[Al_The_Man]

The old original DC drives were velocity mode drives and used a tach for the 'inner' loop, the encoder was the 'outer' loop, the inner loop was always tuned first then the outer loop, now DC motors are rarely if ever, supplied with tachs, as the torque mode drive replaced them.
Al.

[Teknic_Servo]

In my quote above, you are missing my last sentence

Hi Jim,

Thanks for the detailed post. That’s exactly the kind of precise data I was looking for.

I did read the last sentence of your paragraph, but I misunderstood it. I thought you had no inner velocity loop. That can be done, but it’s tricky and depends on very good mechanics and servo control.

If you don’t mind, I have one more question. Can you describe the control of the Z-axis on the BP clone mill in a little more detail. Is it correct to assume that the analog input to the stepper drive is a velocity command? If so, then your velocity loop is around the load, correct? That is particularly interesting given the compliance of the air spring in the transmission. What are the specs on that? Does it provide a lot of force relative to the Z-axis mass? How was this axis to tune?

Ok, I guess that’s five questions. Or, one question with five parts

I appreciate your time on this. It’s always interesting to hear what other knowledgeable people are doing with their systems.

Best regards,
Warren

[Jim Dawson]

My pleasure to brag a bit.

A couple of pictures here is worth at least several paragraphs of explanation.

The black cable is connected to a 1 9/16 air cylinder on the other end. The machine end is connected to a special quill stop that I designed for this purpose, also has the mounting bracket for the reader head built in.

The magnetic scale is mounted in place of the original scale.
https://www.cnczone.com/forums/attac...d=431588&stc=1

And the air cylinder on the back of the red mount bracket.
https://www.cnczone.com/forums/attac...d=431594&stc=1

Here is a close up of the original quill stop and the new one.
https://www.cnczone.com/forums/attac...d=431590&stc=1

The quill is driven by the gear train through the normal quill pinion. I just removed the original power feed hardware and replaced it with the gearbox I built.

The idler gear is on a cam that allows it to be disengaged from the quill bull gear for manual use. Shown in the disengaged position. The motor (upper left) drives the the idler gear.
https://www.cnczone.com/forums/attac...d=431592&stc=1

Can you describe the control of the Z-axis on the BP clone mill in a little more detail. Is it correct to assume that the analog input to the stepper drive is a velocity command?

Yes, a velocity command. I am using an Automation Direct SureStep drive. https://www.automationdirect.com/adc.../stp-drv-80100

If so, then your velocity loop is around the load, correct?

Yes, the feedback from the linear encoder is connected to the Galil DMC-1846 board, which also provides the +/- analog command signal.

That is particularly interesting given the compliance of the air spring in the transmission. What are the specs on that?

The cable/cylinder can exert > 100 lbs up pressure on the quill but I normally have at set for about 50 lbs. This is normally enough to overcome the down pressure exerted by any end mill I might use. The air spring cylinder is paralleled with a 5 gallon air tank to keep the pressure constant, not relying on the regulator to react as fast as the cylinder can move.

Does it provide a lot of force relative to the Z-axis mass?

The quill weighs around 25 lbs or so, so by setting the up force on the quill to about 2X the motor is forced to overcome about 25 lbs of force just to move the quill down. Now the question is what happens when the quill is moving up? Well it works just fine, just requires a little less motor torque to move up. The controller keeps the speed constant in either direction. I did have to reduce the motor torque, because I was afraid of breaking something. I only want a maximum of about 300 lbs of downforce possible, that's about what it takes to drive a sharp 1/2'' drill bit through mild steel under normal conditions. If everything goes horribly wrong, I do have a mechanical shear point built in to limit the down force to about 150% of maximum. I don't want to have to replace the quill pinion.

How was this axis to tune?

Very easy, only took about 10 minutes to dial it in. But due to the short travel, I only allow 60 IPM maximum, so running pretty slow. The total gear ratio is about 15:1 so the motor is only turning at about 900 rpm at maximum.

[Teknic_Servo]

Hi Jim,

Very cool. The pictures are very helpful.

A lot of controls PhDs will tell you that what you did will never work because of the lack of the stabilizing inner velocity loop

Your design is very stiff (especially when driven “pretty slow”) and has essentially no backlash due to the air spring (and tank). Also, the mounting of the linear encoder read head is done well (very rigid and minimal cantilever). Nice work!

Best regards,
Warren

[Jim Dawson]

.........

A lot of controls PhDs will tell you that what you did will never work because of the lack of the stabilizing inner velocity loop

Thank you for the kind words Warren

That's the advantage of not having a PhD, I don't ''know'' what won't work. I just go ahead and do it and make it work.

[jittechnologies]

Sorry to revive a dead thread, but I'm having issues with a ClearPath SDSK implementation. 3 axis router, with the 4th axis slaved to the 1st (for the x axis gantry). The slave motor keeps faulting for Step Input Timing Error. Only the slave does this. It is being operated by an Acorn 4 CNC control. All cables have been replaced. New ground wires have been terminated. All connections and shield terminations have been checked. The machine has not completed a single cycle without faulting, as the package was purchased last fall, and only completely finished installing last week. I even disconnected power from the spindle drive, to make sure it was not the culprit. Jogging is not normally an issue, but it had failed during jog. It usually only faults when running a cycle. Any help world be greatly appreciated.

Sent from my SM-N975U using Tapatalk

[unterhaus]

Can you swap motors without swapping anything else?

[islaww]

Have you installed EMI filters on power supply power lines?

How about installing resistors on the step/direction signal terminals as per the Teknic manual?

[servomotors]

read so many post,only learned one point from the article only, it is hard to manage,maybe this work really only qualified for a PhD

[jittechnologies]

Have you installed EMI filters on power supply power lines?

No, I haven't. Do you mean on my 70V bus supply, or on the main lines coming into the machine. Why would this only affect 1 motor out of 4?

How about installing resistors on the step/direction signal terminals as per the Teknic manual?

Yes, to no avail.

Sent from my SM-N975U using Tapatalk

[jittechnologies]

Can you swap motors without swapping anything else?

Yes... and it was my plan to try that next. Just have having to remove covers, remove pulleys, belts, etc. But again, that's my next move. Thanks!

Sent from my SM-N975U using Tapatalk

[Teknic_Servo]

Hi Jittechnologies,
I'm sorry that you are running into trouble with your build and hopefully this post will help your troubleshooting efforts. The ClearPath motor inputs are optically isolated inputs which are regularly used in the most electrically noisy environments including plasma cutting and high powered laser cutting applications.

A step input timing error in a ClearPath motor is caused when the motor detects electrical noise on the step input. Generally this error is caused by one of the following:
-The controller is outputting less than 4VDC and is unable to reliably turn on the ClearPath motor's input.
-The controller uses open collector outputs and resistor specified in the ClearPath manual is not installed correctly. Generally this will only effect machines with cable runs longer than 10ft however in applications with a large amount of electrical noise (such as plasma cutting applications) the resistors may be needed on shorter cable runs.
-The Controller's maximum step pulse frequency is higher than 700kHz or the minimum step pulse time is less than 715 nS (link). These settings are typically configurable in the controller.
-The wiring to the ClearPath inputs does not maintain a separate signal pair for the step and direction signals between the motor and controller. Sometimes this can happen when a controller doesn't have enough connections for V+ and GND.

Hopefully these suggestions will help you narrow down the possible causes of your shutdowns, but if you have any additional questions please give us a call at 585-784-7454, or use our "Contact Us" form online (https://www.teknic.com/contact/)
Brendan F. Teknic Servo Systems Engineer