[SI-LIST] Re: IPC-2152 trace current/temperature calculator
- From: Jack Olson <pcbjack@xxxxxxxxx>
- To: Arjun Bingipur <arjun.bingipur@xxxxxxxxxxxxxx>
- Date: Thu, 14 Jan 2010 10:04:44 -0600
Thanks.
I had the same difficulty, that's why I did it.
To everyone else, sorry for the interruptions
and thanks for your patience.
.
and now,
back to our regularly scheduled broadcast:
SIGNAL INTEGRITY!!!!
j
.
On Thu, Jan 14, 2010 at 9:59 AM, Arjun Bingipur <
arjun.bingipur@xxxxxxxxxxxxxx> wrote:
> Thanks Jack. Personally, I feel without the modifier charts, it would
> have been a difficult task analysing the temperature curve on its own. I was
> on it for a quite some time without much clarity. But with modifiers
> and temperature curves together, it becomes so much simpler. It would have
> helped if it was part of the the IPC spec.
>
> Regards,
> Arjun
>
> ------------------------------
> *From:* Jack Olson [mailto:pcbjack@xxxxxxxxx]
> *Sent:* 14 January 2010 15:11
> *To:* Arjun Bingipur
> *Cc:* Doug Brooks; SI-LIST
> *Subject:* Re: [SI-LIST] Re: IPC-2152 trace current/temperature calculator
>
> No, I created those charts myself, using the data in the new spec (which
> was obtained from the Lockheed Martin and Naval Surface Warfare Center,
> Crane Division experiments)
>
> The reason I created the "modifier" system is because the way the document
> is laid out, you have to read all kinds of extra stuff (including the
> Appendix!!!) just to get a decent result. The charts give you a starting
> number, but then the text describes all kinds of things that can influence
> the number. I tried to condense what was there into an introductory approach
> to deriving a better answer than the historical charts could have ever
> provided. It is the SIMPLEST, in other words.
> I couldn't put too many charts in the article (and PCDesign magazine was
> already complaining about the length!) so for example I showed how to use
> the universal chart (based on 3oz copper) and modify the result for any
> other copper
> thickness to get a decent approximation, but of course you will NOT have to
> do this if you have all of the temperature curve charts in the spec itself,
> because it contains ALL the copper thicknesses for both internal and
> external.
>
> And maybe I don't have to say this, but there are other considerations that
> may influence your result. For example, this data is provided for traces in
> STILL AIR, and since most of our designs are in sealed housings these
> numbers are good enough for us, but if you have designs with good air flow,
> you may get better results with even smaller traces. On the other side of
> the coin, if you use a specific distance to your nearest plane but your
> nearest plane is chopped and split,
> your results may vary. And finally, I didn't even address parallel traces
> at ALL, and just about every design has those.
>
> so...
>
> To summarize I can only say
> 1) this is a better starting point than IPC has ever provided before
> 2) It is only an approximation, so add some derating "wiggle room"
> 3) You may need thermal simulation to prove your own results
> 4) get the spec, it contains much more than I could put into that little
> article
>
> (that's why I don't envy Doug the task of trying to make a calculator.
> everything needs a disclaimer!)
>
> http://pcdandf.com/cms/magazine/209/6850/
>
> hope that helps,
> Jack
>
>
> .
> On Thu, Jan 14, 2010 at 5:06 AM, Arjun Bingipur <
> arjun.bingipur@xxxxxxxxxxxxxx> wrote:
>
>> Jack,
>>
>> As per your document, If we are trying to obtain the final trace width
>> using the new universal chart rather than the conservative chart, we
>> need to be able to modify the values according to the modifier charts.
>> But, having gone through IPC 2152, I do not find the modifier chart
>> information. However, I'm currently having access only to the IPC 2152 -
>> final Draft document, dated December 2008. Does the released IPC 2152
>> spec contain this information or am I missing anything here?
