[Date Prev][Date Next][Date Index]
Re: OMS VME58 anyone?
Tony:
I have purchased 3 of these VME 58-8 (8-axis stepper motor controller).
To date, each of them can work individually under EPICS.
I have used them to drive the full complement of 8-axes simultaneously.
We'll be running 3-7 of these cards per VME crate on our beam line.
We'll have 3 or four VME crates on our beam line.
When I tried to operate these three boards _together_ in one VME crate,
two of the boards have given me fits: 'Access Error' is reported at
iocInit time and vxWorks dies then. These boards have been back to OMS
a couple of times; each time OMS tells me that they find no problems
but have replaced one of the chips. Joe Sullivan at APS is convinced
that it's a timing problem that OMS can't see because they haven't had
a 25 MHz crate processor. OMS is purchasing a mv167 so that particular
situation should be fixed soon. Maybe then, OMS can figure out what's
wrong with two of my cards.
I borrowed five VME 58-8 cards from APS/XFD here at Argonne and
booted my crate with all of them no problems. Talked to all 40 motor
axes individually. This confirms that my VME 58-8 problems are with
the two individual cards, not in my crate.
About EPROM Firmware versions:
Joe and I talked about this the other day. Version 1.8
is the most recent version of the EPROM firmware that can be used
with EPICS. DO NOT TRY TO USE v1.9. Your crates will get
'Access Error' (or was it Bus Error).
Other versions that work are: 1.3 & 1.4
(Most of these version upgrades are to provide the address pipelining
feature to a non-EPICS user of VME 58-8 boards. EPICS does not
use this feature. Apparently, we are quite sensitive to changes
in the OMS firmware.)
As far as running these motor controllers for long periods
of time, such as an overnight data scan, can't say I've been
in a position to do that yet.
Pete
Jemian@anl.gov
UNICAT
> From: Tony Cox - (415)926-3105 <tcox@ssrl01.slac.stanford.edu>
> To: TECH-TALK@aps.anl.gov, beamline_controls@aps.anl.gov
> Subject: OMS VME58 anyone?
>
> Has anyone done extensive tests on the Oregon Micro Systems' 58 8-axis
> stepper motor controllers?
>
> Pierre & I have had a number of `revisions' of the firmware (3 or 4 so
> far), all claiming to solve various problems. We're now up to version 1.8,
> and I hear that 1.9 is around the corner. We understand from the engineer at
> OMS that they don't have a VME bus analyzer, because its too expensive (and
> yes, he wants one but the accountant won't let him have one). I also here
> rumours of problems elsewhere in control freak land, but see nothing on the
> mailing lists.
>
> So, to allow me to rest better at night, has anyone:-
>
> a) Run more than one axis simultaneously
> b) Run with more than one OMS in the VME crate at the same time
> c) Done both a and b at once and seen nothing unusual or
> unexpected happening for a time period (say) of about how
> long it takes to obtain an EXAFS or crystallography
> data set??
>
> We'll be running typically 4-5 OMS58's in each of three VME crates,
> and need to identify problems (if any).
>
> Tony