Re: USB interface with EPICS

[Kate Feng <feng1@bnl.gov>]

"D. Peter Siddons" wrote:

> The only point here is that these little USB motor controllers are very
> attractive for micropositioning of optical systems, they fit nicely with
> most optical accessories and are small, including the controller module
> which is designed not to get in the way when mounted on an optical
> bench. Motorized micrometers are slow by definition, so who cares if the
> interface is not super-snappy? It would be nice to have them as part of
> the scientist's toolkit within the EPICS framework. THere are also USB
> CCD cameras which are relatively inexpensive.

Regarding the video,
USB does not have the DMA channels as what firewire does.
USB is not isochronous, and needs the host CPU to move data,
which ties up the CPU resource.
USB maybe OK  when one does not care if one lost video,
or do not need to process the video without losing frames.
Meanwhile, the firewire camera is inexpensive, too.

Kate

> No, USB is not the answer
> to all, but more and more stuff comes with it.
> Pete.
>
> Till Straumann wrote:
> > Eric Norum wrote:
> >> True, devices on the USB can not make unsolicited requests for
> >> service.  But that's not a show-stopper.  Consider all the GPIB
> >> devices in use which do not use the SRQ line -- the situation in that
> >> case is quite similar.
> >>
> >> USB controller chips do have interrupts so it's not necessary for a
> >> device driver to continuously poll some register in the controller to
> >> see when a message has arrived.
> > True - I just wanted to make the point that this in some
> > cases [when low-latency is critical] could be a show-stopper
> > for USB.
> >
> > T.
> >>
> >> On Dec 18, 2006, at 4:06 PM, Till Straumann wrote:
> >>
> >>> Just a side note before becoming excited about USB:
> >>> AFAIK (but I could be wrong), USB has no real interrupts
> >>> but always works in some sort of 'polled' mode. There could
> >>> be latency issues with USB devices if you need really fast response.
> >>>
> >>> -- Till
> >>>
> >>> Chestnut, Ronald P. wrote:
> >>>> Kevin,
> >>>>
> >>>> At SLAC, one of the RF physicists started using USB. To keep matters
> >>>> simple, he developed subroutines with which he could write to and
> >>>> read from USB. We then hook these in with subroutine records when
> >>>> time comes to put EPICS on it.
> >>>>
> >>>> I'd also like to see some real USB driver-supported stuff.
> >>>>
> >>>> Ron Chestnut
> >>>> -----Original Message-----
> >>>> From: Kevin M. Peterson [mailto:kmpeters@anl.gov] Sent: Monday,
> >>>> December 18, 2006 7:44 AM
> >>>> To: tech-talk
> >>>> Subject: USB interface with EPICS
> >>>>
> >>>>
> >>>> Is anyone using the USB interface with EPICS?  Of particular
> >>>> interest is the ThorLabs APT series.
> >>>>
> >>>> Thanks,
> >>>>
> >>>> Kevin Peterson
> >>>> kmpeters@anl.gov
> >>>>
> >>>>
> >>>
> >>
> >> --Eric Norum <norume@aps.anl.gov>
> >> Advanced Photon Source
> >> Argonne National Laboratory
> >> (630) 252-4793
> >>
> >>
> >
> >