Re: Step resolution
Posted by
metlmunchr
on 2004-08-19 08:29:04 UTC
The belt pitch doesnt come into the picture. The leadscrew pitch
seems to be missing from your calculations. Let's assume you have a
metric leadscrew with 5mm/rev pitch. Your encoder generates 2000
pulses per revolution of the leadscrew. 5mm/rev X 1rev/2000pulse =
.0025mm/pulse. Dividing by 25.4 to convert to inches gives .0000984,
or essentially .0001"/pulse. How good is this figure? Well, it's
mathematically correct, so it's perfect in that respect. In the real
world its got very little to do with the accuracy of the machine.
Depending on the type of screws, fit of components, and rigidity of
the machine, the machine will probably be able to make parts with an
accuracy somewhere between 5 times and 100 times the resolution.
Manufacturer's claims notwithstanding, it is difficult to hold parts
to +/-.0005" on an ongoing basis on a machine that costs $100,000
without a lot of continual attention to the process. These machines
today typically have resolution of .00001, or ten times greater than
your resolution, and .0005 is 50 times their least resolution. In
general, these machines could have a least resolution of .0001 instead
of .00001, and the parts would come out identical, yet the accuracy
would be 5 times the least resolution instead of 50 times.
seems to be missing from your calculations. Let's assume you have a
metric leadscrew with 5mm/rev pitch. Your encoder generates 2000
pulses per revolution of the leadscrew. 5mm/rev X 1rev/2000pulse =
.0025mm/pulse. Dividing by 25.4 to convert to inches gives .0000984,
or essentially .0001"/pulse. How good is this figure? Well, it's
mathematically correct, so it's perfect in that respect. In the real
world its got very little to do with the accuracy of the machine.
Depending on the type of screws, fit of components, and rigidity of
the machine, the machine will probably be able to make parts with an
accuracy somewhere between 5 times and 100 times the resolution.
Manufacturer's claims notwithstanding, it is difficult to hold parts
to +/-.0005" on an ongoing basis on a machine that costs $100,000
without a lot of continual attention to the process. These machines
today typically have resolution of .00001, or ten times greater than
your resolution, and .0005 is 50 times their least resolution. In
general, these machines could have a least resolution of .0001 instead
of .00001, and the parts would come out identical, yet the accuracy
would be 5 times the least resolution instead of 50 times.
--- In CAD_CAM_EDM_DRO@yahoogroups.com, "azzi_b" <azzi_b@y...> wrote:
> I am wanting to calculate the resolution of my machine and using the
> calculation in the mach2 instructions get a figure of .0025mm per step
> working on 400 steps per mm.
>
> the servo has 1000 pulses per rev (250 step diff encoder), 2:1 belt
> reduction (26:13) and 5mm belt pitch
>
> is the .0025mm per step right, and if so how good is this figure.
> Also what would this be in imperial ?
>
>
> Thanx
>
> Aaron
Discussion Thread
azzi_b
2004-08-19 06:39:33 UTC
Step resolution
Tom Hubin
2004-08-19 07:20:11 UTC
Re: [CAD_CAM_EDM_DRO] Step resolution
metlmunchr
2004-08-19 08:29:04 UTC
Re: Step resolution
Fred Smith
2004-08-19 09:15:01 UTC
Machine accuracy, was Re: Step resolution
JanRwl@A...
2004-08-19 14:23:22 UTC
Re: [CAD_CAM_EDM_DRO] Step resolution
Aaron
2004-08-19 16:21:59 UTC
Re: [CAD_CAM_EDM_DRO] Step resolution
Stephen Wille Padnos
2004-08-19 16:39:59 UTC
Re: [CAD_CAM_EDM_DRO] Step resolution
JanRwl@A...
2004-08-19 17:24:39 UTC
Re: [CAD_CAM_EDM_DRO] Step resolution
JanRwl@A...
2004-08-19 17:25:43 UTC
Re: [CAD_CAM_EDM_DRO] Step resolution
Marv Frankel
2004-08-19 23:50:48 UTC
Re: Step resolution
JanRwl@A...
2004-08-20 11:20:28 UTC
Re: [CAD_CAM_EDM_DRO] Re: Step resolution