FreeLists / si-list / [SI-LIST] Re: soldering plastic balls

["Lars Juul" <write2larsj@xxxxxxxxx>]

Interesting.
Having designed 10Gbps wirebond BGAs myself, I know the challenge.

Is your design based on low-cost, lossy BT-resin substrates with
conservative design rules, or do you cheat, using those nice $20 ceramic
BGAs with plugged vias and nice well-defined thin-film etched lines on the
top side? ;-)

Lars.



2007/6/5, steve weir <weirsi@xxxxxxxxxx>:
>
> Lars ( shameless plug alert ), if you would like a blazing fast package
> design that is 50 ohms through and through using BGA packaging,
> Teraspeed will be happy to help you out.  A 1.27mm pitch is near
> optimum, but we can work with 1 mm and still stay pretty close.
>
> Regards,
>
>
> Steve.
>
> Lars Juul wrote:
> > Just went one link deeper.
> > It seems that they're also trying to flog their product to us, claiming
> the
> > solder ball sphere is much more uniform in size for these balls, making
> it
> > possible to make a controlled impedance for the package/board
> interconnect.
> >
> > However, I'm not convinced, as
> > a) You need to place the balls awfully close to come even close to a Z0
> of
> > 50 Ohms. My bet is that you'd have a significant risk of shorting the
> balls
> > in a production environment. They might aim for a higher impedance,
> which
> > might lessen the Z0 mismatch, I suppose.
> >
> > b) In order for this to make sense, you need to somehow avoid the huge,
> > clumsy through vias that are normally seen on bga substrates. I am
> guessing
> > that the impedance mismatch caused by the package through vias is much
> more
> > significant than what you gain from having a controlled impedance across
> > your solder joints to the board. The vendor does in fact show a
> > microwave-like module with the circuitry on the same side as the solder
> > balls, which makes sense. But, honestly, how many would employ such a
> design
> > for standard high volume surface mount components? (apart from TBGAs, I
> > guess)
> >
> >
> > 2007/6/5, Lars Juul <write2larsj@xxxxxxxxx>:
> >
> >> Hi Steve,
> >>
> >> I'm surprised to hear that the primary objective with the plastic core
> >> balls is to equalize the resistance over frequency, as Sekisui write on
> >> their product page:
> >>
> >> ( http://www.sekisui-fc.com/eng/html/jissou/product/jipr_02.html)
> >> ...
> >> Micropearl SOL is a solder plated true sphere with a plastic core and
> was
> >> designed for mounting chips to circuit board. Unlike conventional
> solder
> >> balls which often cracks under the heat cycle test,
> >> ...
> >>
> >> They go on to show graphs with how much longer these solderjoints will
> >> hold in temp cycle tests comparison to SnAg alloys, which I feel is
> very
> >> relevant in these RoHS times.
> >>
> >> However, it's true that at high frequencies the current flows on the
> edge
> >> of the conductor (in closest possible proximity to your return
> current), and
> >> that the lack of conductor diameter will lead to increased resistance
> closer
> >> to low frequencies, but I doubt that it will act as a good loss
> equalizer (I
> >> might be wrong, I guess)
> >>
> >>
> >> 2007/6/5, steve weir <weirsi@xxxxxxxxxx>:
> >>
> >>> Lars, the idea of the plastic balls is that they have only  a thin
> outer
> >>> conducting layer and therefore the resistance remains much more
> uniform
> >>> and therefore equalizing loss across a broad frequency range.
> >>>
> >>> Steve.
> >>> Lars Juul wrote:
> >>>
> >>>> Hi David,
> >>>> I would call up your packaging vendor and ask them to make a test
> >>>>
> >>> build with
> >>>
> >>>> ball samples from Sekisui on existing 10 Gbps BGAs, provided the ball
> >>>>
> >>> attach
> >>>
> >>>> tool is suitable for this. This is usually the stumbling block, as
> >>>>
> >>> this is
> >>>
> >>>> very expensive in NRE.
> >>>>
> >>>> Alternatively, it's worth exploring if you make a manual ball attach
> >>>> (possible if the packages are for small form factor 10Gbps CDRs, for
> >>>> instance) by printing solder paste on the package lands first, and
> >>>>
> >>> then
> >>>
> >>>> manually place the balls  on the paste by a trained person with a
> >>>>
> >>> steady
> >>>
> >>>> hand. I assure you, crazier things have been made.
> >>>>
> >>>> On the other hand, my only concern with this kind of balls would be
> >>>>
> >>> the
> >>>
> >>>> suitability for them in the power supply, as I imagine the current
> >>>>
> >>> carrying
> >>>
> >>>> capabilities are somewhat limited if you have a plastc core on the
> >>>>
> >>> inside.
> >>>
> >>>> (I don't think it's possible to attach solid solder and plastic balls
> >>>>
> >>> on
> >>>
> >>>> specific lands on the package in a high volume setup, yet)
> >>>>
> >>>> PS: you're asking whether they work at 10GHz, which is a single
> >>>>
> >>> frequency
> >>>
> >>>> point, I think it's feasible to use plastic core balls for a
> >>>>
> >>> narrowband
> >>>
> >>>> application, but if it's broadband from 0 to 10e9 Hz you're driving
> >>>>
> >>> at, I'd
> >>>
> >>>> say, try it out in practise first to see if it flies, then make a
> full
> >>>>
> >>> wave
> >>>
> >>>> or macromodel of package afterwards to feed the chip I/O designers.
> >>>>
> >>>> Have fun.
> >>>>
> >>>> Best regards,
> >>>> Lars
> >>>>
> >>>> 2007/6/5, david stern <dan1_st@xxxxxxxxx>:
> >>>>
> >>>>
> >>>>> hello all
> >>>>>
> >>>>>   I was looking for bga soldering balls  technology for an
> >>>>> application working at 5GHz to 10GHZ and  and found
> >>>>> soldering plastic balls from SEKISUI. Did somebody work with them at
> >>>>> 10Ghz?
> >>>>>
> >>>>>   Thanks
> >>>>>
> >>>>>   David
> >>>>>
> >>>>>
> >>>>> ---------------------------------
> >>>>> Moody friends. Drama queens. Your life? Nope! - their life, your
> >>>>>
> >>> story.
> >>>
> >>>>> Play Sims Stories at Yahoo! Games.
> >>>>>
> >>>>> ------------------------------------------------------------------
> >>>>> 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 technical documents are available at:
> >>>>>                 http://www.si-list.net
> >>>>>
> >>>>> 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
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>> ------------------------------------------------------------------
> >>>> 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 technical documents are available at:
> >>>>                 http://www.si-list.net
> >>>>
> >>>> 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
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>
> >
> >
> > ------------------------------------------------------------------
> > 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 technical documents are available at:
> >                 http://www.si-list.net
> >
> > 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
> >
> >
> >
> >
> >
>
>


------------------------------------------------------------------
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 technical documents are available at:
                http://www.si-list.net

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