>>
>> Regards,
>> Arjun
>>
>> -----Original Message-----
>> From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
>> On Behalf Of Jack Olson
>> Sent: 13 January 2010 19:38
>> To: Doug Brooks
>> Cc: SI-LIST
>> Subject: [SI-LIST] Re: IPC-2152 trace current/temperature calculator
>>
>> Doug,
>> I apologize for using the word WRONG in my last post.
>>
>> You have my sincere gratitude for the papers you have written and the
>> tools you have developed and shared so unselfishly. I really appreciate
>> those contributions and I intend to make good on my promise to buy you a
>> lunch someday for letting me use your utilities for FREE for so long
>> (smile)
>>
>> While I agree that it would be very difficult to implement all the new
>> data in IPC-2152 into a simple tool, the one thing that jumped out at me
>> was the proximity to planes. Most of the other factors can't influence
>> the result by more than a few percent. And I'm not a programmer either,
>> but it seems like it would be easy to add the distance to nearest plane,
>> even if you want to derate it somewhat for safety margin. Although I'm
>> not a "thermal guy", I was on the IPC committee (did you see my name on
>> page iii? twice? My mom did! Ha) anyway (sorry about that shameless
>> puffery) I would cheerfully help you any way I can, if you ever try to
>> implement more of it into your tool. I even have a few formulas that
>> were interpolated (?) from the curves in an excel spreadsheet.
>> By the way, we were considering incorporating the fusing data on your
>> (UltraCAD.com) site into the standard, but we didn't want to delay the
>> publication. maybe next revision....
>>
>> Finally, I totally agree with you that the historical chart has
>> withstood the test of time (because in most cases it is very
>> conservative), but what is need is a safer way to "push the envelope" if
>> the designer needs to use smaller features.
>>
>> Do you have any interest in developing a free thermal simulator for us?
>> (grin)
>>
>> thanks, Doug
>>
>> Jack (aka "the new guy")
>>
>> p.s. PLEASE accept my apology for implying that your calculator is wrong
>>
>>
>>
>> .
>> On Wed, Jan 13, 2010 at 1:13 PM, Doug Brooks
>> <dbrooks9@xxxxxxxxxxxxxxx>wrote:
>>
>> > Thanks for the link to your paper. Its a good contribution. Well done.
>> > I'll consider trying to implement some of it in the next version of
>> > the calculator.
>> >
>> > But please understand that this is now the third version of our
>> calculator.
>> > We have never ever suggested that it is *RIGHT* or *WRONG*. In fact,
>> > given that there are almost an infinite number of combinations of
>> > stackups layer dimensions trace dimensions and form factors swiss
>> > cheese effects component placement component thermal effects
>> > environmental effects materials properties etc
>> >
>> > anyone would be a little foolish to suggests they even knew or could
>> > know what *RIGHT* and *WRONG* really was.
>> >
>> > Also, note that we did not help write the standard, nor were we a
>> > member of any subcommittee
>> >
>> > We don't even necessarily endorse the standard (and the IPC doesn't
>> > necessarily endorse our calculator, at least as yet)
>> >
>> > The calculator merely makes using the standard (right or wrong,
>> > somehow
>> > defined) easier.
>> >
>> > The calculator is actually based on 5, user selectable data sources.
>> > These sources differ by as much as 40%. The new IPC-2152, Vacuum, is
>> > the most conservative. The old IPC-D-275 is the most aggressive. The
>> > old standard has been criticized through the years (with some
>> > justification) BUT nevertheless many people think that it HAS
>> > withstood the 50-year test of time! The data from another study,
>> > reported in Design News in 1968, provides results that are in between
>> > the other standards. The final decision of what results to use are the
>>
>> > system designer's responsibility, based on the risks he/she
>> > understands and is willing to take. We don't endorse any one of the
>> data sources.
