MesaCT configuration of axis limits

I now have the X axis moving with LinuxCNC Axis. I'm using ac servos that accept step and direction pulses like a stepper. I'm using a Mesa 7195t driver board.

Ball screws are DFU2005mm so 5mm pitch or 0.1969" per rev
Servos (direct drive) are setup for 1000 pulses per rev, so 0.000197 inches per pulse
Max input rate at Servo driver is 500kHz
I have the scale in Drive0 setup for 5082.85, and this yields correct displacement.
Speed limit (signaling wise) should be (500000/1000)*0.1969 = 98.45'' per second, so really really fast

In the Axis setup of MesaCT:
1) I'm assuming Min and Max limits are for displacement limits for example -15" and +15" is that correct?
2) Based on my calculations above Max Vel should be 98.45, but when I enter say 50, I get the following error:
hm2/hm2_7i95.0: stepgen.00.maxvel is too big for current step timings and position-scale. Clipping to max possible.

So clearly I'm misinterpreting the settings, or don't understand something.
I did move the table at F10, as well as F100 (F100 is quite fast)
I also did a g0 x 10 and the table moves at about 1" per second, plenty fast for me.
Please advise.

 

The maximum step rate is 1/(steppace+steptime) so
with the current values of 2000 ns (2 usec) for each, you have
a physical limit of 1/(4 usec) = 250 KHz.

You would need to reduce the steptime and or stepspace
to increase the maximum possible step rate.

PCW wrote:

The maximum step rate is 1/(steppace+steptime) so
with the current values of 2000 ns (2 usec) for each, you have
a physical limit of 1/(4 usec) = 250 KHz.

You would need to reduce the steptime and or stepspace
to increase the maximum possible step rate.
 

I'm trying to understand the following error:

hm2/hm2_7i95.0:stepgen.01.maxvel is too big for current step timings & position-scale, clipping to max possible.

My calculation is as follows:

Step Time, and Step Space are both set to 1us, so 500kHz max

Servo is configured for 10,000 step pulses per revolution
Direct drive to ballscrew, so 1 revolution = 0.1969 inches of linear displacement or 1 inch = 5.079 revolutions
96 inches per minute = 1.60 inches per second
1.6 * 5.079 * 10,000 = 81264 Hz, so well below the 500kHz maximum rate

There must be something else I'm missing?

I tried turning down the present ~50 pulses per mil to 10 for 0.0001 resolution, same error.

I would just like to understand what limit I'm actually hitting?


 

 

Can you post your hal and ini files?

This looks like something I need to check for in Mesa CT.

JT

PCW wrote:

Can you post your hal and ini files?
 

Please see attached, note I bumped down the Y axis steps from 10000 to 2000 however I still get the same error.




 

Looks like a IPM/IPS confusion
You have the maximum velocity set to 96 IPS (a bit scary...)

96 IPM would be 1.6 IPS
A decent starting acceleration value would be 5X
the velocity setting so acceleration to full speed takes
0.2 seconds = 8


[JOINT_0]
CARD = 0
TAB = 0
AXIS = X
MIN_LIMIT = -9
MAX_LIMIT = 9
MAX_VELOCITY = 96 <<<<<<<<<<<<<<
MAX_ACCELERATION = 2
TYPE = LINEAR
SCALE = 5095.18544

PCW wrote:

Looks like a IPM/IPS confusion
You have the maximum velocity set to 96 IPS (a bit scary...)

96 IPM would be 1.6 IPS
A decent starting acceleration value would be 5X
the velocity setting so acceleration to full speed takes
0.2 seconds = 8
 5095.18544
 

That was it. Thanks again!


I'm really happy with my decisions so far:
Linux vs. Windows, no explanation required.
ac servos vs. steppers, fast, powerful and silent.
Mesa vs. other options, seems well thought out with room to grow.

Pretty soon this thing will be all singing all dancing!

I've improved MesaCT to show the units for velocity settings based on your experience . So thanks for helping me improve MesaCT.
 

JT

JT,

Thank you and I'm a so glad I decided to go the LinuxCNC-Mesa route!