Add stepper backlash post.
authorW. Trevor King <wking@drexel.edu>
Fri, 19 Nov 2010 13:33:06 +0000 (08:33 -0500)
committerW. Trevor King <wking@drexel.edu>
Fri, 19 Nov 2010 13:33:06 +0000 (08:33 -0500)
posts/MultiMode/stepper/backlash.mdwn [new file with mode: 0644]
posts/MultiMode/stepper/backlash/step_backlash.png [new file with mode: 0644]

diff --git a/posts/MultiMode/stepper/backlash.mdwn b/posts/MultiMode/stepper/backlash.mdwn
new file mode 100644 (file)
index 0000000..69dbc77
--- /dev/null
@@ -0,0 +1,37 @@
+There is a good deal of drive-train [backlash][] in the stepper
+mechanism.
+
+Detection procedure
+===================
+
+1. Step towards surface until there is plenty of surface contact.
+2. Step away until you leave the surface.
+3. Repeat until you have too much data :p.
+
+[[!img step_backlash.png
+  alt="Data from 17 counterclockwise approach-retreat cycles"
+  title="Data from 17 counterclockwise approach-retreat cycles"]]
+
+Conclusions
+===========
+
+About 47 steps of backlash. Since we never touch the stepper while any
+critcal surface interactions are going on, simply replace *step back 1
+step* with *step back 60 steps, then forward 59*.
+
+Supporting evidence
+===================
+
+Noise on reversal
+-----------------
+
+When we reversing the stepper (by any means), the movement is “clearly
+audible” for “a bit” and then becomes “almost silent”.
+
+To give a rough idea, I usually have keep my finger on the drive-shaft
+to tell if the motor is moving in “almost silent” mode, but can hear
+the motor from around 5 ft away in “clearly audible” mode.  “A bit” is
+about 0.5 s when controlling the motor using the manual Up/Down
+switch.
+
+[backlash]: http://en.wikipedia.org/wiki/Backlash_%28engineering%29
diff --git a/posts/MultiMode/stepper/backlash/step_backlash.png b/posts/MultiMode/stepper/backlash/step_backlash.png
new file mode 100644 (file)
index 0000000..28dd30f
Binary files /dev/null and b/posts/MultiMode/stepper/backlash/step_backlash.png differ