View Single Post
  #64   Report Post  
Posted to rec.crafts.metalworking
Gunner Asch[_6_] Gunner Asch[_6_] is offline
external usenet poster
 
Posts: 10,399
Default Take apart - put together syndrome

On Sat, 28 Aug 2010 00:00:03 -0400, wrote:

On Fri, 27 Aug 2010 05:32:33 -0700, Gunner Asch
wrote:



Try it in my line of work..fixing high precison machine tools..CNC.

Client will call..telling me that out of 50 parts...2 of them are .010
off..the rest are perfect. And they take 2 minutes each to run. And its
random....

So I have to watch the ****ing machines for a minimum of 100 minutes
just to see if I can catch what the hell is going on. This after doing
a mechanical checkout that takes 15 minutes minumum..then tuning the
control, pulling all that sheetmetal off, checking couplers, ballscrews
etce etc etc.

Can **** away a whole day doing that. And you cant bill the client for
all that time either.


Gunner


I feel your pain, Gunner.
Client had (still has, I believe) a standard modern CNC lathe - would
crank out 100 good parts, then 5 parts 2 thou small, then 7 parts 4
thou small, then another 2 or 3 2 thou small then 105 good ones then 1
2 thou big, and 1 4 thou big, and then another 50 good ones, then50 or
more between 2 and 4 small.

You get the picture.
Darn thing used "resolvers" instead of quadrature encoders and they
were really finicky as far as noise in the lines. We ran sheilded
cable, did all kinds of tricks with grounding, separated the feed to
the computer from the feed to the motors, used isolation transformers,
and in combination finally got it down to the point he could get 1000
good parts out of a run of 1025.

I spent HOURS on that thing, a couple hours a day, a couple days a
week, for a period of several months.
He still owes me a few hours of free machining!!!!


Chuckle...in this case..it was the welding department. Seriously.

To hook up this Omniturn..they ran 100+ feet of #10!! wire from a
breaker panel in the welding department..along with another run of 14ga
from the same breaker panel for the control. The control powers the
servos.....

When they would fire up the grinder..hand grinder..on the same
circuit...it would spike the control..100' away.

The spike would only "occasionally" make it into the control.....shrug

Took me 2 days to figure it out...heard the grinder..saw a tiny flicker
of the control video out of the corner of my eye..measured the part...it
was off. Had them bury the grinder in raw stock over the next dozen
parts....all were off.

Whew...!

Gunner


I am the Sword of my Family
and the Shield of my Nation.
If sent, I will crush everything you have built,
burn everything you love,
and kill every one of you.
(Hebrew quote)