Gallery, Projects and General > The Design Shop

Lathe Screwcutting Idea

<< < (2/3) > >>

John Hill:
Weston Bye, I can understand how you could turn the leadscrew at the right rate in relation to the spindle but how do you keep accurate relationship between the two?

Weston Bye:
John,

Indeed, you have identified the fundamental issue in threading.  Being able to generate the proper ratios between spindle and leadscrew is of little value if you can't follow the same path on successive cuts.  The "trick" is to identify the starting point and return there every time a new pass is started.

The starting point includes all elements involved; obviously, the saddle, but also the spindle angle.  These are the fundamental factors.  Knowing the starting leadscrew angle also will improve accuracy.  This can be demonstrated on a mechanically geard lathe by using a dial or digital indicator to determine the starting position of the saddle, and a pointer and hash mark on the spindle.  But, the half nuts won't engage?  Indeed, the gear train between the spindle and the feed screw will only have a certain number of points where the half nuts will drop in with the spindle hash mark aligned with the pointer.  Manually rolling the spindle over until the 'nuts engage with the spindle aligned gives you a starting point.  With everything aligned and locked up, you start the spindle to make your threading pass.

What I have described is threading without benefit of the threading dial - it can be done but is cumbersome.  With electronic threading, rather than using indicators, hash marks and pointers, a few strategically placed switches are used.


Weston

NickG:
I had thought of this for slow feeds but I can see it being too complex for threading (at least for me!), that relationship being the stumbling block. When you load the spindle it won't slow down at the same rate as the leadscrew so there'd have to be some sort of wizardry that would react quickly enough to maintain the ratio. Obviously not impossible but difficult. would be an interesting project to see.

Nick

djh82uk:
Ok you got me thinking now, If you had an encoder wheel on the spindle, the MCU would always know what angle the chuck is at, aswell as where it was at the beginning, however you would little to zero control over the spindle, so the spindle would either need to be replaced with a powerful servo/stepper so that you did have control, or you would need an equivalent of an absolute dro on the feedscrew so it would work out where the carriage needs to be so that it engages the workpiece at the correct place.

But as soon as you factor in Backlash you have a problem, this then starts to become expensive and so you may aswell just CNC the damn thing.


My initial idea was just to have it so that the electronic feedscrew just replaces the change gears, other than that you do the screwcutting exactly as you did before.  The mcu would be quick enough to react to rpm & direction changes on the spindle and adjust the feed motor speed/direction to the correct proportion.  But to do this the Feed motor has to be a stepper or servo.


But now that you mention it, just adding a slow rpm motor for getting a very fine finish would be a welcome addition, it could be mounted the opposite end of the feed screw so you don;t even have to mes with the gears, just disengage the feedscew from the spindle before firing up the feed motor.  Although you would obviously need some sort of saftey device to ensure the feed motor is off if the feedscrew is engaged with the spindle

DJH

Weston Bye:


--- Quote ---...however you would little to zero control over the spindle,...
--- End quote ---


Control over the spindle speed is not necessary - as long as the leadscrew can remain in sync with the spindle.  Any variations in spindle speeds must be proportionally matched by the motor driving the leadscrew. 


--- Quote ---My initial idea was just to have it so that the electronic feedscrew just replaces the change gears...
--- End quote ---

This is quite feasable.  However, I suggest putting the encoder somewhere upstream of the back gear rather than directly coupled to the spindle.  This way, an encoder with a lower pulse count (less expensive) can be used but still provide enough pulses to divide down to drive the leadscrew.  But, because you don't have the fixed relationship inherent in the gear train, some sort of starting synchronization between the spindle and the leadscrew would be necessary for threading.

Weston

Navigation

[0] Message Index

[#] Next page

[*] Previous page