Re: Fw: Contributions (was Re: disallow distributed switch branches)

Geoffrey.Coram wrote:
> Kevin Cameron wrote:
>   
>> Geoffrey.Coram wrote:
>>     
>>> It's an error because the LRM says so; it's a switch branch.
>>>
>>>       
>> Sorry, it isn't. A switch branch is where it changes within a block, not
>> between blocks.
>>
>> Driving a current from one place and a voltage from somewhere else is
>> perfectly legitimate electronics.
>>     
>
> Please point me to the section of the LRM that you're reading.
> In 5.1.5 (Switch branches) of LRM 2.2, there is no mention of
> blocks (sequential blocks?  analog blocks?).
>
>   

Switch branch as a term in simulation (not the LRM) means a branch which 
can be either a flow or potential contribution - you can't determine 
which statically so the simulator has to be handle it changing at runtime.

> In 5.3.1.3, it says "It is illegal to contribute to an external
> switch branch from within an analog block."  As I suggested
> earlier, the LRM could be clearer on this point: if a contribution
> to an external branch would have the effect of switching it, this
> contribution would be illegal.  But it would be somewhat odd to
> spend a paragraph describing an external switch branch just to
> say, by the way, this thing is illegal.
>   

I think this one depends on whether you are looking at the node-pair or 
a particular branch-contribution, i.e. for the branch you want all-flow 
or all-potential contributions, but as a node-pair you don't care. But I 
think that just takes us back to the debate on whether you want to use 
'<+' for OOMRs etc. - another operator which explicitly operates on the 
node-pair rather than branch disambiguates the problem.

Kev.
 
> -Geoffrey
>   


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.