odd behavior on haas minimill

Every couple of hours of operation one of the axes's move .0004. I mean exactly .0004.
So I make a .0004 move and they come out deadly for a couple of hours, then
bam...0004 out.
The Z axis and the X axis don't move even a tenth in 10 hours of full rpm operation and 100% rapid. I'm figuring the 4 tenths is a metric thing.
Anybody ever run into something like this before?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

also, my oiler seems to only work manually, and sometimes I pull it in the begining of the day, sometimes not. Could oil make the thinng shift location, then stay there perfectly for hours then shift again?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

Sounds like heat build up
Gunner
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Thu, 02 Oct 2008 02:56:06 -0700, Gunner

which could be due to lack of oil.
is the move always plus or always negative?
when you start the next day do you need to shift it back?
Thank You, Randy
Remove 333 from email address to reply.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Today I came in, warmed up the spindle, ran the program empty a few times to warmup and cut a part. Had to make a .0008 move in Y+. It's always Y, and when it jumps its .0004, seems random +-. Never ventures more than .0008 from original number.
Im experimenting with pulling the oiler every three hours. See what happens.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I wonder if it's a backlash compensation variable that's been activated- double check all the controller settings and make certain they are correct
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Oct 2, 7:45am, snipped-for-privacy@gmail.com wrote:

Call your HFO, it is a thermal comp parameter, they can either walk you through it, or have a tech take care of it for you.
"D"
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

Call your HFO, it is a thermal comp parameter, they can either walk you through it, or have a tech take care of it for you.
"D"
I looked in the settings and all the thermal control settings were zeros. If I could predict when this thing would move I'd use that to fix it. But it's random. Who knows, maybe at night the power company changes the voltage? Sometimes 5 volts makes a difference in how the computer thinks. It changes the math. Or its the oil. Hell I dont know. I do know its level and sturdy. Maybe one axis isnt getting any oil, its only got hoses feeding it proberbl.y The mill was only 30,000, prolly used the cheapest hoses china would sell.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Chip in the timing belt between the Y motor and the ballscrew.
--
Anthony

You can't 'idiot proof' anything....every time you try, they just make
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

Possible, but the chip would have to be pretty much the same size every time
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Nah, just create the same tension increase in the belt.
--
Anthony

You can't 'idiot proof' anything....every time you try, they just make
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
vinny wrote:

yea, I had a leak in the way oil system that was alarming out the machine. All the hoses (in a VF-4) are either plastic, or copper tubing.
turned out that one of the plastic hoses for the X axis ways was cut a bit short, so it eventually weaseled its way out of the track, and got caught on the nut on the ballscrew.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I pumped the oiler by hand every 2-3 hours for 12 hours strait, worse move I had to make was .0002 all night. Thats like thru 20 cycle runs. I'm feeling pretty confident its oil based, question is why.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

Sometimes
math.
sell.
Put a chiller on your coolant tank and when not in cycle always leave it running in a warmup program.
--




Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

No coolent tank. And the z repeats perfect, so no need to keep thatwarm. As far as cutting, I try to time the minutes before cycle starts as consistant as I can. Seems to work great.
But the 4 tenths seems an aweful like .0003937. Is that one turn of the screw?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Not even close. Probably near the least resolved increment though:)
Actually I like the thermal compensation parameter answer the best.
--

Dan

CNC Videos - <http://tinyurl.com/yzdt6d
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

thermal growth isnt a huge problem in most modern tools--( the lower bearing set is captured axially and the upper set is allowed to float )

NO
Measure three places along the axis ( front center rear) I would bet that error is less when the ball nut is closest to the thrust bearing.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Check backlash before and after the shift at the same place on the screw. I'm wondering if you have an axis bearing pair that is moving.
How far does on mark on your encoder move the axis?
Is the drive motor bolted down firmly?
Wes
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
snipped-for-privacy@lycos.com wrote in wrote:

Good point.
I've seen a loose ball nut and a loose servo motor both cause similar problems. But in both those cases it would jump fairly random like.
--

Dan

CNC Videos - <http://tinyurl.com/yzdt6d
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

Im too lazy for that one.

But this I can check. Thanks.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.