Note on recent problems with the Anorad table
There was an Anorad failure in early January 2005 which might repeat.
The symptoms were that SD1:
a) would not move
b) felt spongy when locked
c) the HG1- line on pin 7 of the Hall Sensor cable was at -.65V
when disconnected from the Anorad Control/Power chassis.
It was somehow shorting to the tank.
This is the same as 22Z on the extender card and "c" on the 55 pin
vacuum connector.
The Problem vanished when the cable was moved for examination
by removing the 4" flange near the exit tubes. It could have been
a splinter piercing the jacket of the wire, a stray wire clipping
could also have fallen on the exposed connection on the cart and
shorted the line to the cart and tank. If it happens again
we have a better guess of where to start fixing the problem and
hopefully we can find the exact problem area and correct it.
The tank used to be floating and is now grounded to the electrical conduit.
Marc Lacasse
2/1/2005
======================================================================
### 2005-01-08, PASch, MGL, EP, Julien Coyne
### 2005-02-01
#
Short Delay Line 1 Malfunction
==============================
Symptoms
--------
* while operating the ICS, SD1 would not move, error signal acts as if
cart is stuck
* in open tank,
- if laser is locked,
SD1 feels spongy when pushed,
when pressing mechanical limit switches without blocking metrol. laser,
both carts move quite easily apart,
yet still same behavior towards each other
- if laser is unlocked,
SD1 still has considerable current flowing through coils,
has prefered positions;
Does "prefered positions" suggest that Hall sensors do not
commutate motor coil current properly?? Not sure about conclusiveness...
Checks
------
* measured all poss. permutations of mutual resistance between connection
lines in Hall cable (DB9) from Anorad rack to tank/cart;
connector on cart is plugged onto Hall circuit board; did that for both SDLs;
value table (in kOhm):
func: IC+ HG1+ HG3+ 24D/Z IC- HG1- HG3-
Eltr: 28B 22D 24B x shield 28D 22Z 26Z x
DB9: 1 2 3 4 5 6 7 8 9
SD2 \\ SD1
1 | \\ 0.75 2.3 - - 2.85 0.78 2.45 -
2 | 0.78 \\ 2.4 - - 2.9 1.0 2.5 -
3 | 1.5 1.5 \\ - - 1.3 2.4 1.05 -
4 | - - - \\ - - - - -
5 | - - - - \\ - - - -
6 | 1.4 1.5 0.75 - - \\ 2.95 1.5 -
7 | 0.76 1.0 1.5 - - 1.5 \\ 2.5 -
8 | 1.5 1.5 1.0 - - 0.76 1.5 \\ -
9 | - - - - - - - - \\
[ where three digits given, two measurements with opposite polarity
differed by 1 in the 2nd significant digit, e.g., 1.1 and 1.2 render as
1.15 ]
--> no obvious large difference; greatest difference 2.85kOhm vs. 1.5kOhm
many SD1 connections are ~1kOhm larger then corresp. SD2 conn.
* disconnected Hall sensor on cart1
measured resistances between lines (like above) and incl. Ground
with extender card in Eltrol slot on front end of Anorad electronics:
everything open, except for connection between line 7 [22Z, Hall line HG1-]
and Ground: -0.65V_dc; (maybe caused in back plane?)
checked: tank to ground -0.65V
track to ground -0.15V
* measuring wires in Hall cable from Anorad rack connector to connector on
Hall sensor board on cart1:
lines 1, 2, 3, 6, 7, 8: 0.006Ohm to 0.007Ohm
line 5 : shield
--> reconnected Hall sensor on cart1
* swapping various cables, i.e., connectors at rack:
motor current and Hall together: error stays with SD1 cart
limit swap: no change, no effect;
suspecting possible ground loop through limits?
