Re: Z axis Offsets

[Ray <rehenry-at-up.net>]

Hi Tim

On Saturday 29 September 2001 23:42, Tim wrote:
> Ray,
<s>
> Also found that the handling of offsets is not as solid as I first
> though in xemc. I truthfully wonder how anyone is doing any real
> machining with the current builds? It just can't seem to keep straight
> what mode it is in and I feel very luck to not have ruined a part or
> broken a tool when it would be displaying something like X1.5 and I
> enter a X1 and the machine moves way more than .5 inch because it
> decided I wanted to be a X1 absolute, not relative to the offset that is
> displaying.

The release that you are using is experimental.  There are a lot of 
"features" that were added right after the first BDI release that I was 
trying to move to some BDI powered machines for test.  Although I don't 
see quite the same set of problems here, your posts really help get to the 
heart of problems.  There are a couple possible sources of problems like 
yours that I am trying to get a handle on.

To me g92 is a bastard child left laying around as a part of some 
experiments with features based programming.  IMHO, it was intended to be 
the set position command at the end of a series of probes for the position 
of the part to be machined.  The intent was to make the control smart 
enough to find the part and its orientation and then apply the ngc program 
to it. 

1 - I wonder if you'd try running using the standard coordinate systems 
rather than g92.  Or is there some difference between them that is 
essential to your operation?  

I put a small script (Set_Coordinates) under that menu that will let you 
locate your start position manually or however you like, select the 
coordinate system to use, and then press the set pos button and it will 
record the current values.  These go directly into the var file and are 
then read into the interpreter and task stuff. 

2 - A couple people have posted a problem with var numbers being 
interpreted as metric on first pass rather than being read using the 
currently selected units.  When you get the goofy kinds of moves like you 
describe, think about the possibility that they might be the mm version of 
the value that you entered.  

> Maybe I am just confused how I should be using EMC now. I never had this
> problem with the older builds for the RH 5.2. Can anyone give me some
> advise how to actually use the current build for really making parts
> before I nuke the install and just go back to a stable version?

All those in favor of Tim being confused, raise your hand...  I don't see 
any hands.  

The EMC in BDI-2.10 is fairly conservative, more so than my tgz.  Paul did 
build in the larger buffers so that it doesn't crash nearly as often.  But 
his leaves out the settable debug and the units stuff so it might work 
better for you than mine.  I believe that all you would need to do is run 
from the /usr/local/emc directory using the ini you wrote for bridget.

Hope this helps.

Ray