Dave

The use of PCB traces as a fuse is highly unreliable.  During my stint at
Racal-Vadic in the early 80's I was assigned the task of determining the
reliablility of PCB traces as a fuse.  The tests were made under very 
controlled conditions.  We had each of our 6 PCB vendors make 10 boards 
each.  Each PCB had a number of different trace widths connected to the 
RJ-11's.  The resulting variations in the fusing action was so wide and 
varied we abandoned the concept and wrote an edict "Thou shall not use 
PCB traces as a fuse."  Just the variation in opening up the trace to 
over current was totally unusable.  Not to mention the problem of 
shorting to other traces and components when a trace buckled and peeled 
from the PCB.  This was the case even with MIL grade material and PCB 
design specs.  

The best argument against this idea:  The company is finacially libel for
all injuries, damages and the board becomes a throw away.  It's simply
best to use a fuse.  I would be surprised if UL would accept a trace as a
fuse anyway.  I think it's a false economy.

Duane Marcroft
Telecom Consultant

___________________________________________________

On Mon, 29 Jul 1996, Dave Spalding wrote:

> Hi TREG -
> 
> We have the "opportunity" to purchase a cheaper transformer which will meet 
> the requirements of UL 1459 instead of UL 1950.  I understand that we can 
> comply with either standard until March 2000 or 2001 and UL 1459 is getting 
> talked about around here.  Our product is an ISA card that will fit into a 
> PC.  I would rather go with UL 1950 because the PC would be tested to that 
> (right?), because I have heard that we would need to install a fuse in the 
> front end to comply with UL 1459 and because UL 1950 is more compatible with 
> international standards..  My boss is looking to save cost so he wants UL 
> 1459 and answers my desire with the fact that we will have multiple boards 
> (one for USA/Canada, one for international), so we could use this transformer 
> for the domestic board, although the issue of whether Canada would accept UL 
> 1459 is not resolved. Also, the "fuse", he says, would be the etch on the 
> board because they would just peel off with the overvoltage testing  of UL 
> 1459.  I am not comfortable with that, though.  Anyway, I want to go with UL 
> 1950 but I need ammunition.  I was wondering if you could help me understand 
> the difference between UL 1459 and UL 1950 and also give me your 
> recommendation on which to test to.  Any help that you can give would be 
> greatly appreciated.  Thanks in advance.
> 
> - Dave Spalding
> 
> 
> David Spalding
> Motorola ISG
> 20 Cabot Boulevard
> M2-250
> Mansfield, MA 02048-1193
> Phone: (508) 261-4742
> FAX: (508) 339-2346
> EMail: [email protected]
> 

Reply via email to