Author Topic: Differential screw  (Read 4254 times)

Offline sshire

  • Full Member
  • *****
  • Posts: 3044
    • LS Editions
Differential screw
« on: September 04, 2013, 09:42:35 PM »
Just got this from Village Press via email. New to me. VP encourages sharing the email so...

http://email.villagepress.com/pub/HSM/20130904/20130904.html
Best,
Stan

Offline smfr

  • Full Member
  • ****
  • Posts: 1195
  • San Francisco Bay Area, California
Re: Differential screw
« Reply #1 on: September 04, 2013, 11:01:25 PM »
That's neat!

Online b.lindsey

  • Global Moderator
  • Full Member
  • *****
  • Posts: 13695
  • Dallas, NC, USA
    • Workbench-Miniatures
Re: Differential screw
« Reply #2 on: September 05, 2013, 12:25:43 AM »
 I got it too Stan...thought it was interesting too.  Seems like i recall some discussion on here about a similar concept.

Bill
« Last Edit: September 05, 2013, 11:16:06 AM by b.lindsey »

Offline sshire

  • Full Member
  • *****
  • Posts: 3044
    • LS Editions
Re: Differential screw
« Reply #3 on: September 05, 2013, 02:48:26 AM »
I'm filling that away. Could be useful for something.
Best,
Stan

Arbalest

  • Guest
Re: Differential screw
« Reply #4 on: September 05, 2013, 10:03:02 AM »
Seems like I recall some discussion on here about a similar concept.

Bill

Yes there was Bill. I bought an item recently that had one on as well but I can't remember what it was, something DIY related though I think.

Offline Captain Jerry

  • Full Member
  • ****
  • Posts: 1026
  • Summerfield, FL USA
Re: Differential screw
« Reply #5 on: September 05, 2013, 12:29:23 PM »
Here is the discussion from this forum.  I mentioned my use of a differential screw to adjust the effective length of a reach rod and Alan Haisley chipped in with a more complete analysis of the subject.

http://www.modelenginemaker.com/index.php/topic,1389.msg18999.html#msg18999

Jerry
NOTARY SOJAK

There are things that you can do and some things you can't do. Don't worry about it. try it anyway.

Offline Noitoen

  • Full Member
  • ****
  • Posts: 103
  • Setúbal - Portugal
Re: Differential screw
« Reply #6 on: September 05, 2013, 12:38:47 PM »
Some "horizontal bar type" tap holders have the concept built in on the moveable jaw.

Online b.lindsey

  • Global Moderator
  • Full Member
  • *****
  • Posts: 13695
  • Dallas, NC, USA
    • Workbench-Miniatures
Re: Differential screw
« Reply #7 on: September 05, 2013, 12:41:03 PM »
Thanks Jerry... I knew I had seen it and thought it was connected to one of the Corliss threads, just couldn't get it to come around in the rolodex :)

Bill

Online sco

  • Full Member
  • ****
  • Posts: 1635
  • Location: Northants UK
Re: Differential screw
« Reply #8 on: September 05, 2013, 01:16:03 PM »
My mill uses this principal for the draw-bar - tightening the bolt draws the arbor in, slackening the bolt ejects the arbor from the morse taper without the need for a hammer.

Simon.
Ars longa, vita brevis.

Offline swilliams

  • Full Member
  • ****
  • Posts: 896
  • Canberra Australia
Re: Differential screw
« Reply #9 on: September 05, 2013, 01:23:27 PM »
Some "horizontal bar type" tap holders have the concept built in on the moveable jaw.

Indeed, but all the ones I've seen have a left and right hand thread to make the effective pitch larger rather than smaller. Of course the concept is still the same, just as you say.

Steve

Offline mklotz

  • Full Member
  • ****
  • Posts: 2404
  • LA, CA, USA
    • SOFTWARE FOR PEOPLE WHO BUILD THINGS!
Re: Differential screw
« Reply #10 on: September 05, 2013, 05:16:03 PM »
There's a program on my site (DIFFTHRD) for designing differential thread screws.  While calculating the effective pitch if the pitch of the two threads is known is straightforward, the inverse problem of finding the two threads needed to approximate a desired pitch is a bit trickier.  Then there's the problem of calculating the length of threads needed to get a desired motion.

Listed below are, respectively, the information page that accompanies the program and an example run.  Hopefully, one of you will find it helpful.

=================================================

   For very precise motion or, alternatively, a very high torque
advantage, very fine threads are required.  Cutting very fine threads is tricky
at best and the resulting thread tends to be very fragile.  A better solution
is to use a differential thread.

