Hello Guest it is December 12, 2019, 02:56:06 AM

Author Topic: AC Servo Spindle and ESS  (Read 6215 times)

0 Members and 1 Guest are viewing this topic.

Offline Hood

*
  •  25,856 25,856
  • Carnoustie, Scotland
    • View Profile
Re: AC Servo Spindle and ESS
« Reply #20 on: April 15, 2015, 07:32:45 AM »
Jeff, you seem to be misunderstanding, I am talking Spindle, not axes.
If I press feedhold the axes decelerate to a stop, the spindle keeps turning. If I then press Stop the spindle will, with the ESS, have its pulses withdrawn immediately and thus will come to a sudden stop.
The USB SS does it correctly by decelerating the spindle at the rate of acceleration set in Motor Tuning, the ESS should do the same.

Hood
« Last Edit: April 15, 2015, 07:34:53 AM by Hood »

Offline Hood

*
  •  25,856 25,856
  • Carnoustie, Scotland
    • View Profile
Re: AC Servo Spindle and ESS
« Reply #21 on: April 15, 2015, 07:48:06 AM »
Dan, also I am sure you are aware as this happens with the USB version as well as the ESS, but just in case you are not. Make sure you edit your M3 and M4 macros so that they check whether the spindle is running before doing the reversal, if it is have the macro stop the spindle first or you will probably get the same effect as you are experiencing with the Stop issue.

I am not sure how the parallel port handles things as it has been a long time since I have used it but the CSMIO/IP-A and -S will first decelerate before reversing, however as I am unsure how the parallel port handles things I can't say whether this is a ESS/USB SS fault or just emulating the parallel ports actions.

Previously I had edited my CAM to do a M5 before a reversal but I feel it is actually safer to do it via the M3/M macros.

Hood

Offline Dan13

*
  •  1,244 1,244
    • View Profile
    • DY Engineering
Re: AC Servo Spindle and ESS
« Reply #22 on: April 15, 2015, 09:44:23 AM »
Yes, I am aware of this Hood. In the kind of work I do, I rarely reverse the spindle just like that. So usually there will be an M5. But it does happen seldom. It's a good idea to have a  macrto take care of this. Thanks.

Dan

Offline Jeff_Birt

*
  •  1,107 1,107
    • View Profile
    • Soigeneris
Re: AC Servo Spindle and ESS
« Reply #23 on: April 15, 2015, 11:55:54 AM »
Hood, a S/D spindle is treated just like an axis in that Mach 3 generates a pulse train just as for the other axis. When you press the 'Stop' button Mach 3 (parallel port driver) will immediately stop outputting pulses, in other words instant deceleration. I don't think this is desirable but it is how it works and with stepper motors you are sure to loose position if the machine is moving.

The SmoothStepper tries to emulate the parallel port driver so it to immediately stops outputting pulses. Using Feedhold is the only way to decelerate to a stop, then press the Spindle button to turn off the spindle safely, then you can hit Stop. The issue I was talking about before is that if the SmoothStepper tries to decelerate the axis (including S/D spindle) on its own, it has no idea of the correct path the machine should be following so the best it will be able to do is decelerate in a straight line (whereas Feedhold will stay on the correct path.)

Mach 4 is so much quicker at talking to the plug-ins that I think we'll see a lot better behavior in this type of scenario and we can make the Stop button to a feed hold, which will work properly with external devices, and then to a spindle off and then a Stop.
Happy machining , Jeff Birt
 

Offline Hood

*
  •  25,856 25,856
  • Carnoustie, Scotland
    • View Profile
Re: AC Servo Spindle and ESS
« Reply #24 on: April 15, 2015, 12:08:18 PM »
Jeff,
 I can not recall how the parallel port works in respect to the spindle. I saw Ger saying pressing feedhold does not halt the spindle, so that is the same as the ESS. However I do not know if subsequent press of Stop suddenly halts the spindle without deceleration or not and I have no way of testing with the PP to see.
There is however one pertinent fact, the USB SS handles things as you would expect so it would be reasonable to presume the ESS should do it that way as well.

Greg said he would look into it many times but either he could not find the problem or just was too busy to get round to fixing it and I gave up mentioning it.
I still like the ESS but it does have problems with Step/Dir spindle and this is just one of them, they are not massive problems but can be alarming at times.

Hood

Offline Jeff_Birt

*
  •  1,107 1,107
    • View Profile
    • Soigeneris
Re: AC Servo Spindle and ESS
« Reply #25 on: April 15, 2015, 12:19:39 PM »
What problems with the S/D spindle are you referring to? I have several customers who use S/D to Analog cards to control their spindle speed and that seems to work fine.
Happy machining , Jeff Birt
 

Offline Jeff_Birt

*
  •  1,107 1,107
    • View Profile
    • Soigeneris
Re: AC Servo Spindle and ESS
« Reply #26 on: April 15, 2015, 02:23:24 PM »
I was chatting with Greg about this. The issue is that in Mach 3 that a Step/Direction spindle is sent to external devices as a jog command, when you hit 'Stop' Mach cancels the jog processor so all the sudden you have no S/D output to the spindle servo controller. The axis are also cut off suddenly.

What I might think of doing is customize the screen set with my own stop button. The stop button would do a feed hold, when the axis stop then do an M5, when the spindle stop then do a 'Stop'.
Happy machining , Jeff Birt
 

Offline Dan13

*
  •  1,244 1,244
    • View Profile
    • DY Engineering
Re: AC Servo Spindle and ESS
« Reply #27 on: April 15, 2015, 02:32:49 PM »
With the USB SS, Greg somehow made it work properly. A custom screen is a possible workaround (pressing the reset button would still result in the same), but a fundamental solution would be solving it at the controller - at no occasion it should allow for a sudden stop of the pulse train stop to the spindle. Even when it recognizes a sharp interrupt in the pulse train, it should take over and safely decelerate, regardless of Mach3 state.

Dan

Offline Hood

*
  •  25,856 25,856
  • Carnoustie, Scotland
    • View Profile
Re: AC Servo Spindle and ESS
« Reply #28 on: April 15, 2015, 03:19:36 PM »
What problems with the S/D spindle are you referring to? I have several customers who use S/D to Analog cards to control their spindle speed and that seems to work fine.

The ones that come to mind are
Very occasionally the pulstream will stop for a split second, this also happens with the USB SS but as I have that on a mill it is hardly noticable. On the lathe however with the heavy chuck it is more noticable as you get a clunk from the backgear. As said however this is very occasional and can be fine for long periods. I only ever once managed to capture it on a scope and that was just pure luck that I was walking past at the time as it had been connected up for days without an issue.

The spindle override is poor, if you override the spindle it will be consistent for a short period of time then it will get very erratic, again this is amplified due to the weight of the chuck. I have my spindle override disabled due to this. At the start I wondered if it was the external pot that was the issue or possibly  noise  but I can watch the PLC and also the Brain and the numbers in it are very steady and the values in the PLC are big so any variance is easily seen.

The pulses themselves from the spindle output is not like the axes in that the width is not controlled with the velocity but rather it is fixed. This causes issues if you try to use a high steps per unit as the pulses are just too close together at higher RPMs. My servo drive can really respond to a high rate of feedback but I can not take advantage of it due to the way the spindle pulses are output. If I uses a high interpolation on the feedback encoder then I get nice control but I then have to use a large electronic gearing for the steps per unit input, this then can cause problems if stopping from higher RPMs. To confirm that it was definitely an issue with the way the spindle pulses were output I configured the spindle as an axis and the problem dissapeared.




Hood

Offline simpson36

*
  •  1,374 1,374
    • View Profile
Re: AC Servo Spindle and ESS
« Reply #29 on: April 22, 2015, 09:03:41 AM »
Coincidentally, I have been messing with screens lately and it occurs to me that this issue might be easily fixed by replacing the action behind the stop button.

Both my mill spindle and 4th axis have enable/disable states and pneumatic locks, so I have long ago replaced the spindle ON/Off with macros which check the condition of the lock and enable state prior to starting the spindle and check the speed of the spindle prior to engaging the lock and so on. Both spindles are AC servos, but the drives themselves take care of deceleration in the event of a sudden pulse stream loss, so that was never an issue that I needed to resolve.

I jumped into the Mach3 screen editor to confirm that the red 'stop' button can indeed be changed to trigger a VB script. (see attached image)

From there it should be an easy task to program any sequence you want including checking the spindle speed and decelerating it to zero prior to MACH abruptly shutting off the pulse stream. I am unsure if the 'ismoving' function in MACH works with the spindle, but in case it did not, I recall Hood saying that there is a 'speed reached' signal available from the drive.  I use this signal to prevent Mach from moving the axis prior to the 4th axis reaching speed (for turning) and it works fine.

There are probably a lot of ways to skin this cat, but off hand, I don't see why commanding zero speed and then monitoring a 'speed reached' signal from the drive prior to issuing a 'stop' would not be a complete solution. There may be a way to stop the axis movement while waiting for the spindle to decelerate, but I would need to noodle on that for a while.

« Last Edit: April 22, 2015, 09:15:01 AM by simpson36 »