Hello Guest it is June 26, 2019, 04:59:56 AM

Author Topic: CS Labs CSMIO/P-A  (Read 15176 times)

0 Members and 1 Guest are viewing this topic.

Re: CS Labs CSMIO/P-A
« Reply #360 on: September 19, 2018, 03:39:45 AM »
Hi,
good news.

Quote
Set the drive to max rpm of 1000 which give me 5m minute and re-autotuned the CSMIO
I think this is a mistake, the drive and servo should be programmed for maximum speed (3000rpm)  but the 5m/min axis speed limitation
should be in Mach.

A max servo speed setting is useful if there is some VNE (velocity never exceeded) like leadscrew whip or voltage overload on deceleration
for instance. Normal axis speed settings are rightly the province of the motion planner, Mach in this case.

Craig
My wife left with my best friend...
     and I miss him!
Re: CS Labs CSMIO/P-A
« Reply #361 on: September 21, 2018, 02:37:26 AM »
Hi
Can anyone explain the acceleration setting in Mach.
E.g. 1000/s/s
What exactly does that equate to?
Is it 1000/60/60 Which would be 0.27 seconds to reach velocity target.
My drive is factory set to reach desired velocity in 30mili secs.

Mick



Offline Hood

*
  •  25,847 25,847
  • Carnoustie, Scotland
    • View Profile
Re: CS Labs CSMIO/P-A
« Reply #362 on: September 21, 2018, 03:21:41 AM »
Hi,
good news.

Quote
Set the drive to max rpm of 1000 which give me 5m minute and re-autotuned the CSMIO
I think this is a mistake, the drive and servo should be programmed for maximum speed (3000rpm)  but the 5m/min axis speed limitation
should be in Mach.

A max servo speed setting is useful if there is some VNE (velocity never exceeded) like leadscrew whip or voltage overload on deceleration
for instance. Normal axis speed settings are rightly the province of the motion planner, Mach in this case.

Craig



Not sure I agree with this. I have had first hand experience of mismatched velocities causing issues. I was helping a friend retrofit a lathe and we had a servo on the turret. He had the drive set at the default 4000rpm but in Mach it was set to the equivalent of maybe 1000  RPM. He just could not tune it, when I told him to drop the drive down to the equivalent rpm the tuning was easy. Now there was a large disparity between the two rpms initially and a smaller difference may not matter but I personally think a match is the best route.

Hi
Can anyone explain the acceleration setting in Mach.
E.g. 1000/s/s
What exactly does that equate to?
Is it 1000/60/60 Which would be 0.27 seconds to reach velocity target.
My drive is factory set to reach desired velocity in 30mili secs.

Mick






The acceleration in the drive will be of no consequence as you are using Mach to command the drive so the acceleration will be at the rate set in Mach and not the drive.
As for 1000mm/s/s it simply means the axis will accelerate and reach 1000mm/s in one second, 2000mm/s in 2 seconds etc. Your Velocity is set in Mach in mm/min so 1000mm/min velocity will be 1000/60 = 16.7mm/s  With an accel of 1000mm/s/s you would reach your 1000mm/min in 0.017seconds
Re: CS Labs CSMIO/P-A
« Reply #363 on: September 21, 2018, 01:40:16 PM »
Hi Hood

Cheers on both comments. Sort of makes sense. Need to read it a few more times had a busy day.

Mick
Re: CS Labs CSMIO/P-A
« Reply #364 on: September 22, 2018, 03:10:01 PM »
Hi Hood

So using 1000rpm as an example as this is what my drive max velocity is set to. The calculation is 1/60 =0.017 seconds. Should I set my drive to reach max speed in 17ms also as opposed to 30ms?

Cheers
Mick

Offline Hood

*
  •  25,847 25,847
  • Carnoustie, Scotland
    • View Profile
Re: CS Labs CSMIO/P-A
« Reply #365 on: September 22, 2018, 04:55:37 PM »
Acceleration is velocity over time, so in Mach velocity is in Units per minute  and acceleration is units per second  per second.

So first thing you need to do for the calculation is change the set velocity from units per minute to units per second.
So for your example 1000 units per minute it would be 1000/60 = 16.666666667 units per second.

Ok so now we know the velocity in Units per second we can work out how long it will take to get there with the acceleration you have set in Mach.

