TECHNET Archives

March 2001

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Marty Brooks <[log in to unmask]>
Reply To:
TechNet E-Mail Forum.
Date:
Thu, 29 Mar 2001 15:11:43 -0800
Content-Type:
text/plain
Parts/Attachments:
text/plain (221 lines)
Thanks everyone for your input.

To respond to some of the comments:

Phil - Yes, we removed the chip and resoldered the same chip to the pwb.
Problem is How can I tell the vendor to perform a labour intensive rework
procedure in a mass production environment without the vendor charging me
for it.  I will argue their process is causing the problem, they will argue
it's our design.  And yes, we did replace the poorly performing part with a
'new' one and performance was much better with the new part.  This all leads
me down the path of solder joint quality - I just need some way to prove
that to the vendor.

Steve - I think the bottom only terminations of the chip does pose a
problem.  Have other assembly shops had problems with devices like this and
how have they overcome them?  The chip is a ceramic base with gold plated
terminations.

Dave - I've considered the trace dimensions.  They are all optimized as best
they can, except for the Xceiver chip which has pads 10mil longer than
recommended by the manufacturer.  I wonder if that extra length, as small as
it is, would be enough to throw the impedance out of whack?  Something that
I will look into.

Ingemar/James - I'm no RF engineer either.  The device operates at 900MHz.
Board is FR4 with a dielectric constant of 4.1 or 4.2.  Board is impedance
controlled to 50ohms.  Surrounding components are 0603 resistors and caps -
except for the RF output line which goes through a 15nH wire wrapped
inductor.  We have optimized the layout and moved all the components as
close together as possible to minimize any transmission line anomalies that
may want to appear.  As far as your other questions, naturally I don't have
those at my fingertips, but I'll see what I can dig up and present them to
the first RF guru I can track down.

You have provided some new points for consideration though.

Thanks
Marty


-----Original Message-----
From: Ingemar Hernefjord (EMW)
[mailto:[log in to unmask]]
Sent: Thursday, March 29, 2001 6:51 AM
To: [log in to unmask]
Subject: Re: [TN] RF Assembly Issues


Marty, sounds like you have to use a fishbone modelling. So many parameters
that could have impact. You don't say what kind of part you are making,
something for under 2GHz or much above? And nothing about dielectric and
surrounding components, ground planes, distriubution kind of traces, noise
level requirements, spurioses, I/O-loading etc. Not easy for an RF engineer
to give advice. Have you modelled the circuit and simulated variations in a
computer? If you are talking very high frequencies, even length of bondwires
on inside of package plays a role. And what's the package made of, ceramic
or plastic? If you ask an RFspecialist to engage you need prepare to give
information about:

epsilon as function of temperature for your board material, as well as for
all the components
RF-parameters (S e.g.)as function of temperature
possible hysteresis of your ceramic caps, if you use such (behave different
on way up and down T)
identification of the parameter that gives your freq shift (Temp sensitive
part on inside of 20leg)
checking if you really use valid test set up (microwave is tricky)
checking possible ferroelectric phenomenons in ceramic components, if you
use such
what kind of tuning components do you use? Are they temperature sensitive?
As James said, check all return paths, also crosstalking and feedback
traces.
and many more questions, depending on the complexity of your circuit
diffusion phenomenons that can change chip data
Tin, Silver migration that can cause bridging


I recommend you to see nearest RF engineer in person. RF problems are rarely
solved by mail. If it is a RF-problem at all! Try not to solder the 20pin at
all. Mount some 10 boards using silverepoxy mounting the 20pin after first
soldering all other components. Aftermounting with other words. If all 10
get center freq where it should be, then the 20pin could be temp sensitive.

In fact your questions seems to reveal that you are not making an advanced
HF circuit but something for the commercial 'low freq' market, a radio board
or so. Important for you to tell TN, because the 'RF' guys are either
ordinary and sound radio/TV/equipment makers..or Telecom/MIL/SPACE 'real' RF
neurds with doctor's grade and at least 500 empty Coca bottles on the
wall..2GHZ is called DC among the later...

Ingemar

-----Original Message-----
From: Marsico, James [mailto:[log in to unmask]]
Sent: den 29 mars 2001 14:50
To: [log in to unmask]
Subject: Re: [TN] RF Assembly Issues


Most of the time, RF components need the body (under the component) soldered
to the PWB for proper grounding.  I'm no RF engineer, (God forbid) but I
understand that this grounding can be critical for performance.  Perhaps
this is a good place to start.

Jim Marsico
Senior Engineer
Production Engineering
AIL/Electronics Systems Group
An EDO Company
[log in to unmask] <mailto:[log in to unmask]>

        -----Original Message-----
        From:   Marty Brooks [SMTP:[log in to unmask]]
        Sent:   Wednesday, March 28, 2001 4:41 PM
        To:     [log in to unmask]
        Subject:        [TN] RF Assembly Issues

        We are experiencing a concerning variation from board to board in
the
        performance of our RF product.  We have investigated several causes
of the
        problem and are currently of the opinion that soldering quality of
the
        transceiver chip is the most significant contributing factor.  As an
        example, on the last run of prototypes that was manufactured, the
signal
        frequency was at the upper limit of the allowable range.  We removed
and
        replaced the transceiver chip in-house and decreased the frequency
closer to
        what we would expect.

        I'm going to be making a visit to our contractor's facility next
week to
        investigate this problem and would like some feedback on possible
causes or
        process controls that I should look into.

        The board is 4 layers, HASL finish, single side smt, solder paste
used is
        Alpha's UP78 no-clean.  The transceiver chip is a 20-pin device with
        terminations flush along the bottom of the component.

        Thanks for any advice.

        Marty Brooks
        Manufacturing Engineer
        IDENTEC Solutions, Inc.
        102, 1860 Dayton Street
        Kelowna, B.C., Canada V1Y 7W6
        Ph:(250)860-6567, Fax:(250)860-6541
        www.identec.com

        This email contains confidential information and is intended for the
sole
        use of the addressee. Any other distribution, copying or disclosure
is
        strictly prohibited. If you are not the addressee, please notify the
sender
        by email and delete this message.


----------------------------------------------------------------------------
-----
        Technet Mail List provided as a free service by IPC using LISTSERV
1.8d
        To unsubscribe, send a message to [log in to unmask] with following
text in
        the BODY (NOT the subject field): SIGNOFF Technet
        To temporarily halt delivery of Technet send the following message:
SET Technet NOMAIL
        Search previous postings at: www.ipc.org > On-Line Resources &
Databases > E-mail Archives
        Please visit IPC web site (http://www.ipc.org/html/forum.htm) for
additional
        information, or contact Keach Sasamori at [log in to unmask] or
847-509-9700 ext.5315

----------------------------------------------------------------------------
-----

----------------------------------------------------------------------------
-----
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt delivery of Technet send the following message: SET
Technet NOMAIL
Search previous postings at: www.ipc.org > On-Line Resources & Databases >
E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700
ext.5315
----------------------------------------------------------------------------
-----

----------------------------------------------------------------------------
-----
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt delivery of Technet send the following message: SET
Technet NOMAIL
Search previous postings at: www.ipc.org > On-Line Resources & Databases >
E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700
ext.5315
----------------------------------------------------------------------------
-----

---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt delivery of Technet send the following message: SET Technet NOMAIL
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------

ATOM RSS1 RSS2