[Soekris] The best way to check for ground problems? (was: Strange problem with net4801)

Wolfgang.Anger@t-systems.com Wolfgang.Anger at t-systems.com
Thu Jan 20 13:25:08 UTC 2005


> -----Original Message-----
> From: soekris-tech-bounces at lists.soekris.com 
> [mailto:soekris-tech-bounces at lists.soekris.com] On Behalf Of 
> Henry Spencer
> Sent: Thursday, January 20, 2005 5:41 AM
> To: Soekris Tech List
> Subject: Re: [Soekris] Strange problem with net4801
> 
> 
> On Thu, 20 Jan 2005, Horst Laschinsky wrote:
> > 1.) I still do not understand, how this noise can be triggered by
> >     the application layer protocol (as I said, in my case: scp).
> 
> As I said in my earlier posting, ground problems can cause 
> *very* strange symptoms.  In particular, it can be 
> protocol-specific, because different protocols can send 
> packets at different rates or in different patterns, and that 
> can change things like how much current the line drivers are 
> pushing into the cable.  In normal operation, something like 
> that doesn't matter, but what happens when you've got a 
> grounding problem depends so much on so many details that 
> even something this small can affect it. 
> 
> (I can't explain this particular effect in detail, but I have 
> seen enough weird things to check for grounding problems 
> immediately when a data link is doing bizarre, inexplicable things.)
> 
>                                                           
> Henry Spencer
>                                                        
> henry at spsystems.net
> 

Hi Folks,

I understand the checking for grounding problems in theory but in the praxis?
(the think what cam me in mind is to check the R between ground<->shield with a Ohm-meter, some else?)

best Wolfgang



More information about the Soekris-tech mailing list