ball because of the total height increasing.

The optical center of the sensors are at 35 degrees from horizontal plane. I wonder the twisting can be detected at this angle.




With the model above, local 3D printing shop tells me the price is about $50 with acrylic or ABS kind materials, size about 60 mm x 60 mm x 23 mm.

[1] http://www.isb-iguchi.com/e012_03is_n.html
PFU HHKB JP, Sanwa MA-TB38 trackball

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #39 on: Fri, 27 June 2014, 04:07:32 »
That is some excellent work! And yes those transfers seem to be clones of Alwayse's.

It seems intuitive to me that as long as there are two sensors that are at right angles to eachother, you can place them almost anywhere you want and capture the entirety of the ball's rotation. The only things we need to be conscious of is the distance from the ball (which would be dictated by the sensor's range) and the amount of bulk.

Is the enclosure designed for a 57.2mm ball? Did they tell you how much bulk purchase would be? of that enclosure?

You might want to design some sturdy mounting studs for the container, that way this system could be used with just about any outer housing shape/design as long as the housing has the holes for mounting the container.


Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #40 on: Fri, 27 June 2014, 04:14:56 »
Actually the two sensors can be placed and oriented almost anywhere on the ball and you can capture full 3d rotation. For optimal accuracy you probably want them in some specific alignment though. This paper suggests that you get better numerical stability if the two sensors are at 90° to each-other http://3map.snu.ac.kr/courses/2001/cg/lecture/trackball.ps but I think their analysis seems a bit simplistic.

Offline dorkvader

  • Posts: 6288
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #41 on: Fri, 27 June 2014, 12:52:03 »
Actually the two sensors can be placed and oriented almost anywhere on the ball and you can capture full 3d rotation. For optimal accuracy you probably want them in some specific alignment though. This paper suggests that you get better numerical stability if the two sensors are at 90° to each-other http://3map.snu.ac.kr/courses/2001/cg/lecture/trackball.ps but I think their analysis seems a bit simplistic.
I think we should "throw away" any unneeded data and forget about added accuracy. Those sensors are good enough for me!
 



The more I look at this, the more I really like it! even the price isn't bad!

Offline hanya

  • Posts: 132
  • Location: Japan
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #42 on: Sat, 28 June 2014, 08:17:13 »
I put data files for testing only at: https://github.com/hanya/btutb

Is the enclosure designed for a 57.2mm ball? Did they tell you how much bulk purchase would be? of that enclosure?
Its for 2.25" ball. They have no discount for bulk purchase.

Quote
You might want to design some sturdy mounting studs for the container, that way this system could be used with just about any outer housing shape/design as long as the housing has the holes for mounting the container.
Good idea.

I uses trackball in left hand, so I would make left handed one for mine.
PFU HHKB JP, Sanwa MA-TB38 trackball

Offline dorkvader

  • Posts: 6288
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #43 on: Sun, 29 June 2014, 10:25:09 »
I put data files for testing only at: https://github.com/hanya/btutb

Is the enclosure designed for a 57.2mm ball? Did they tell you how much bulk purchase would be? of that enclosure?
Its for 2.25" ball. They have no discount for bulk purchase.

Quote
You might want to design some sturdy mounting studs for the container, that way this system could be used with just about any outer housing shape/design as long as the housing has the holes for mounting the container.
Good idea.

I uses trackball in left hand, so I would make left handed one for mine.

2.25" is a very uncommon trackball size: I don't know of any that use it. DO you know where to source balls in that size for it?

More common trackball sizes are 2.125" or 2" (which are usually a drop-in compatible with 54mm and 50mm respectively).

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #44 on: Sun, 29 June 2014, 11:32:39 »
What?? 2 1/4th is a very common size. The older Kensingtons, the CSTs, the dt225...

It's the same size as most billiard balls so it's easy to find quality replacements at decent prices.

Offline dorkvader

  • Posts: 6288
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #45 on: Sun, 29 June 2014, 22:52:49 »
What?? 2 1/4th is a very common size. The older Kensingtons, the CSTs, the dt225...

It's the same size as most billiard balls so it's easy to find quality replacements at decent prices.
Just re-checked: my mistake.

Fortunately I have a plethora of trackballs in that size.

Thanks very much for the correction :)

Offline do_Og@n

  • * Maker
  • Posts: 988
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #46 on: Sun, 29 June 2014, 23:24:32 »
@Synnöve - I have two older (off white) CST Trackballs if you would like them for testing and building your model. PM me if you are interested.

Offline bueller

  • MX baller
  • * Esteemed Elder
  • Posts: 3769
  • Location: Perth, Australia
  • Church of the Ergo Clear
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #47 on: Sun, 29 June 2014, 23:26:11 »
Watching this closely, love older trackballs but all the current models on the market are uncomfortable for me to use.
It's a good width!  If it's half-width it's too narrow, and full-width is too wide. 

[WTT] bueller's trade thread - CLACKS WANTED

Offline Synnöve

  • Thread Starter
  • Posts: 26
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #48 on: Sun, 29 June 2014, 23:38:08 »
Watching this closely, love older trackballs but all the current models on the market are uncomfortable for me to use.
What is uncomfortable about them?

@Synnöve - I have two older (off white) CST Trackballs if you would like them for testing and building your model. PM me if you are interested.
Will do!

Offline dorkvader

  • Posts: 6288
  • Location: Boston area
  • all about the "hack" in "geekhack"
Re: Trackball via Ball Transfer Units *Preliminary Test Results*
« Reply #49 on: Mon, 30 June 2014, 00:50:22 »
Watching this closely, love older trackballs but all the current models on the market are uncomfortable for me to use.
What is uncomfortable about them?

@Synnöve - I have two older (off white) CST Trackballs if you would like them for testing and building your model. PM me if you are interested.
Will do!

for me at least its those awful bearings they use that are full of friction and hate.

This should fix that nicely.