Vasile Surducan wrote:
> On 7/29/06, stef mientki <s.mientki@mailbox.kun.nl> wrote:
>   
>> Dumitru Stama wrote:
>>     
>
> snip
>
>   
>>> first to be sure there is no hardware connection error. Have you tried
>>> with more than one ENC chip ?
>>>       
>> Yes, ...
>> ... but I know I'm violating the SPI errata,
>> but I've seen no problems there  ;-)
>>     
>
> Just a thought Stef :
>
> I know you're using USART for the hardware SPI protocol.
> I don't know how well it works that (mostly because USART and SPI are
> shared in the same library). Contrary to Gods, the mortal humans are
> using MSSP for hardware SPI.
>
>   
well Vasile,
good idea, but in this case I behaved like a normal human, ...
... and as normal human I make mistakes ;-)

I just discovered an interesting phenomena,
adding extra capacitors changed the behavior !!
 - error rate increased
 - ping response time decreased
 - error detection / recovery increased
So I guess it's a real hardware problem.
As all is done on a prototyping module,
wiring is not optimal. But I now think power-supply wiring and 
decoupling is very important for the ENC-chip.

I also read the microchip TCP/IP stack sources,
and discovered there is error ("single SPI bit error", not described in 
the errata) that I was not aware of.

cheers,
Stef

> Could be this one be a problem ? (It's not the PIC-ENC communication
> problem after all?)
>
> greetings,
> Vasile
>   

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist