lin-epics-modules/motorBase
0
0
Fork 0
forked from epics_driver_modules/motorBase
Code Pull Requests Activity

Copied bug fix from R3.13_branch.

Browse Source 
Added a final poll of motor status if drvMM4000ReadbackDelay is
non-zero.  This is required to work around a bug in the firmware
(versions 2.40 and 2.44) that the motor is reported as done moving
on the first poll after a move is begun with the motor power off.
This commit is contained in:
Ron Sluiter
2003-11-07 22:24:21 +00:00
parent 33eb1357d4
commit 2763a78a7b
1 changed files with 18 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
motorApp/NewportSrc/drvMM4000.cc
+18 -2
View File
@@ -2,9 +2,9 @@
FILENAME... drvMM4000.cc
USAGE... Motor record driver level support for Newport MM4000.
Version: $Revision: 1.3 $
Version: $Revision: 1.4 $
Modified By: $Author: sluiter $
Last Modified: $Date: 2003-05-28 14:57:08 $
Last Modified: $Date: 2003-11-07 22:24:21 $
*/
/*
@@ -47,6 +47,10 @@ Last Modified: $Date: 2003-05-28 14:57:08 $
* - set_status() sets RA_PROBLEM along with CNTRL_COMM_ERR to terminate
* node.
* .07 05-19-03 rls Converted to R3.14.x.
* .08 11-04-03 mlr Added a final poll of motor status if drvMM4000ReadbackDelay is
* non-zero. This is required to work around a bug in the firmware
* (versions 2.40 and 2.44) that the motor is reported as done moving
* on the first poll after a move is begun with the motor power off.
*/
@@ -350,12 +354,24 @@ STATIC int set_status(int card, int signal)
* motor is done moving, which means that the "jerk time" is done. However,
* the axis can still be settling. For now we put in a delay and poll the
* motor position again. This is not a good long-term solution.
*
* Another TEMPORARY FIX, Mark Rivers, 4/13/03. The MM4000 has reported that
* the motor is done moving. However, if the motor power was off, and the firmware
* version is 2.43 or 2.44 then the controller "lies" and says the motor is done
* on the first poll, when it is really moving. We work around this by reading
* the status again after the delay, in case it is really still moving.
*/
if (motor_info->pid_present == YES && drvMM4000ReadbackDelay != 0.)
{
epicsThreadSleep(drvMM4000ReadbackDelay);
send_mess(card, READ_POSITION, (char) NULL);
recv_mess(card, cntrl->position_string, 1);
pos = signal*5 + 3; /* Offset in status string */
mstat.All = cntrl->status_string[pos];
if (mstat.Bits.inmotion == true)
motor_info->status &= ~RA_DONE;
send_mess(card, READ_POSITION, NULL);
recv_mess(card, cntrl->position_string, 1);
}
}
else