Advanced Search

I am looking for guidance on how to setup the C axis as a second spindle.  This is for a rose engine lathe (otbok.info/index.php?n=Main.RoseEngineLathe) where we need to use:

1. the Spindle to rotate the piece, and
2. the C axis to rotate the rosettes used to impart the design

We turn these quite slowly, certainly less than 10 RPM, and at times speeds approaching 3 or 4 minutes per revolution.  Thusly we are using stepper motors for both axes.

The end goal is to drive the spindle and the C axis simultaneously, but specifying a relationship percentage for their rotational speeds.  For example,  Is there some documentation I can read which describes 

1. How to setup LinuxCNC to treat the C-axis as a secondary spindle, and
2. The G-code I would use.

Apologies if this is already been asked and answered:  I searched the forum and could not find it.

Personally I would close the feed back loop with linear encoders, this will give actual feed back position, with a mesa FPGA this is a very well worn track with plenty of info available. Same for the spindle and encoder that also employs an index.
Sorry I'm too broke to have experience with Ethercat so really can't make a comment that maybe of use.
What kind of lathe are you working with ?
I'm kind of lucky with my Frankenstein Myford ML7\Super 7 in that I have plenty of room for the stepper motors, the only issue was the X axis, but I was able to use 10mm ballscrew and unlike most other lathes the nut is fixed to the carriage casting rather than running under the cross slide.

And for my techno babble mate, lathes conventionally use a XZ setup, Z being the carriage and X being the cross slide.
Also very few lathes, I can't think any that the average Joe would have access to lathes that use linear rails, mostly dove tails for cross slide and top slide and either plain rectangular ways for the bed,eg Myford, rectangular single prism, think 7x14 mini lathe or dual prism for the more industrial type. These are just the ones I can of off the top of my head.

Sorry to the OP for getting a bit off track.

Luckly I installed a disk brake on my lathe's spindle during the retrofit (brake is operated with pressurized air). This works pretty well for holding the spindle during cutting in a specific angle.

Not a problem, hope it was helpful.

Regarding spindle orient... I'd hoped to use orient for my lathe as a 'poor-man's c-axis' once I realized that caxis.comp wasn't doing what I wanted.

I quickly realized that is orient by itself won't hold a position.  You either need a spindle brake, or you need a servo drive that is in position mode (internally).  Servo drives in velocity-mode don't necessarily hold a position rigidly, but LCNC is perfectly capable of treating a position-mode servo drive as a spindle.

LCNC can orient a plain old VFD to a pretty exact position, but that does no good if you're trying to use live tools once the spindle is in the appropriate position.

I think you're on the right track for your first steps.  Get the spindle working as well as you can, then add the orient function.  Orient is simple in concept, but tuning and nailing the position can be tricky and time-consuming.  Especially if you have a heavy spindle/chuck.

Something else to consider... your spindle encoder need not be super-high resolution for spindle/orient functions, but for eventual c-axis work the higher the resolution the better.  And that can become a rabbit-hole when trying to balance high encoder speeds with high resolution... while remaining budget friendly.

Thanks a lot for the detailed reply. Really appreciated.

I know that this is not an official LCNC function. I was expecting this to be troublesome, but with your feedback, I might consider leaving it be for the moment.
If I find time I will try to understand ACIERA's setup and try this out.

I might just setup spindle orient function for now. This should be fairly straight forward.

Cheers,
Dave

Caxis.comp was originally written by Andy Pugh to control a single motor for velocity and positioning mode (spindle & c-axis).  I don't think it was intended to control two separate motors.

It's not part of the mainline LCNC components, nor are there many people using it... so I don't think you're going to find a wealth of info on implementation.  For your situation, there may be better approaches to managing your dual-motor arrangement.

The main hurdle to overcome for any spindle/axis (dual or single motor) is how to manage the axis encoder.  Once you switch from axis to velocity mode the axis will generate a following error when LCNC sees the 'axis' encoder moving away from the commanded position.  So we have to 'trick' LCNC and adjust the encoder value somehow.

