weird results when homing on large format machine

41 views
Skip to first unread message

Randy Zimmer

unread,
Oct 2, 2012, 4:59:50 PM10/2/12
to vintage-...@googlegroups.com
I have two robots made with Makerbot electronics
On the first, with a 18" Z height, the homing times out before it gets 3/4 of the way.
I just reversed the Z to MIN - it takes less time anyway.
On the new machine, I set calibration to 0,0,0, set Z to MIN as on the other.
(this thing is 500mm in every direction)
When the robot homes, it looks OK, then, the distances seem to double and the Z goes straight up!
I'm afraid to let it run that way very long, but it sure is weird.
When starting from 0,0,0 without homing, it runs just fine.
I don't need to fix this, I'm just wondering if it has happened to anyone else?
Boards are last generation 12V before replicator
RepG 34
Skein 35

Dan Newman

unread,
Oct 2, 2012, 5:19:42 PM10/2/12
to vintage-...@googlegroups.com
On 2 Oct 2012 , at 1:59 PM, Randy Zimmer wrote:

> I have two robots made with Makerbot electronics
> On the first, with a 18" Z height, the homing times out before it gets 3/4
> of the way.

If you're using the default homing command in the start gcode, then that's
to be expected: the timeout is too short. While the firmware itself accepts
a timeout and "feedrate" for the command, the gcode parser in RepG only accepts
a "feedrate" and uses a hard coded timeout value. The "feedrate" is actually
how many microseconds to pause between steps and thus the smaller the value
the faster the apparent motion (and thus it's inversely related to the speed as
its time units per step and not steps per time unit).

At any rate, the typical thing to do is to put in multiple homing commands.

> On the new machine, I set calibration to 0,0,0, set Z to MIN as on the
> other.

It's not clear to me what you mean by "set calibration to 0,0,0". Showing
your actual gcode would be helpful.

> (this thing is 500mm in every direction)
> When the robot homes, it looks OK, then, the distances seem to double and
> the Z goes straight up!

The distances double or the speed doubles or the steps seem to double?
At any rate, if you told it to move to Z min (down) and it moved up, then
it sounds like you need to reverse your Z axis stepper motor or invert the
Z axis in the firmware's EEPROM via the "Machine Onboard Preferences" panel
of ReplicatorG.

As to anyone else seeing this, the timeout on homing is common place even
with stock thing-o-matics. The axis running in the opposite direction is
also common and hence the ability to invert the axes from software or by
unplugging the wiring between the stepper and stepper driver and turing it
180 degrees and plugging back in.

Dan

Randy Zimmer

unread,
Oct 2, 2012, 10:55:42 PM10/2/12
to vintage-...@googlegroups.com


> On the new machine, I set calibration to 0,0,0, set Z to MIN as on the
> other.

It's not clear to me what you mean by "set calibration to 0,0,0".  Showing
your actual gcode would be helpful.

Hi Daniel,
What I meant was I used the standard calibration script but reversed the Z 162 (MAX) to 161 (MIN) and saved it.
So when I run a part, after the bot goes to all three MINs in homing, the distances double (or more!) and Z starts rising without a pause.
The speeds are normal but a 20mm distance is at least 40, maybe 80, too dramatic to measure in detail, I'm searching for the abort button.
Its pretty strange to see it go so far.

When I delete the homing and start the part from the center, everything is sized correctly and it runs as it is told.


Reply all
Reply to author
Forward
0 new messages