So you say you have 1000 units/s/s acceleration. That means it will get to a velocity of 1000 units per second  in 1 second.

That is way faster than your velocity however as your velocity is 1000 upm so only 16.66667 units per second.
So to get the time it will take to get to your velocity you can work it out a few ways, one is to just simply divide your units per second velocity by the acceleration, so 16.66666667/1000 = 0.016666666667 seconds to get to max velocity.

You mention RPM however so I am not sure exactly what you are meaning. Obviously the rpm is the max value in the drive but what do you actually have in Mach? If you have 1000 set for the Velocity then  that will not be 1000rpm at the motor unless your ballscrew was 1mm pitch which is unlikely. So say it is 5mm pitch then that would mean for 1 rev of the motor you would move 5mm so the velocity in mach for a motor rpm of 1000 would be 5000mm/min.
If that is the case then the time to get to max velocity would be
5000/60 = 0.08333333333333 seconds.



Regarding the value for accel set in the drive, I do not think it will matter at all. I think that setting will be used only when there is an internal drive command such as indexing or internal speed command etc.
 As you are controlling the  drive via the analogue inputs (step/dir would be the same) then you are also controlling the velocity via that method and thus the setting in the drive is irrelevant.

Offline Hood

*
  •  25,847 25,847
  • Carnoustie, Scotland
    • View Profile
Re: CS Labs CSMIO/P-A
« Reply #366 on: September 22, 2018, 05:13:47 PM »
 My last calc above has a bit missing ::)
It should read 5000/60/1000 = 0.08333333333333
Re: CS Labs CSMIO/P-A
« Reply #367 on: September 23, 2018, 03:43:49 AM »
That makes total sense now.

I have the max rpm set in the drive at 1000rpm. I have the velocity set in Mach at 5000mm min which gives me 1000rpm max speed of the motor using 5mm ball screws.

So like you have worked out the 5000/60/1000=0.083. Which is slower than the drive setting of 0.030.

The manual is saying parameter 1-34 sets the acceleration if in velocity or PT mode. Screenshot attached.

I'll try setting 1-34 to something a lot higher to see if it does affect the acceleration.

Am I right in thinking if it does then settings in Mach would not be accurate outside the drive limits.  

Mick
Re: CS Labs CSMIO/P-A
« Reply #368 on: September 23, 2018, 04:06:43 AM »
Hi,
if you set Machs velocity and/or acceleration higher than the servo/drive then when Mach issues a max velocity positional move,
ie a G0 then the servo cant keep up and will fault out 'following error'.

At all times the servo/drive must be at least as capable as Mach expects. In fact the higher the acceleration of the servo/drive
the better it can follow the commanded input.

It is for this reason that I advocate the servo/dive be programmed for its maximum practical velocity and acceleration.

Note that as the forward gain (velocity/acceleration) is increased if there is a fixed phase lag in the drive chain, the encoder for instance,
then the phase margin reduces and may become unstable. Hood has already described that situation. Even if the velocity/acceleration
commanded by Mach is less than the servo/drive and that the servo/drive loop has an instability (low or vanishing phase margin) the
instability will be excited. It is not because Mach has commanded an impossible move but rather the servo/drive has an instability.
It is for this reason that much is made of self tuning notch filters...they can restore phase margin without undue reduction in forward gain.

By reducing the maximum velocity (or acceleration) of the servo you are in effect throwing away forward gain......ie you bought a servo that has 'sports car'
performance but you throttle it back so it performs like a 'Morris 1000'.

Craig
My wife left with my best friend...
     and I miss him!
Re: CS Labs CSMIO/P-A
« Reply #369 on: October 04, 2018, 03:14:15 PM »
Hi All

Had to take the wife away for a week so sorry for no update.

All good set the X-Axis up now and no issues what so ever. Done several tests with my 0.01mm gauge and see no error at all on both Z and X. Would be interesting to get a 0.001 gauge to test.

As previously discussed with acceleration. I can set the acceleration way above what is set in the drive and it does not fault out. So I think Hood was right in his previous comment.

Need to get the spindle sorted now and torn between exploring further the two motor options or 3 phase Lexium 32. This is totally off-topic so will start a new topic over the weekend.

Thanks again to Hood, Craig and everyone else who commented. 

Cheers
Mick