Richard, 

Great discussion.  No, I haven't found trace thickness 
to be of any significant clearance issues.  But, then, 
I've only worked as high as 2 oz Cu.  Nothing higher. 

I have no experience with IEC 664-1.  Sorry.  

Regards,  Doug 


Richard Steele wrote:
> 
> Doug,
> 
> > Situation A :  Board is continuous between trace A
> >                and trace B.  That is to say, there
> >                is FR4 material between the traces.
> >
> >  ----------------------
> >    ------    ------
> >   |  A   |  |  B   |
> >    ------    ------
> >  ----------------------
> 
> We have treated this as a cemented joint and have achieved an electric
> strength >5kV for a 6 thou spacing (inner layers only). It's the outer
> layers where creepage (and to some extend clearance) becomes applicable
> with around 1kV for a 6thou spacing.
> The creepage increased from 1kV @ 6thou to 2.3kV at 60thou whereas the
> clearance increased from 980V @ 6thou to 5kV @ 54thou.
> Different samples gave slightly different results, this may be to do
> with the fibreglass weave grain and surface contamination.
> 
> >
> > Situation B : Board is NOT continuous between trace A
> >               and trace B.  That is to say, there is an
> >               air gap cut through the board between
> >               the traces leaving the traces exposed.
> >
> >  --------+  +-----------
> >    ------|  |------
> >   |  A   |  |  B   |
> >    ------|  |------
> >  --------+  +-----------
> >
> > Clearly, even on an unpopluated board fab,
> > Situation B is strictly a clearance issue.
> 
> Haven't used this scenario and can't think when we would, but yes in
> this situation clearance is more applicable than creepage, dependant
> upon the distances involved.
> 
> > I assume you're talking 1 or 2 oz copper for your traces?
> 
> Generally 1oz but greater on PSU's, have you found that the thickness
> alters the creepage/clearance greatly?
> 
> What relevance are the creepage distances quoted in IEC 664-1 Table 4
> column 2 for printed circuit material, do these assume a CTI of >600
> (Material Group I)?, if so this has not been stated.
> The IEC 664-1 creepage figures differ greatly from those in IEC 950 for
> PCB material. Where did these come from and what can they be used for?
> 
> Are there any creepage and clearance experts out there?
> 
> Regards
> 
> Richard Steele
> Equipment Engineering
> Fujitsu telecommunications Europe limited

Reply via email to