Experimental Physics and
Industrial Control System

Ron Sluiter <[email protected]> · Mon, 02 Jun 2008 12:10:12 -0500

Hello Lifang,

I was planning on notifying the EPICS community of a recently discovered
problem with the MAXv device driver.
Although it appears that this is NOT the problem you are having, I will
take this opportunity to communicate the driver
problem.

There is a problem with the MAXv device driver when it is configured for
more than one board, and, either A24 or A32
addressing is specified. The driver is not sizing the address space
occupied by each MAXv correctly. MAXv users
should use A16 addressing until this is fixed in the next motor record
release (R6-4) or the next bugfix release.

Lifang,

You can confirm the MAXv configuration by executing the command,

dbior "drvMAXv"

at the VxWorks prompt. You should get something like the following,

Driver: drvMAXv
Oms MAXv motor card #0 @ 0xfbfff000, id: MAXv-8000 ver:1.29, s/n:000122,
FPGA:B1:A6 BOOT:1.1 - Oregon Micro Systems

The VxWorks BSP I use maps the A16 address space as follows, 0xFBFF0000
- 0xFBFFFFFF => 0x0000 - 0xFFFF

Hence my MAXvSetup(1, 16, 0xF000, 190, 5, 10) results in the MAXv base
address; 0xFBFF F000.

I don't know the AB ControlLogix 5561, but you will have to make sure
that it is not using
A16 memory from 0xFBFF F000 -> 0xFBFF F100. Maybe you can use the
"dbior" command
without any argument to determine this?

Hope this helps,
Ron

Zheng Lifang wrote:
> Dear all,
> For some reasons, I want to combine the source codes of AB
> ControlLogix with the ones of motorR-6-2-2 in one VME/MVME5500.
> The hardware environments are:
> VME/MVME5500
> MAXv-8000
> AB ControlLogix 5561
> The software environments are:
> Tornado2.2/VxWorks5.5.1
> EPICS base 3.14.8.2
> ether_ip.tgz (from SNS:EtherIP link)
> motorR6-2-2.tar.gz
> I have tested with the following:
> 1. If IOC runs ether_ip or motorR6-2-2 individually, everything works OK.
> 2. After combining these two kinds of source codes together, there are
> some strange things happened:
> 1) Usually MAXv-8000 can be initialized normally, but there are some
> problems with PLC. All PVs exists in IOC, but the displayed values on
> EDM/OPI are zeros.
> 2) Occasully I made a mistake: the address of MAXv-8000 is set to
> 0xE000, but I use "MAXvSetup(1,16,0xF000,180,6,10) in st.cmd. So
> MAXv-8000 can't be initialized, but this time the EPICS system of PLC
> works OK.
> I am not sure if there is something conflict with these two source
> codes. Any suggestion and idea are welcomed.
> Thanks&Best Regards,
> Lifang

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System