   What's that?  Imagine a cylinder that has one half threaded with a
coarse thread of pitch Pc (tpi - threads per inch).  The other half is threaded
with a fine thread Pf (tpi).  The coarsely threaded portion of the cylinder is
threaded through a matching nut that is fixed to some unmoving surface.  A
"movable" nut is threaded onto the fine threaded portion of the cylinder.  This
movable nut is free to move axially along the cylinder but is constrained from
rotating relative to the cylinder.  In other words, it can translate but not
rotate.  This movable nut is the "output end" of the device that serves to
position or move whatever the differential thread is "driving".

   Now, if we turn the cylinder through one revolution it will move
forward (relative to the fixed nut) by a distance 1/Pc.  The movable nut will
move in the opposite direction by an amount 1/Pf.  The net movement of the
movable nut relative to the fixed nut is then (1/Pc)-(1/Pf).  In other words,
the movable nut acts as if it were threaded on a shaft with an effective pitch
of:

      Pe = 1/((1/Pc)-(1/Pf))

   For example, with the relatively coarse values of:

      Pc = 20 tpi
      Pf = 25 tpi

we obtain a perfect micrometer thread:

   Pe = 1/((1/20) - (1/25)) = 1/(.05-.04) = 1/0.01 = 100 tpi

   An attractive application is a leveling screw for a precision level.
Here the fixed nut is a threaded hole in the level base.  The movable nut is a
threaded "foot" that is prevented from rotating by friction with the surface
on which the level sits.  Turning the threaded cylinder will tilt the base by
minute amounts determined by the geometry of the base and the placement of the
adjusting screw.  Obviously, the circumference of the threaded cylinder can be
divided into suitable subdivisions to achieve accurate partial rotations for
an even finer adjustment.

   For a given desired range of motion of the movable nut, the computation
of the various dimensions involved can be tricky.  DIFFTHRD does all the work
for you.

   Moreover, you may not have to hand the required screwcutting gear for
just any thread.  (25 tpi isn't easily available on a lathe or via a tap and
die.)  The ascii data file supplied (DIFFTHRD.DAT) allows you to specify what
threads you CAN cut (either via a screwcutting lathe or with hand threading
gear) and DIFFTHRD will decide what combination comes closest to producing the
effective pitch you desire.

   For a more compact geometry, the entire assembly can be "folded" into
itself.  Consider a bolt of pitch Pc threaded into a fixed nut.  Bore this
bolt out and thread the hole with a thread of pitch Pf.  A second bolt fitted
into this hole and prevented from rotating (say by a spring wire inserted into
a cross-drilled hole) will move with an effective pitch of Pe when the Pc bolt
is rotated.

Update 12/99:

   Don't know why I didn't originally include metric threads, since mixes
of Imperial and metric threads offer the possibility of coming closer to a
desired effective pitch.  You can now specify metric threads in the data file.
However, the program still operates in Imperial notation.  I'll change that
when someone complains loudly enough.





=================================================


DIFFERENTIAL THREAD CALCULATIONS


Number of data items read = 64

Desired effective pitch of differential thread [100 tpi] ?

Of available threads, best match to 100.000 tpi is:
Coarse thread = 22.000 tpi = 1.155 mm/thrd
Fine thread = 28.222 tpi = 0.900 mm/thrd
with an effective pitch of 99.786 tpi

Pitch of coarse thread [22 tpi] ?
Pitch of fine thread [28.222222222 tpi] ?
Thickness of coarse (fixed) nut [0.375 in] ?
Thickness of fine (movable) nut [0.25 in] ?
Desired motion of movable nut [0.25 in] ?

Effective pitch = 99.786 tpi
Motion for one revolution = 0.01002 in
Total turns to obtain desired motion = 24.946
Minimum length of coarse thread needed = 1.509 in
Minimum length of fine thread needed = 1.134 in
Maximum distance between nuts = 1.134 in
Minimum distance between nuts = 0.884 in
---
Regards, Marv


Home Shop Freeware
http://www.myvirtualnetwork.com/mklotz

Arbalest

  • Guest
Re: Differential screw
« Reply #11 on: September 05, 2013, 08:44:26 PM »
My mill uses this principal for the draw-bar - tightening the bolt draws the arbor in, slackening the bolt ejects the arbor from the morse taper without the need for a hammer.

Simon.

This is something I'd like to do on my Mill at some point, do you have any pics?