[SI-LIST] Re: Fw: Problem with Ni plated transmission line

  • From: Aubrey_Sparkman@xxxxxxxx
  • To: kuchta@xxxxxxxxxx, si-list@xxxxxxxxxxxxx
  • Date: Fri, 14 Mar 2003 10:55:52 -0600

Daniel,
You didn't mention the number you used for Ur.  Ni is one of the 3 or 4
metals that have a magnetic permeability that matters.  While electroless Ni
(if my memory is correct) has a lower Ur than Electroplated Ni, both
conductivity and permeability should be considered for your skin depth
calculations.

Due to your difference in DC resistance, I would suspect significant
over-etch in your plating process.  Your Ni plated line should have lower DC
resistance.  You didn't mention what the differential impedances were.  How
close to target were they?

Aubrey Sparkman 
Signal Integrity 
Aubrey_Sparkman@xxxxxxxx 
(512) 723-3592


> -----Original Message-----
> From: Daniel Kuchta [mailto:kuchta@xxxxxxxxxx]
> Sent: Friday, March 14, 2003 9:29 AM
> To: si-list@xxxxxxxxxxxxx
> Subject: [SI-LIST] Fw: Problem with Ni plated transmission line
> 
> 
> >The plating process will
> prevent oxidation of the copper lines; does the soldermask 
> have the same
> functionality?
> 
> Yes, I believe it does.  The reason for the plating was to prevent Cu
> oxidation.
> 
> 1/2 oz Cu is about 18um thick.  The 100u in of Ni adds another 2.5um.
> There is a pre-etch in the plating process which is supposed 
> to remove Cu
> oxide and prepare the surface for plating.
> It seems that ~1um of Cu may have been removed by this pre-etch.
> From our skin effect calculations, we expected the fields to 
> be mostly in
> the Cu even for really poor Ni conductivity.
> These consideration still do not explain our results.
> 
> The Er of Pyralux is 3.1.  The loss tangent of Pyralux is 0.005.
> 
> Daniel M. Kuchta
> IBM T. J. Watson Research Center
> 1101 Kitchawan Rd. Rt. 134
> Yorktown Heights NY 10598
> 914-945-1531 (voice)
> 914-945-4134 (fax)
> kuchta@xxxxxxxxxx
> 
> ----- Forwarded by Daniel Kuchta/Watson/IBM on 03/14/2003 
> 10:24 AM -----
> |---------+---------------------------->
> |         |           Silqun Leung     |
> |         |           <SLeung@xxxxxxxxx|
> |         |           m>               |
> |         |                            |
> |         |           03/14/2003 10:15 |
> |         |           AM               |
> |---------+---------------------------->
>   
> >-------------------------------------------------------------
> --------------------------------------------------------------
> ------------------|
>   |                                                           
>                                                               
>                     |
>   |       To:       Daniel Kuchta/Watson/IBM@IBMUS            
>                                                               
>                     |
>   |       cc:                                                 
>                                                               
>                     |
>   |       Subject:  RE: [SI-LIST] Problem with Ni plated 
> transmission line                                             
>                          |
>   
> >-------------------------------------------------------------
> --------------------------------------------------------------
> ------------------|
> 
> 
> 
> 
> Daniel,
>  Those are some interesting results (which I should have considered
> when I was designing things before).  If memory serves me 
> correct, 1/2oz
> copper produces traces with 17um (or was it 35um) thickness 
> correct?  I'm
> assuming these are microstrip transmission lines, in other 
> words at higher
> frequencies one would expect current to be concentrated on the bottom
> portion of the conductor.  However the electroless process 
> would affect not
> only the conductor thickness but also the conductor width, so 
> your fields
> are no longer propogating along a 'homogenous' conductor.  
> Even higher up
> in
> frequency, the current will actually be concentrated along 
> the edges of the
> conductor, which is no longer copper, but Ni.  These are only 
> some things
> off the top of my head and not something I've tested 
> extensively, however I
> believe the reasoning is there.
> 
>  I have another question for you then.  The plating process will
> prevent oxidation of the copper lines; does the soldermask 
> have the same
> functionality?  Also what is the Er of the substrate you're using?
> 
> Thanks,
> 
> Silqun
> 
> 
> > -----Original Message-----
> > From: Daniel Kuchta [mailto:kuchta@xxxxxxxxxx]
> > Sent: Friday, March 14, 2003 9:58 AM
> > To: si-list@xxxxxxxxxxxxx
> > Subject: [SI-LIST] Problem with Ni plated transmission line
> >
> >
> >
> >
> >
> >
> >
> > I have run  into an unexpected problem with a Ni plated
> > transmission line.
> >
> > I have fabricated 2 types of 100 Ohm differential
> > transmission lines on a flex circuit. They are both
> > approximately 3mil width and 3 mil space.  The substrate is
> > 3mil Dupont Pyralux.
> >
> > The metallurgy on one pair is 1/2oz copper covered with
> > PSR9000 solder mask. The metallurgy on the other pair is 1/2
> > oz copper plated with 100u in Electroless Ni and 10u in
> > Electroless Au. The plated line is not covered with solder mask.
> >
> > The expected result was that the plated line would have
> > higher bandwidth since the solder mask has a loss tangent of 0.02.
> >
> > The measured result is 20GHz for the covered lines and 7 GHz
> > for the plated lines! The plated lines appear to suffer from
> > high conductor losses.
> >
> > The dc resistance of the lines is also different: 0.5 Ohm for
> > the unplated and 1.1 Ohm for the plated.
> >
> > Does anyone have a similar experience with plated
> > transmission lines? Does anyone have an explanation for this effect?
> >
> > Thanks
> >
> > Daniel M. Kuchta
> > IBM T. J. Watson Research Center
> > 1101 Kitchawan Rd. Rt. 134
> > Yorktown Heights NY 10598
> > 914-945-1531 (voice)
> > 914-945-4134 (fax)
> > kuchta@xxxxxxxxxx
> >
> > ------------------------------------------------------------------
> > To unsubscribe from si-list:
> > si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the 
> Subject field
> >
> > or to administer your membership from a web page, go to:
> > //www.freelists.org/webpage/si-list
> >
> > For help:
> >
> > si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
> >
> > List archives are viewable at:
> >           //www.freelists.org/archives/si-list
> > or at our remote archives:
> >           http://groups.yahoo.com/group/si-list/messages
> > Old (prior to June 6, 2001) list archives are viewable at:
> >           http://www.qsl.net/wb6tpu
> >
>  >
> 
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
> 
> or to administer your membership from a web page, go to:
> //www.freelists.org/webpage/si-list
> 
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
> 
> List archives are viewable at:     
>               //www.freelists.org/archives/si-list
> or at our remote archives:
>               http://groups.yahoo.com/group/si-list/messages 
> Old (prior to June 6, 2001) list archives are viewable at:
>               http://www.qsl.net/wb6tpu
>   
> 
> 

------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List archives are viewable at:     
                //www.freelists.org/archives/si-list
or at our remote archives:
                http://groups.yahoo.com/group/si-list/messages 
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu
  

Other related posts: