Re: Z Axis motor speed

> ==
>
> Software is DeskCnc via their dedicated 'PIC' based serial

controller - not

> sure
> about max encoder frequency input - but Fred Smith states on their

forum:

>
> 'At the present time our Serial and usb interfaces provide the

same max

> throughput of 300 blocks per second, and 40,000+ Hz, step rates to

each

> axis'
>
> I am sure the pitch is correct, I have the setting for 'Steps per

MM ' at

> 200 and the shaft encoder is rated at 2000 pulses per rev + the Z

axis is

> moving the correct distances under software control.

Your first post listed the shaft encoders at 2000 *lines* which would
produce 8000 pulses. Which servo motor control are you using? Is it
the SV500 card they sell?....can't be with the motor you are running.

I understand the pulse generator is the DeskCNC card and software but
the servo encoders are referenced back to the motor control. The
Gecko 320 for example has an upper step input limit of 250K
steps/sec, well within your range/

I need confirmation on two things: The actual pitch of the leadscrew
measured on the screw and the actual specs on the encoders. I would
expect 2000 pulse encoders to actually be 500 line encoders.

The 300 "blocks" refers to blocks of code the engine can process and
the 40K pulses per axis is the max step rate.

200 steps per mm means you should be able to make one shaft
revolution with 2000 pulses or .005mm/step (resolution). I doubt the
rest of your machine can hold that tolerance. Thermal expansion will
account for more than twice that alone. If it takes 2000 pulses to
do one shaft revolution then you should in theory be able to get 20
rev/sec (except the motors won't do that).
10mm pitch seems pretty course and should yield some pretty good
rapids. Somehow the numbers are not working out.

I don't know how they set their velocity in DeskCNC but in MACH2 it's
in mm/sec. Make sure you have the velocity for Z set to something
like 45 or 50. Lower numbers would not effect your steps per mm but
would limit your max speed.

>
> I must admit that I have never considered the input frequency of

the system

> to be a limitation ?
> I would not really want to decrease the resolution of the encoder,

as if my

> understanding is correct I will only be sacrificing accuracy ? BUT,

I would

> like to get the Z axis moving faster.

If you did drop the encoder count by a factor of 4 you would still
have .02 mm resolution (.0008 inch). You probably have 2 to three
times that amount in column or gantry flex! We call this false
accuracy. The accuracy of your machine is not the resolution but a
combination of all the error factors under a range of cutting loads
and temprature changes. Only the largest most ridgid machines can
hold consistant .03mm and most will be closer to .04 or .05.
Resolution is the smallest distance (in theory) your load will move
given a single step. Having a resolution much finer than the actual
accuracy you can hold given all the forces at work is just more steps
you have to provide with no increase in overall accuracy.

>
> Funny, now I have re read your very concise and helpful answer - the
> encoders on my X any Y axis's are only 400 pulses per rev and

although

> different motors - the x and y move at very fast speeds.
>
> Dave Fisher UK
>
>
>