Taig lathe DC motor

[John HaineParticipant @johnhaine32865]

Solved problem with stepper motors. Requires an acceleration/deceleration algorithm that looks ahead and slows down smoothly.

[Robert Atkinson 2Participant @robertatkinson2]

As discussed on another thread using ramp up and down on a ELS conversion (assumes a standard lathe where spinde speed is free-running) causes errors in the thread pitch. The error may or may not be acceptable depending on the application.
Counting a quadrarure encoder without missed steps, particuarly if it stop – starts or reverses, is not a trival task for software. It is normal for encoder interfaces to use hardware counters. Many microprocessor families have devices with these interfaces buit in but they tend not to be used in the general purpose models. Ther is lots of code examples for quadrature encoder interfaces out there but most are intended fir applications like control knobs where there is a “human in the loop” to correct any errors.
An ELS conversion controller is not a trivial task. I’ve been invoved with the design of encoder interfaces as part of my “day job” but when I saw this conversion kit:

Presentation

I decided it wasn’t good use of my time to design my own and just bought one. Not finished installing it yet. It’s been a bit cold in the workshop.

Robert.

[John HaineParticipant @johnhaine32865]

“As discussed on another thread using ramp up and down on a ELS conversion (assumes a standard lathe where spinde speed is free-running) causes errors in the thread pitch. “

A bit dogmatic that Robert.  Mach3 threads just fine.  It’s the crazy way the ELS systems work that gives problems.

[BazyleParticipant @bazyle]

Why is a ramp down needed?

[Robert Atkinson 2]

On Robert Atkinson 2 Said:

[…] An ELS conversion controller is not a trivial task. I’ve been invoved with the design of encoder interfaces as part of my “day job” but when I saw this conversion kit: […] I decided it wasn’t good use of my time to design my own and just bought one. Not finished installing it yet. It’s been a bit cold in the workshop.

That’s a very impressive little box-of-tricks, Robert … and your endorsement brings comfort.

I will be interested to see your installation, and particularly if you find the need for any additional EMC shielding.

MichaelG.

[John Haine]

On John Haine Said:

“As discussed on another thread using ramp up and down on a ELS conversion (assumes a standard lathe where spinde speed is free-running) causes errors in the thread pitch. “

A bit dogmatic that Robert.  Mach3 threads just fine.  It’s the crazy way the ELS systems work that gives problems.

Mach3 running on what hardware? Mach3 is a CNC control program for PC’s. I’m sure it cuts great threads on a 5 axis Hass….

No mater what the hardware if you change the feed rate relative to the spindle speed you will get a pitch error. As most conventional lathes do not allow precise control of the spindle speed by external signal ramping the feed rate must give a pitch error.

Robert.

[John HaineParticipant @johnhaine32865]

This is based on my own practical experience, running on a standard PC with a Super 7 actually, using a standard induction motor.  The tool starts at a fixed position to the right (for a standard thread) of the work for each cut and accelerates up to feed rate when the controller gets a spindle index pulse.  The spindle speed, monitored by M3, sets the feed rate accordingly before the cut starts. If you don’t go mad on DOC it threads just fine. At the end of the cut the tool pulls out, the feed decelerates, the tool is moved back to the start position, fed back in, and the next cut starts again timed by the index pulse. Like all CNC operation, this works through precise control of the acceleration and deceleration.  Which is why I said it was a solved problem.

OK, later controllers (Mach4 I think, UCCNC, etc) also have a spindle encoder to monitor its speed to give real-time adjustment of the feedrate, but to reiterate, Mach3 threads well.

[Robert Atkinson 2Participant @robertatkinson2]

Hi John,
I agrre with all you say. But that is a 2 axis CNC lathe conversion not an ELS. From your description the X axis does not accelerate when the cutter is engaged with the work so of course it does not change pitch. A typical ELS does not have a Y axis to automatically engage the tool. They are an electronic gearbox with extras.
Of course even a 2 axis CNC lathe has issues if working up to a shoulder or similar feature.

By encoder on the spindle do you mean a servo driven spindle or an absolute encoder? Even my basic ELS has a encoder connected to the spindle.

Robert.

 

[John HaineParticipant @johnhaine32865]

Er, actually the Z axis is longitudinal and X axis across on a lathe.  Just having an electronic gearbox seems a waste of time to me when without that much extra hardware you get full CNC.

I’ve never had an issue working up to a shoulder, what is the problem?

Neither a servo nor absolute encoder.  My spindle has a 1 ppr index (slotted disc + opto-interrupter), systems such as UCCNC or Mach 4 also need a higher tooth count wheel.  The index tells the controller when to start the threading passes, the other sensor keeps track of instantaneous spindle speed. With these two the controller “knows” where the spindle is all the time.

Worth noting that there are versions of GRBL now that can do lathe threading with an index + speed sensor running on Arduino variants, which apart from the extra X-axis stepper and driver should cost not much more than an ELS.

[Robert Atkinson 2Participant @robertatkinson2]

Sorry on getting the axis letters wrong.
All you are refering to are CNC machines. Not everyone wants CNC or uses CAD or wants to mess with G-Code etc.
An ELS requires none of that, just a encoder (not even an index pulse) coupled to the spindle and a motor (stepper or servo) connected to the leadscrew. That is easilly done and equally easy to revert back.
I don’t want to add a motor to the cross-slide on my ML7 or use CAD at home.

Robert.

[Author]

Posts