swap all cables (metrol., motor, limit, Hall), drive carts with ICS:
cart 2 works fine, cart1 not working properly (same failure mode as above)
--> returned all cables to correct connectors
swap motor, but not Hall cables, i.e., motor on cart1 controlled by
Eltrol circuit2 and v.v.:
sponginess moves from cart1 to cart2 which would not lock;
pushing cart1 causes cart2 to oscillate,
pushing cart2 causes cart1 to move <= 1mm
swap Hall, but not motor cables, i.e., ...:
sponginess moves from cart1 to cart2, i.e., same effect as just before;
--> since we do know that motor2 works fine,
suspicion is that fault is in Hall sensor
* measured SD1 coil currents:
not switched off completely when laser unlocked, which
normally should be the case
* checked various limit switches (home, soft limit, hardware switch)
by moving carts manually and observing LED panel:
--> seems okay, and both carts behave the same way
* from previous knowledge:
IC+/- on Hall cable carries error/command signal from metrology
which is directed through the Hall sensors of a cart
--> set up DB9 probe cable and oscilloscope
to look at different signal pairs IC+/-, HG1+/-, HG3+/-
in Hall cable for each cart
* on Hall cable, when pushing locked cart,
the following static voltages appear:
SD1 SD2
IC+ 11V 12V
IC- 14V 12V
HG1+ 11V 12V
HG1- 11V 12V
HG3+ 7V ??
HG3- 7V ??
* in some of the following notes, when looking at the two phases of a signal,
e.g, IC+ and IC-, at the same time, then two different cases can be observed
when moving the carts:
"+" and "-" move _in_ phase (i-ph): _ _
\_/
_ _
\_/
"+" and "-" move _counter_ phase (c-ph): _ _
\_/
_
_/ \_
whereby positive or negative slopes/edges occur
when direction of the cart movement is reversed.
Amplitudes of voltage changes are quite large, whereas in the flat regions
there are additional (sinusoidal, ~~~~~ ) wiggles present
with comparatively small amplitude.
Let's call the large amplitudes A, the small amplitudes a.
_ _ _ _
\_/ \_/
_ _ _
\_/ _/ \_
~~~~~~~~~~
* unplug motor coils of each cart, leave Hall cables plugged in
(probed by DB9 cable), determine A for each cart when moving:
SD1 SD2
IC+ 13V 10V _no_ perceivable a
IC- 14V (c-ph) 10V (??)
HG1+ 8V 2V small a
HG1- 8V (??) 2V (??)
HG3+ 8V 3V small a
HG3- 8V (??) 3V (??)
* idea: switch metrology cables SD1 <-> SD2
to disentangle metrol. signal (locked) from Hall signal
when moving only one cart at a time;
have motor coils disconnected for that ("open loop" operation)
= looking at signals on cable from SD1
(A resp. a omitted if not perceivable):
moving SD1 moving SD2
IC -- A=17V (c-ph)
HG1 a=0.5V (c-ph) A= 8V (i-ph)
HG3 a=0.5V (c-ph) A= 8V (i-ph)
= looking at signals on cable from SD2
(A resp. a omitted if not perceivable):
moving SD1 moving SD2
IC A=10V/7V not balanced! (c-ph) --
HG1 A=2V (i-ph) a=0.5V (c-ph)
HG3 A=6V (i-ph) a=0.5V (c-ph)
--> unplug Hall cable of SD2:
moving SD1, SD2-IC shows A=21V (c-ph)
* plug metrology cables back to correct channel, unplug Hall cables:
moving SD1, SD1-IC shows A=21V (c-ph)
moving SD2, SD2-IC shows A=21V (c-ph)
* With correct metrology, but switched Hall cables and moving SD1,
SD1-IC shows A=8V (c-ph).
With correct metrology _and_ correct Hall cables and moving SD1,
SD1-IC is unbalanced A=12V/20V (c-ph)!
* Hum, unbalanced signals...?
--> unplug 60-pin ribbon cable from cart1,
measure some potentials/resistances:
SD1 SD2
IC+/P1 to Ground 1.23V 0V (very high resistance, both ways)
IC-/P6 to Ground 1.17V floating
cart to Ground floating floating
Tank to Ground: 1V
SD1
IC+/P1 to Tank 0.8kOhm
IC-/P6 -"- 3 kOhm
HG1+/P2 -"- 1 kOhm
HG1-/P7 -"- 4 Ohm <-- !!
HG3+/P3 -"- 2.5kOhm
HG3-/P8 -"- 2.6kOhm
Followed HG1-/P7 to pin "c" on 55-pin connector on flange to tank.
Took flange off for inspection, but did not see anything out of the ordinary.
Yet handling and moving the cables inside and outside the tank
somehow made the problem go away, i.e, resistance between HG1-/P7 and tank
was in the kOhm range again...