What's new
What's new

OT: SIP rotary table index adapter etc.

Billtodd

Titanium
A while ago I picked up a nice SIP P1-2 20mm tilting rotary table. This is an opportunity to thank Chuck of this parish for sending me a copy of the manual and to show you my indexing adaption.

I had a couple of double sided index plates and a set of sector arms so while under house arrest for some thing I'm supposed to have done to Pam Demick (I swear I don't know the woman!), I thought I'd fit them to the RoTab.

There's no where to mount an adapter other than into the dial shield casting , so I made the adapter a slip fit and registered it against the fiducial quadrant - which isn't the 90 degrees that I assumed it was! hence the bronze shim/clamp (it adjusts out with a tapered grub screw from behind the plates.)

The handle was a major pain , since all I had to make it out of was some unknown stainless steel. When first assembled - it gauled and seized after a couple of turns , so i had to add a bunch of fiddle-arse bronze bearings .
 

Attachments

  • sip rotab and indexing adapter handle.jpg
    sip rotab and indexing adapter handle.jpg
    80.7 KB · Views: 122
  • sip rotab  existing dial fitting .jpg
    sip rotab existing dial fitting .jpg
    88.9 KB · Views: 104
  • sip rotab adapter back.jpg
    sip rotab adapter back.jpg
    91.2 KB · Views: 105
  • sip rotab and indexing adapter 1.jpg
    sip rotab and indexing adapter 1.jpg
    92.2 KB · Views: 116
  • sip rotab and indexing adapter 2.jpg
    sip rotab and indexing adapter 2.jpg
    89.5 KB · Views: 115
Next job is to make a setting pin (ball) , but in the mean time I designed this copy of the SIP 'dial indicator' for centring and locating the table.

[edit] forgot to add...
Does anyone here have a real SIP one? if so is the top centre (the bit the spindle centre fits into) sprung loaded?
 

Attachments

  • sip centring indicator sectioned.jpg
    sip centring indicator sectioned.jpg
    48.6 KB · Views: 83
  • sip centring indicator a.jpg
    sip centring indicator a.jpg
    56.6 KB · Views: 80
Nice! How true does the ball run to the table axis?

I rough turned the shaft then , drilled and reamed the stainless ball in a 20mm dead length collet before bonding it onto the shaft with loctite 638. Then, with the ball end in the collet , turned the shaft between centres

Pretty good . I was having trouble trying to hold my 1um Mahr millimes dti still enough , but it seems to be less than 10um . Once I drag the thing onto the mill I should be a to measure it properly .

I was/am chuffed to buggery how well I got the shaft to fit (pure luck :-))
 

Attachments

  • Setting Pin d.jpg
    Setting Pin d.jpg
    89.4 KB · Views: 52
Neat setup! If you try this again, you might rough, bond the ball, then make a female taper in a chunk of stock held in the chuck. Seat the ball in the taper, kludge a drive dog, and use a dead center in the tailstock. Should be as accurate as the spindle bearings on the lathe...

I have a SIP RT too, but single axis. At ~30" table and around 460kG it's not an everyday tool... :D
 
Sorry to post so late to the party. But I have a real one for a PI-5 that I picked up today. The top centre is spring loaded.
IMG_1049.jpg

I didn't really know what it was when I bought it, I thought that the shaft looked like it would make a good indicator stand and the box looked well made.
When I picked it up and saw the SIP logo on the box, I knew it was special. Looked back at the catalogue and its matching PI-5 was sold as another lot. It's frustrating that the people running these auctions don't put more effort in to keeping tooling together. I contacted the auctioneer to see if they could put me in contact with the new owner of the PI-5 so I could re-unite them, but I don't hold out much hope. There are a couple of these for the PI-5 on eBay at the moment (search SIP PI-5)
 
Thanks for posting :-)

I concluded that making the top centre spring loaded would make it unpredictable (since any up/down motion will be read on the dial), so I'm actually quite surprised that the sip one is spring.

I can pop the top centre out and add a spring , so I may well test my theory.
 
OK, I give up. Can someone explain just how this device works for "centering and locating the table"? I assume that means getting the center of the table in line with the spindle's center line. But even that may not be the case.

In the photos below, it appears that a center in the (milling machine?) spindle sits in the 60° center depression of the top pin in the device. But what turns?

Do you turn the table and therefore the ball? If the ball is actually centered on the table, turning the table would result in no motion since the ball would only revolve about it's own axis.

Do you turn the spindle? I don't see how that would do anything either. The center in the spindle would just rotate and since it is on the center line of the spindle, no motion would result.

Do you turn the device? It would rotate about the center mounted in the spindle and the ball. Both of these are fixed in their location. But the distance between the ball and the center would not change so the indicator would not change?

Do you use the X-Y controls to move the table under the device? This seems like the most likely explanation. But simple trigometry tells me that the amount of displacement of these axis would be many times greater than the amount that the DI would change. The lever in the device would amplify that motion, but is it enough to overcome the trigonometric reduction. I doubt it. A quick calculation: The vertical distance between the contact at the top center and the ring of contact with the ball seems to be about 50mm. That is the hypotenuse of our triangle. The distance that the table is off center is the small side. Since 0.025mm (about 0.001") is a reasonable number for the desired accuracy, and that is our small triangle side and the distance we wish to reduce. Then the movement of the device in the vertical distance will be given by

The movement = hypotenuse - long side

Movement = Hypotenuse - ( sq rt(Hypotenuse^2 - Short Side^2))

Movement = 50mm - (sq rt(50mm^2 - 0.025mm^2))

Movement = 0.00000625mm

Now, assuming that the lever in the device amplifies this by a factor of 10. This is a generous amount, but I will use it anyway. Then the movement that the DI sees at the other end of that lever will be:

Movement = 0.0000625mm

That is about 400 times less than the amount of error that we assumed in the table's position. The DI shown in post #8 seems to have a sensitivity of 0.001mm. It would not even wiggle. We seem to be off by a factor of 400 here.

This may be a way for roughing in the table position, but it would seem that something else would need to be used for the final positioning.

What am I missing? Is the actual SIP device more sensitive? Is there a You Tube video on this? I would love to see it in action.



Next job is to make a setting pin (ball) , but in the mean time I designed this copy of the SIP 'dial indicator' for centring and locating the table.

[edit] forgot to add...
Does anyone here have a real SIP one? if so is the top centre (the bit the spindle centre fits into) sprung loaded?
 
To locate the spindle above the ball, it works much like a traditional DTI in a spindle mounted sweep.

An arm mounted in the spindle carries a centre point to locate the upper centre on the indicator, and an offset pin that sweeps out a plane that is perpendicular to the spindle axis.

A lever in the indicator body contacts the pin and transfers the reference plane to the DTI pointer.such that any movement of the indicator body the plane of the centre and pin will register on the DTI. It has a gain ,in my design of about 2:1

In use, the indicator is sweeped around while viewing the DTI in a traditional manner.

I suppose it's main advantage over a traditional spindle indicator setup , is stability - it is locked between centres so isn't as sensitive to handling- and because of the lever, it has some mechanical gain -i.e a 0.01 error will be read as 0.02 .

Once the ball is located below the spindle, simple trig will locate the plane and centre of the table.
 








 
Back
Top