• caxis.comp does this by intercepting the encoder output, and then subtracting the encoder change from the encoder starting value before sending that to the axis PID feedback input.
• Another method is to disconnect the encoder output from the axis feedback in spindle mode

The problem with #2 above is that when the encoder is re-connected (switch to axis mode) there is a significant following error instantly present.  The switching scheme @Aciera came up with a while back for his lathe (using a single motor) looks like this:

• Step 0
  • LCNC starts up with main spindle motor disabled
  • Home C-axis before any spindle moves (at start of LCNC session). 
• Step 1
  • Command C-axis to C0
  • Disconnect spindle encoder
  • Enable spindle mode
    • in Aciera's case that means switching his drive from position to velocity mode
    • for a two-motor arrangement that means retracting the axis motor and enabling the main spindle motor via M-codes & digital inputs/outputs
• Step 2 - Run spindle as a spindle
• Step 3
  • Use orient.comp to orient the spindle to 0
  • Disable the spindle motor
  • Enable the axis motor
  • Reconnect the spindle encoder

If everything is adjusted properly (orient offset, C0 actual position, etc.) there should be a very small difference between Orient-0 and Caxis-0, resulting in (hopefully) a small twitch of the spindle at most.

For the past year I've been fiddling with caxis.comp and have recently come to the conclusion that it isn't going to work for me.  There are a couple of annoying things I don't like (can't disconnect the spindle easily to manually rotate the chuck), but the serious problem I found was that spindle synchronized motion wasn't working properly.

Something - and I'm not smart enough to figure out exactly what it was - in the caxis.comp scheme was causing erratic behavior (big motion jumps) during tapping and threading commands.  Once I removed caxis.comp and use the spindle as just a spindle the tapping/threading stuff started working as expected.  So now I'm re-doing a bunch of stuff in my config and will eventually wind up with something very similar to Aciera's arrangement.

So to sum up, I'd suggest abandoing caxis.comp for your dual-motor lathe and search for Aciera's config.  I know he posted his files in the past year or so, although I think he's been working on a glitch with the encoder disconnect function recently.

Hi all,

I am trying to understand and implement the caxis.comp with my lathe. 

My setup is:

VFD with ac motor for "normal" spindle mode (analog signal --> pwmgen)
C-axis is connected via a reducing gear to the spindle, the c-axis motor is a closed loop stepper (step/dir). The c-axis drive is engaged to the spindle with a pneumatic cylinder. Normally it is disengaged

I am trying to better understand the caxis.comp and was wondering if someone can shed some light on this:

pin in float spindle-revs --> connected to spindle.0.revs
pin in float spindle-velocity-in --> connected to encoder velocity signal
pin out float spindle-velocity-out --> ? velocity signal for pwmgen and/or step/gen ?
pin out float position-fb-out --> is this position feedback calculated from encoder velocity signal ? is this connected to joint.xx.pos-fb?
pin in float pid-in; --> connected to PID output of c-axis
pin out signed state; --> what is this for?

Has anyone used this in similar setup (spindle with pwmgen, and c-axis with step/dir). Should the step/dir generator be run in velocity mode?
Would probably help to have some insight.

I have also tried a similiar thing using the mux component and connecteing the encoder signal to the c-axis feedback or not (depending on the mode). But as others have experienced there is an ussed with joint following after switching back to c-axis mode. Even when reseting the encoder position to 0, this will only work if the c-axis is also at 0.
I also could not achieve proper orientation (homing) when switching to c-axis mode.

Will keep trying ----


Regards,
Dave

AFAIK those breakouts do not have a 5V regulator, you must supply 5V.
So the behavior you see is pretty much as expected. That is, the 5V side is being
powered (weakly) via the PC parallel port data pins.

The regulator just supplies the inputs and the analog spindle interface
(regulates to 12V)

I converted my lathe from parallel port to mesa when I added linear encoders. Prior to that, I had zero problems threading with the parallel port using this 65 tooth, 260 count disk using a quadrature software encoder on an ancient PC. I only used the spindle encoder for threading (<150RPM) so I'm not sure what the limit for this setup was. I still run 4 identical ancient PCs for my LCNC machines but they all have mesa cards now.

3d printed timing pulleys are a great low cost solution for the low load timing application.

The slotted disk (I used thin copper coated pc board I had laying around) is easy if you have a working CNC router or mill. The finicky part for me was aligning the sensors to the teeth. To get quadrature and index signals all three encoders need to be positioned just right.  

I mounted my ballscrew bearing block directly to the apron. (original mount was damaged) This gives me just enough room for a 12 or 13T timing pulley to pass under the X slide. I may have modified the bearing block for clearance. When you remove all the manual clutter, there is room to mount the x stepper directly underneath so you end up with a shortened x axis that doesn't get in your way.

