Re: GreenSpring TIP810 CANbus system

[John Maclean <jfm@aps.anl.gov>]

Bill,

At APS we've just moved to VIPC626 IP carriers. These are software 
compatible with the previous VIPC616 model which has EPICS support. The 
VIPC626 are a lot easier to configure I believe, no wire wrapping required.

I've used the TIP810 modules with Andrew Johnsons driver code on a 
couple of projects and have found them to be absolutely rock solid. IIRC 
they use an early generation of CAN controller chip. Later generations 
of CAN controller mainly do more to offload processing from CPU. This is 
probably not an issue unless you have many heavily loaded CAN links 
coming into your ioc. I don't think 3 will be a problem, but it will 
depend on the amount of bus traffic you have of course. Gemini have 8 
buses in one ioc. It almost certainly isn't an issue if you use a PPC.

HTH,

John.

wtw00992@creighton.edu wrote:

> Greetings,
>
> I am trying to duplicate a CANbus system here at the STAR experiment. We have
> two distinct modules/controllers and chains.  We want to split off some Wiener
> crates to form a third (to make life easier on some subsytem detectors).  It
> would also give us a backup option!
>
> We presently have two GreenSpring TIP810, which are VIP610 carrier boards. (6U
> VME boards with 4IP and front I/O ribbon connections).  We use two processors,
> one is a 162, and the other 167, otherwise things are pretty identical, except
> for the numbers and kinds of Wiener crates on each chain.   For these systems we
> run an old version of VxWorks 5.2 and an old version of EPICS 3.12-LBL.4 
>
> I did get some information from TEWs regarding the TIP810-10 which replaces the 
> Greenspring 810, and am anxious to order.  I wanted to get some expert advice
> first.  
>
> Would there be significant benefits to trying a "new" approach rather than
> duplicate. This all seems like a quick and easy option but perhaps there are
> others that we should be considering?  (For example other carrier boards, the
> TIP816 etc.)
>
> Thanks in advance
>
> Bill Waggoner
> STAR
> Brookhaven National Lab