Triac not holding offsets

boots · Post by **boots** » Tue 27 Feb , 2007 21:49 pm

Older Triac (July 1998) at Cal U, Pa. Machine was upgraded with Denstep card a few years ago and running VR Milling V.2.30. Feed rates are at Max 300 for X Y Z.
Triac does not hold offsets in X Y Z. Always off just a bit, up to 0.190 inch at Program position "zero". Will also get a 55 sec. time-out error when homing. Traic was checked out in Oct. 2006 was working just fine (home, jog, spindle).

Any Ideas?

Post by **Denford Admin** » Wed 28 Feb , 2007 10:28 am

If the machine is losing its position, it must be stalling.
What is the machine doing inbetween the first home, and the return to home (out of position) ?

The homing timeout is there to check that the home switches are working properly- are there any signs of the home switches sticking or not working ? Or is the home speed just too low to let the axes get from one end of travel to the home positions ?

boots · Post by **boots** » Tue 06 Mar , 2007 15:49 pm

It's not that the machine is not holding offsets, it's when you go to set the offsets.
When I jog the machine to the datum point on the billet and set the X ,Y, Z offsets then close the Offset window the X Y & Z numbers in the Control Panel window Program Position are not 0.000. They read 0.010, 0.056, 0.109 very small amounts, not zero.

Post by **Denford Admin** » Tue 06 Mar , 2007 16:55 pm

On this machine, you can expect an error of +/- 0.005mm as thats the resolution of the the steppers.

I can't say I have seen anything like this before

When you open the offset window, and datum the offset - the software reads the current position from the control and stores it in the offset file.
Whatever number is in this file, is then subtracted from the postion display.
So if you are at machine pos X-100.123 and zero the offset, -100.123 will be stored.
When the position readout calculates what to display, it will read machine pos -100.123 and subtract -100.123 (from the offset file)giving 0.000

The current position is being read continuously via the RS232, so I can only guess something is corrupt with the communications, or it is not reading the current position properly.
Try this:
- Set the X offset to 0.0 manually so the machine, and program positions both read the same.
- Jog to your billet touch on point
- Note down the X machine position from the control panel
- Manually type in the same position into the X offset
- Close the offset window
- The program display should now read 0.000 for the X axis

BTW, what is the control type parameter set to ? It should be 0

boots · Post by **boots** » Thu 15 Mar , 2007 20:03 pm

When setting the offsets, the machine is jogged over to the billet to set the XYZ zero, open the offset window, click on each axis to set the offset. The number in each offset line matches the machine position on the control panel, but the program position is off by very small amounts not all zeros. This will go to zero by just opening and closing the Offset window (lower left of VR Milling screen) 3 or 4 times, and not touching anything else.
This has to be done with every offset or we get an error message when a program is run. This is problem when running students thru with thier projects.
Any ideas on what to look for, software, wiring, mechanics?

Whers is this- BTW, what is the control type parameter set to ? It should be 0

Post by **Denford Admin** » Fri 16 Mar , 2007 10:29 am

Whers is this- BTW, what is the control type parameter set to ? It should be 0

Go to menu:
Setup -> machine parameters -> denny -> communications
Control type: in the parameters on the left should read 0 - if not then try this setting.
(If it is 7 or 9 then leave it as it is)

If you are seeing very small changes in position from 0.0000 then this would be normal - like I said, the resolution of the machine is down to how many stepper motor steps it takes to move an amount (eg 1mm)
If the stepper resolution is 200 steps per mm, then you will have a possible error of + / - 0.005mm (0.0002")

If this is causing problems because the machine then thinks it is out of limits, then you may have to move your workpiece in a little bit.

boots · Post by **boots** » Mon 19 Mar , 2007 14:16 pm

I am sorry, I do not believe you understand the problem.
The machine cannot establish PART Zero via offset icons on the computer display. We have turned the offset icon on....then set the XYZ aspect of this to zero via these icons.....these to not change to zero....they may be off by 3/16 - 1/4 inch......not .005 mm. Please do the math. It is not acceptable to MOVE the part. This cannot be accomplished when the part is clamped in a vice and the vice is bolted to the table.

Post by **Denford Admin** » Mon 19 Mar , 2007 17:25 pm

earlier you said:

The number in each offset line matches the machine position on the control panel, but the program position is off by very small amounts not all zeros. This will go to zero by just opening and closing the Offset window (lower left of VR Milling screen) 3 or 4 times, and not touching anything else.

and now say:

they may be off by 3/16 - 1/4 inch......not .005 mm

Is the display reading 3/16 - 1/4 inch or "very small amounts" when you datum the offset

Have you checked the control type parameter yet

I can only guess that a communication problem is stopping the software from reading the current machine position properly when you hit the offset button.

boots · Post by **boots** » Mon 19 Mar , 2007 19:05 pm

It varies. Everytime it is different, just a few .000tenth's up to a couple .hundreth's, not the same on each axis. Never know what will come up on each readout always random.

boots · Post by **boots** » Tue 20 Mar , 2007 20:37 pm

CalU is scheduling as service call for tis TRIAC so any help on what to look at on the machine would be appreciated.

Thank you.

Post by **Denford Admin** » Tue 20 Mar , 2007 23:04 pm

I'm 90% sure its to do with a communication problem

The machine position is read from the control through the RS232 every 200ms or so.
If there is a problem with the comm port, port buffers, RS232 cabling etc.... then this could cause the position reading to be corrupt, or "out of date" (ie the software is only using the last valid reading)
The control type could be an important part of this as control type 5 had communication problems with NT / win2K / XP machines and was fixed by control type 0

Try the usual RS232 checks:
viewtopic.php?t=544

Also try setting stop bits to 2 in the communication parameters.

Check VR Milling is the latest version.

It would be advisable to take a spare Baldor control card, and a known working laptop / PC to rule out the PC or hardware.

Good luck

Denford Software & Machines

Denford Software & Machines

Triac not holding offsets

Triac not holding offsets