I used a wireless MPG for a while but found the joypad was much better for jogging and feeding at a consistent rate and I use MDI when I need precise movements anyway. As a bonus, it's easy to cut a nice 45 degree chamfer using the analog stick! 

Note that ' s-output => spindle.0.on' does not make much sense in the context of the hal sections posted above as 's-output' is a float signal and 'spindle.0.on' is a digital out pin.

Looks like you are actually missing the input to enable the stepgen.

So in the spindle section of your hal file change this line:
to this

Hi, there,  I am trying to follow step in this thread to set up stepper as a spindle. But the spindle stepper is not turning. I have tried to add this line as you mentioned. The LinuxCNC has a  problem with s-output  => spindle.0.on. Do you have any idea what I should have fix or add to hal in order to work properly. Thank you very much in advance.

Hello,

I am new to the CNC world, and would like to convert my small-ish lathe to CNC.
I think motors with around 2Nm for Z and around 0.6Nm for X will be sufficient, based on some preliminary measurements. For X I am constrained to Nema 17 size to be able to integrate them well.
For I/O and the spindle encoder, I would like to use a Raspberrry Pi and Beckhoff EtherCAT terminals, which I have already gotten to work (thanks for all the great ressources on here!). I would also like to use EtherCAT-connected drives for the axis (and maybe later the spindle), though I am unsure, which path to take there.

Initially, I was looking at closed-loop stepper drivers such as the Rtelligent ECT60 or the Lichuan CL3-E57H , which close the loop on the driver. But then I also found the Beckhoff stepper terminals  EL7041 and  EL7047 , which would make integration quite elegant as I am already using other Beckhoff terminals. With these however, I am not quite sure if they support closing the loop on the driver. The EL7041 manual mentions some closed-loop control aspects (differing between the standard and the -1000 model), but I don't fully understand what it actually can and can't do. The EL7047 supports closed loop in conjunction with FOC, but it is officially only supported with Beckhoff motors and requires them to be unloaded on startup for referencing. I could not find out if non-FOC closed-loop operation similar to the ECT60 or CL3-E57H is possible as well.
What approach would you recommend here?

In general (e.g. if the Beckhoff drivers can only act as open loop drivers with encoder interfaces), is it advisable close the loop in LinuxCNC when using such an EtherCAT based setup without any FPGA for low level control?

Another option would be AC servo drives (probably the  Lichuan ones due to their comparable cost), at least for the Z axis as I haven't found a Nema 17 sized option that would fit X. These would probably provide better performance, but seem more complex to set up and tune.
What do you think about this alternative?

I am looking forward to your comments, and hope the questions are specific enough. Please let me know if any details are missing.

Best Regards
Daniel

I have been reading a lot about options with the RPi and LinuxCNC, and am just getting more confused so I thought I would just ask a couple of hopefully simple questions.

I have a 3-axis (sometimes 4 axis) mill currently running LinuxCNC on an OrangePi. It has been run with GRBL on an arduino, and an old PC with LinuxCNC. The current setup works okay with the OrangePi but the interface lags and I have some mods I want to eventually do to my mill. I have the newest official RPi image running on a RPi 4b. I need to eventually marry this to an external motion board but I decided to start off with software stepping using a Pi hat, but I know that won't last long. Basically I'm done screwing around and want to setup a system that is capable and expandable EASILY.

I've been looking at the Mesa 7C81 but that is probably overkill for my purposes, but if it is truly as easy as it seems it may be my final choice. So first question is if setup of the 7C81 is as easy as it seems with the PNC config wizard?

Second question I have is whether I have missed another option that is less overkill.  Here is what I eventually "complete" my mill

• LinuxCNC running with touchscreen interface without lag or problems. Would eventually like on-screen keyboard with a wireless keyboard/mouse usable if needed.
• 4 closed loop steppers (XYZ and a rotary table)
• 7 home/limit switches (2 each on XYZ, 1 on rotary table)
• Spindle control (VFD)
• Spindle encoder for tapping.
• 3 encoder wheels (for using the machine as a pseudo manual machine)  
• Maybe a dozen or so switches/buttons for input to LinuxCNC
• Only need about 40 kHz stepping rate max.
  
  Should I just stop thinking about it and buy the 7C81 when it comes back into stock? Or is there an almost as easy solution that will fit the above needs.
  
  BTW, I am also planning to CNC my lathe and would want to duplicate the setup the mill has on the lathe so I don't have to relearn anything.
  
  Thanks in advance.