Re: asynMotor: NewFocus 874x Picomotor Controller Driver

[Angelic Ebbers <aebbers@gemini.edu>]

Yes .. that was definitely the problem .. Thanks!!

On Wed, Mar 28, 2018 at 1:54 AM, Mark Rivers <rivers@cars.uchicago.edu> wrote:

Hi Angeleic,


The only error I could spot in your posting was this:


>     file "db/basic_asyn_motor.db"
>     {
>     pattern
>     {P,N,M,DTYP,PORT,ADDR,C,S,DESC,EGU,DIR,VELO,VBAS,ACCL,BDST,BVEL,BACC,MRES,PREC,DHLM,DLLM,INIT}
>     {lis:,1,   "m$(N)","asynMotor",MOT1,1,0,0,"Rot",counts,Pos,1750,500,1,0,0,50,1,0,0,0,""}
>     }
>



The substitutions line there should be M0, not MOT1.   You need to specify the name of the asyn port for the motor driver, not the TCP driver.


>     {lis:,1,   "m$(N)","asynMotor",M0,1,0,0,"Rot",counts,Pos,1750,500,1,0,0,50,1,0,0,0,""}

Mark



________________________________
From: tech-talk-bounces@aps.anl.gov <tech-talk-bounces@aps.anl.gov> on behalf of Torsten Bögershausen <torsten.bogershausen@esss.se>
Sent: Wednesday, March 28, 2018 1:27 AM
To: Angelic Ebbers; tech-talk@aps.anl.gov
Subject: Re: asynMotor: NewFocus 874x Picomotor Controller Driver

Hej Angelic,
I'm pretty sure that the usage of MOT1 is the problem.
You need 2 different asyn ports:
- one to communicate via TCP/IP to the motion controller
- the second to allow the generic model 3 driver (devMotorAsyn)
   to access the C++ classes in asynMotorController.cpp and
   asynMotorAxis.cpp

This is what I use:
#One:
epicsEnvSet("IPADDR",        "127.0.0.1")
epicsEnvSet("IPPORT",        "5000")
epicsEnvSet("ASYN_PORT",     "MC_CPU1")

#two:
epicsEnvSet("MOTOR_PORT",    "MCU1")

#And then for one:
drvAsynIPPortConfigure("MC_CPU1","127.0.0.1:5000",0,0,0)
asynOctetSetOutputEos("MC_CPU1", -1, ";\n")
asynOctetSetInputEos("MC_CPU1", -1, ";\n")

#And for two:
EthercatMCCreateController("MCU1", "MC_CPU1", "32", "200", "1000")

EthercatMCCreateAxis("MCU1", ...)

dbLoadRecords("./db/EthercatMC.template", "PREFIX=IOC:, MOTOR_NAME=m1, R=m1-,
MOTOR_PORT=MCU1,  AXIS_NO=1, DESC=motor1 ")

# And the template looks similar to this:
record(motor,"$(PREFIX)$(MOTOR_NAME)")
{
        field(DESC,"$(DESC)")
        field(DTYP,"asynMotor")
        field(OUT,"@asyn($(MOTOR_PORT),$(AXIS_NO))")
         [snip]
}


#########
# (end of blabla)
So in your code
  nf874xCreateController("M0", "MOT1", 1, 200, 1000)

"M0" is "MCU1" in my example
"MOT1" is "MC_CPU1" in my example

And with that knowledge you should be able to use "M0" and "MOT1"
in the right way.

On 28/03/18 06:55, Angelic Ebbers wrote:
> Hi,
>
>      I'm working on supporting a New Focus 8743-CL controller at Gemini
> Observatory using EPICS/RTEMS.  I was thrilled to see that there already is a
> driver for the 8742 model controller in the motor distribution.  From my
> initial reading, I'm thinking there will be minimal changes needed to support
> closed loop control.
>       I've been trying to set up the driver, connected to the 8743, expecting
> that most functionality is the same as the 8742.  I seem to be doing something
> wrong, however, since the IOC crashes during/shortly after iocInit .. appearing
> to encounter problems during init_record.  The errors I'm getting seem to
> indicate a problem with how I'm setting up the driver support.  I'm following
> the example in "iocBoot/iocNewFocus874x".
>
>       Here are the errors I'm getting:
>
>     lis:m1Offset devAsynFloat64::initCommon findInterface asynFloat64Type
>     Bad init_rec return value  PV: lis:m1Offset ao: init_record
>
>     lis:m1Resolution devAsynFloat64::initCommon findInterface asynFloat64Type
>     Bad init_rec return value  PV: lis:m1Resolution ao: init_record
>
>     lis:m1Direction devAsynInt32::initCommon findInterface asynInt32Type
>     2018/03/27 17:09:45.730 devMotorAsyn::init_record, lis:m1 find int32
>     interface failed
>
>
>       I've done some digging on "devMotorAsyn::init_record find int32 interface
> failed" and found several TechTalk threads about trying to set up model 1
> drivers as model 3 (using asyn).  The specific problem in those cases was
> specifying DTYP as "asynMotor". (link to thread:
> https://epics.anl.gov/tech-talk/2013/msg02425.php
> <https://epics.anl.gov/tech-talk/2013/msg02425.php>)
>        I'm pretty sure I'm setting up the driver with the correct approach
> though and find no other clues as to the root of the problem.
>
> Here is my motor substitution file:
>
>     file "db/basic_asyn_motor.db"
>     {
>     pattern
>     {P,N,M,DTYP,PORT,ADDR,C,S,DESC,EGU,DIR,VELO,VBAS,ACCL,BDST,BVEL,BACC,MRES,PREC,DHLM,DLLM,INIT}
>     {lis:,1,   "m$(N)","asynMotor",MOT1,1,0,0,"Rot",
>     counts,Pos,1750,500,1,0,0,50,1,0,0,0,""}
>     }
>
>
> relevant portion of my startup script:
>
>     ### Motors
>     # Setup IP port for 8742
>     drvAsynIPPortConfigure("MOT1", "10.1.60.111:23 <http://10.1.60.111:23/>")
>     asynOctetSetInputEos("MOT1",0,"\r\n")
>     asynOctetSetOutputEos("MOT1",0,"\r")
>
>     #db asyn debug traces
>     asynSetTraceMask("MOT1",0,0x9)
>     asynSetTraceIOMask("MOT1",0,0x2)
>
>     nf874xCreateController("M0", "MOT1", 1, 200, 1000)
>
>     ## Load record instances
>     dbLoadTemplate("data/newfocus8742.asyn.motor.substitutions")
>
>     dbLoadRecords("${ASYN}/db/asynRecord.db",
>     "P="lis:",R=Asyn,PORT=MOT1,ADDR=,OMAX=80,IMAX=80")
>
>
>
> I've done a bunch of checking to ensure that my dbd has all of the asyn/motor
> support built in .. It seems ok.
> Here's the Makefile component:
>
>     gis-mk-ioc_DBD += motorSupport.dbd
>     gis-mk-ioc_DBD += devNewFocus.dbd
>     gis-mk-ioc_DBD += devSoftMotor.dbd
>     gis-mk-ioc_DBD += motorSimSupport.dbd
>     gis-mk-ioc_DBD += asyn.dbd
>     gis-mk-ioc_DBD += drvAsynSerialPort.dbd
>     gis-mk-ioc_DBD += drvAsynIPPort.dbd
>
>     gis-mk-ioc_LIBS += NewFocus
>     gis-mk-ioc_LIBS += softMotor
>     gis-mk-ioc_LIBS += motorSimSupport
>     gis-mk-ioc_LIBS += motor
>     gis-mk-ioc_LIBS += asyn
>
>
>
> Cheers,
> Angelic Ebbers
>
>