>> >
>> > The trace current/temperature calculations are one of four
>> > capabilities of the calculator. Others include an estimate of fusing
>> > current (see an article on our web site), skin effect calculations
>> > (including skin depth, crossover frequency, and frequency and
>> > temperature adjusted trace resistance, see an article on our web
>> > site), and basic Ohm's Law calculations for traces. The trace
>> > current/temperature calculations and fusing current calculations can
>> be automatically adjusted for skin effects.
>> >
>> > Again, we never suggest our calculators are *RIGHT* or *WRONG*. Just
>> > that they make working with already established formulas and
>> > relationships easier. We developed them for our own use, and make some
>>
>> > of them available to others --- as is and as represented.
>> >
>> > Doug
>> >
>> >
>> >
>> >
>> > At 08:45 AM 1/13/2010, Jack Olson wrote:
>> >
>> >> Not to complain (kudos for developing the calculator!) but since
>> >> proximity to planes has the most drastic effect over any of the other
>>
>> >> factors, and it is a fairly linear relationship, I'm curious why you
>> >> didn't incorporate it.
>> >> (only one extra box!) of course, not knowing about splits and "swiss
>> >> cheesiness" of the plane, maybe you were simply afraid? (grin,
>> >> kiddin' ya) There is a graph in this article if you are interested:
>> >> http://frontdoor.biz/PCBportal/HowTo2152.pdf
>> >>
>> >> Unless I'm misunderstanding your calculator, I think your result will
>>
>> >> be about 50% WRONG in most cases (since most boards have planes these
>>
>> >> days)
>> >>
>> >> just trying to help,
>> >> keep up the good work,
>> >> Jack
>> >>
>> >>
>> >>
>> >> > From: dbrooks9@xxxxxxxxxxxxxxx
>> >> > Subject: [SI-LIST] Re: IPC-2152 trace current/temperature
>> >> > calculator
>> >> > Date: Tue, 12 Jan 2010 22:38:19 -0700
>> >> >
>> >> > I don't know if you have read the standard. There are a couple of=
>> >> pages
>> >> > that address thermal modeling, planes, alternative dielectrics,
>> >> > etc,= in
>> >> a
>> >> > very general manner. There are no quantitative adjustments included
>>
>> >> > (=
>> >> with
>> >> > the possible exception of how to treat parallel conductors which is
>>
>> >> > pr= etty straightforward, and pretty obvious.) There is a Figure
>> >> > A-13 that rela=
>> >> tes
>> >> > to some estimated factor adjustments that can be used based on the
>> >> > dist= ancebetween the conductor and a plane for 1 Oz traces in .070
>>
>> >> > thick
>> >> Polyim=
>> >> > ide Board (and a single curve for 2 Oz traces). The calculator
>> >> > does n=
>> >> ot
>> >> > incorporate these adjustments in its calculations. But these
>> >> adjustments=
>> >> > caneasily be made to the calculated results if desired. Overall,
>> >> > the
>> >> gene=
>> >> > ralities included in the Standard apply equally well to the
>> >> > calculated resu= lts from the calculator.
>> >> > The Calculator is based o= n Conductor Sizing Charts contained in
>> >> > the Appendix (from Figure A-17, p 35= to Figure A-86, p83).
>> >> >
>> >> > Doug
>> >> >
>> >>
>> >>
>> >> ------------------------------------------------------------------
>> >> 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 technical documents are available at:
>> >> http://www.si-list.net
>> >>
>> >> List archives are viewable at:
>> >> //www.freelists.org/archives/si-list
>> >>
>> >> Old (prior to June 6, 2001) list archives are viewable at:
>> >> http://www.qsl.net/wb6tpu
>> >>
>> >>
>> > Check out our resources at http://www.ultracad.com
>> >
>>
>>
>> ------------------------------------------------------------------
>> 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 technical documents are available at:
>> http://www.si-list.net
>>
>> List archives are viewable at:
>> //www.freelists.org/archives/si-list
>>
>> 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 technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
