[SI-LIST] RF loss due to Nickel plating.

In response to the question about nickel plating losses in
transmission line traces, be very careful about translating the
results from Alloy 42 lead frames to copper/NI/Au or Cu/Ni/Solder
traces. In fact, when there is a non magnetic coating, like silver,
over the Alloy 42, the resistance and inductance are dominated by the
silver, due to the very small skin depth of the alloy 42 layer.

While it is true that Alloy 42 is ferromagnetic and shows high
permeability and a strong frequency dependent resistance, and nickel
is ferromagnetic, when there is a copper layer adjacent to the nickel,
the nickel  plays very little role in the circuit board trace.  When
you have a PCB trace that has a layered composition for the metal,
like Cu/Ni/Au, you have to ask where is the current really flowing. At
frequencies above 10 MHz, for 1 oz copper traces, the current
distribution is skin depth limited. With a high permeable material
like nickel, very little of the current will be in the nickel, most of
it will be in the copper.

Since the nickel is only used in the top layer, all such traces are
microstrip traces and most of the current will be in the lower part of
the copper. In measurements we have done of losses in FR4 transmission
lines, all of which have a Cu/Ni/Au plating, we are able to account
for all the losses, from ~ 100 MHz to 15 GHz, from just the dielectric
loss and the conductor loss, assuming just straight copper
metalization. You can see an example of this in my article in the Oct
issue of PC Design Magazine, "Analysis of Lossy Transmission Lines",
reprinted as an application note on our web site, www.gigatest.com.

--eric


**************************************
Eric Bogatin
CTO, Giga Test Labs
v: 913-393-1305
f: 913-393-1306
e: eric@xxxxxxxxxxxx
26235 W. 110th Terr.  Olathe, KS  66061
corporate office:
408-524-2700
134 S. Wolfe Rd Sunnyvale, CA 94086
web: www.gigatest.com
**************************************

Date: Fri, 12 Oct 2001 09:51:47 -0800
From: "Seol Byongsu" <seol@xxxxxxxxxx>
Subject: [SI-LIST] Re: RF loss due to Nickel plating.

Dave,

I think the following references are useful to you although they
include
not PCB-level information but package-level.

1.  C.-T. Tsai, et. al, "High-Frequency Inductance Measurement and
Characterization of Alloy42 and Copper Packages," 43rd  ECTC, pp.
635-640, 1993.  This paper includes comparison results of measured
inductance of five package interconnection materials: bare Cu, bare
Alloy42, Pd/Ni-plated Cu, Pd/Ni-plated Alloy42, and Au-plated Alloy42.
In addition to this, this paper shows estimated relative permeability
of
Alloy42 (Fe-Ni alloy) leadframe material in the IC packaging at 7
different frequencies to give an answer to your question #1.

Frequency (MHz)     Estimated relative permeability
2.6                         75
5                            60
8.85                       52
20                         45
80                         26
200                       12
300                        8

2.  Brian Young, "Determination of the Frequency-Dependent Complex
Permeability of Alloy42," 44th ECTC, pp. 978-980, 1994.
This paper includes complex permeability for Alloy42 derived from 2-D
simulations.

3.  O. Kagaya, et. al, "A Very Thin Power Amplifier Multichip Module
for
1.9 GHz Cellular Phone System," IEEE Trans. Advanced Packaging, pp.
40-45, Feb. 1999.  This paper shows relationship between the Q-factor
and the thickness of the Ni layer in the Cu/Ni/Au conductor for MCM
applications.

I hope this helps you.

-Seol

---------Included Message----------
>Date: Thu, 11 Oct 2001 15:11:10 -0700
>From: "Smith, David TQO" <dsmith@xxxxxxx>
>Reply-To: <dsmith@xxxxxxx>
>To: <si-list@xxxxxxxxxxxxx>
>Subject: [SI-LIST] RF loss due to Nickel plating.
>
>
>I have been asked to look at how much
>loss one might expect from nickel plating
>on a transmission line on FR-4 (or similar
>material with lower loss tangent) PCB.
>Operating frequency is in the 1-3 GHz range.
>
>I have found values published for Nickel
>resistivity and permeability (about 450!)
>and it is reported to be ferromagnetic, with
>a coercivity depending on the metallurgy.
>However, I have not been able to find any
>information about how these characteristics
>depend on frequency.  I know that the
>plating is magnetic because I can pick up
>small pieces with a magnet.
>
>I am left with these questions:
>
>1. Does the high magnetic permeability hold up to
>   GHz frequencies or does it roll off in the MHz?
>
>2. Is there any published references on this
>    phenomena?
>
>3. Is there a suitable commercial simulator that could
>    calculate the expected losses accurately? (I am going
>    to try to get credible results using Ansoft HFSS).
>
>Any sage advice, pointers to publications, or anecdotes
>about using nickel plating  on PCBs operating in the
>GHz frequency range would be much appreciated.
>
>Dave Smith

------------------------------------------------------------------
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:
http://www.freelists.org/webpage/si-list

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

List archives are viewable at:     
                http://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: