Re: How much accuracy is enough?
Posted by
Ken Jenkins
on 2002-03-12 00:35:03 UTC
Chris raises some interesting questions regarding "how much is enough"
and I have been asking myself some of those questions lately. Not just
in relationship to CNC but in terms of general methods of work. Overall
I am finding I err in many projects being too accurate strangely enough.
I suppose the response might be, "How can you be too accurate?!! Surely
it's better to err on that side?" True if you have all the time + $ in
the world. I don't. All the additional time (and money in some cases) I
spend on accuracy which is not required is wasted except in some
sort of "inherent-reward for craftsmanship" or fulfilling some "aesthetic
sensibility". We're not talking about types of commercial work where
the QC requirements of the customer dictate regardless of "real issues"
(translate: "We're GM! .... and you're not ... end of discussion.")
The more accuracy you pursue, the more money and time you spend to
achieve it, in the end does it really make a difference in terms of
the functionality of the end product? (don't start your response to
this with a lecture about pistons and cylinders, I know it depends
on what you're making :-).
The above leads up to the opposite side of another question I posted
regarding "true cost of the ideal conversion project". How cheaply
can someone get into this hobby/work and still be productive and make
"real stuff"? Again, I know the above leaves out a lot in terms of criteria
(required part fits related to function, materials typically machined,
repeatability for multiple interchangeable parts, etc.).
I have another reason for asking this in that I have a friend who works
on footbridge and well projects in Kenya. It is amazing the effect that
the implementation of hi-tech principals and materials in lo-tech, low
cost scenarios can change people's lives. I'm wondering about a
"cottage industry" implementation of lo-cost CNC set-ups in providing
opportunities there and a plasma table made of mainly off the shelf
components might be just the thing.
Ken Jenkins
kjenkins@...
. deleted
.
. deleted
.
and I have been asking myself some of those questions lately. Not just
in relationship to CNC but in terms of general methods of work. Overall
I am finding I err in many projects being too accurate strangely enough.
I suppose the response might be, "How can you be too accurate?!! Surely
it's better to err on that side?" True if you have all the time + $ in
the world. I don't. All the additional time (and money in some cases) I
spend on accuracy which is not required is wasted except in some
sort of "inherent-reward for craftsmanship" or fulfilling some "aesthetic
sensibility". We're not talking about types of commercial work where
the QC requirements of the customer dictate regardless of "real issues"
(translate: "We're GM! .... and you're not ... end of discussion.")
The more accuracy you pursue, the more money and time you spend to
achieve it, in the end does it really make a difference in terms of
the functionality of the end product? (don't start your response to
this with a lecture about pistons and cylinders, I know it depends
on what you're making :-).
The above leads up to the opposite side of another question I posted
regarding "true cost of the ideal conversion project". How cheaply
can someone get into this hobby/work and still be productive and make
"real stuff"? Again, I know the above leaves out a lot in terms of criteria
(required part fits related to function, materials typically machined,
repeatability for multiple interchangeable parts, etc.).
I have another reason for asking this in that I have a friend who works
on footbridge and well projects in Kenya. It is amazing the effect that
the implementation of hi-tech principals and materials in lo-tech, low
cost scenarios can change people's lives. I'm wondering about a
"cottage industry" implementation of lo-cost CNC set-ups in providing
opportunities there and a plasma table made of mainly off the shelf
components might be just the thing.
Ken Jenkins
kjenkins@...
> Message: 13.
> Date: Mon, 11 Mar 2002 18:08:09 -0600
> From: Chris L <datac@...>
> Subject: Re: New Member to be has interest in home-brewing plasma table...
>
> Short and Sweet Version.............
>
> I Built my First CNC Plasma machine on the cheap because I needed something
> quick. I used Cold Rolled Steel for Linear Rails and simple Cam Followers to
> slide the carriages on it. Motion was provided by some very small stepper
> motors using Belt drive to Rack and Pinion.
>
> Having Built a very qualified CNC Router already with the "more appropriate
> components", I was a bit hesitant to use such cheap Linear Motion Techniques.
> But, After using it a while, I am not sure that I would do it much different
> next time.
. deleted
.
> Additionally, I gave some consideration to the necessary accuracy of a Plasma.
> Cutter. We're not talking Laser here..... The accuracy of any project you
> typically use a plasma cutter for does not come into tight tolerances. I've
> made some components for Automotive applications that just simply do not
> matter if they are off a few thou. This is true especially if the components
> get welded together. If the edge is smooth but off as much as .030 no one
> would know. I've never measured anything off that much even with the cold
> rolled rails.
. deleted
.
> Granted, My machine only cuts a 4' X 4' area. so if your building something
> much longer you would have to make sure your rails are straight.
>
> I would also use the Rack and Pinion again...... No way any type of screw
> could last a long time in this application. As long as the rack is placed on
> edge so any Swarf can fall out you are set. Amazingly, I only use some 80oz
> motors. She'll run as fast as 400ipm using a 36 volt supply. Obviously, a
> little bigger motor couldn't hurt but these amaze me.
>
> Finally, I only have it set up for 2 axis. I can see that next time I would
> plan for a 3rd Z axis and some type of software arc control. You do need to
> have the material pretty straight and level to get a good job. Sometimes that
> isn't so easy.
>
> So, as crude as it sounds, Plasma is one application where doing it the
> "cheap" way just might be an excellent way for the hobby low volume user.
>
>
> My experience in a nutshell......................
>
> Chris L
Discussion Thread
Ken Jenkins
2002-03-12 00:35:03 UTC
Re: How much accuracy is enough?
Art Eckstein
2002-03-12 14:09:14 UTC
Re: [CAD_CAM_EDM_DRO] Re: How much accuracy is enough?