Brian,

You make a good point and I apologize if I was not clear.  I was thinking of 
the test in terms of qualifying a process, not as a process monitoring tool. 
Monitoring the process is one thing, but validating the process with this 
method is entirely another.  And I see companies that use it for that 
purpose all of the time. Hence my remark.

Joe Russeau


----- Original Message ----- 
From: "Brian Ellis" <[log in to unmask]>
To: "TechNet E-Mail Forum" <[log in to unmask]>; "Joe Russeau" 
<[log in to unmask]>
Sent: Wednesday, April 16, 2008 10:42 AM
Subject: Re: [TN] SV: [TN] Ionic Contamination Question


It is perfectly true that the 75/25 solution is also an anomaly dictated
by the extremely short integration time of Hobson's reference method. It
makes absolutely no sense in testers where the PCB being tested is
immersed in the solution. The UK Ministry of Defence did tests with 75,
67.5, 50, 32.5 and 25% solutions. The first two lacked sensitivity and
the last one solubilisation power. The USA argued surface tension, but
the difference is very small (~23 dyn/cm for 75% and 24.5 for 50% at
20°C, rising to 30 at 25%). In the end, they decided that 50% v/v (this
is important) was an excellent compromise and easy to formulate for use
with all flux types, including rosin and water-soluble (which they
incidentally preferred, even as far back as 1986). This solution was
specified as having a density of 0.919 g/cm3 ±0.001 g/cm3 at 20°C. It
should be noted that the %/density curve is non-linear. The European
Space Agency also did very comprehensive tests and came to similar
conclusions, as well as proving that the MIL-P-28809A commercial
instruments "fudge factor" was totally unscientific because they had to
be changed radically for each make of flux, even by as much as half an
order of magnitude.

I disagree with you about the usefulness of ionic contamination testing.
I agree it is not the most scientific method and is next-to-useless in
today's context as a qualification method, but it is the only reliable
tool we have, when used properly with statistical analyses, for a rapid
and cheap process control covering everything from flux chemistry,
soldering parameters, cleaning chemistry and cleaning parameters. I
still consider it as having a paramount importance if used properly. The
only problem is that many users haven't a clue how to use it properly:
they see a figure of 1.5 µg/cm² and use that, whether they are making
the latest toy gadget for 3-y.o. kids, an i-Pod, avionics for the A380
or Dreamliner or are sending it into outer space. Not to mention that
they try (often unsuccessfully) to respect this same magic number with
all fluxes and even if they don't clean! What an unintelligent waste of
a good tool!

As an ex-manufacturer of ionic contamination testers, including the most
sophisticated ones ever to appear on the market, I still have many ideas
how the principle and technology could be improved. I sold out my rights
  on the subject in 1991 and have since regretted it. Since that date,
no radical improvements have been made to the technology; au contraire,
the instruments have been pared down by minimalists to lower prices, but
they are no easier to use and they do far less today than my old CM-5
model did 17 years ago.

Brian

Joe Russeau wrote:
> Agreed.  By-the-way, I like blunt,  it saves time.  One thing that I have 
> yet to see mentioned, which complicates this issue more, is that the ROSE 
> test uses a solution of 75% isopropanol and 25% deionized water.  The 
> possible variations in flux formulations are enormous.  With all of this 
> in mind, not all materials will dissolve in 75/25, nor do all materials 
> dissolve at the same rate.  I would think this would be especially true 
> for lead-free materials, given the fact that their residues tend to be 
> more difficult to remove.  Therefore, I ask, how can a value that was 
> developed over 30 years ago, for the higher rosin based fluxes, still 
> apply to today? It absolutely makes no sense to think that it does.  Yet, 
> somehow, that value keeps finding its way back into the standards. 
> Personally, I would like to see the ROSE test go the way of the dodo. 
> Although, I know that won't happen anytime soon.
>
> Best Regards,
>
> Joe Russeau
>
>
>
> ----- Original Message ----- From: "Brian Ellis" <[log in to unmask]>
> To: <[log in to unmask]>
> Sent: Wednesday, April 16, 2008 4:12 AM
> Subject: Re: [TN] SV: [TN] Ionic Contamination Question
>
>
> They may not have been set arbitrarily, but the reference test method as
> per Hobson et al. 1969, in the Naval Avionics Facility, Indianapolis,
> published 1972 as the first text adopted in standards was more than
> arbitrary. It was probably the most unscientific piece of junk ever
> perpetrated on industry. You could get it to read whatever you wanted it
> to read, with a possible fiddle/fudge factor within limits of -100% to
> +500%. In 1969, most of the electronics was on single-sided boards with
> components whose size was larger than many of today's assemblies.
> Typical minimal line widths and spacings were 1 mm each, if not larger.
>
> IOW, that 39-y.o. reference method has absolutely no relevance in
> today's electronics, and nor has Navy report MRR-3-78, which was based
> on it 30 years ago. Mil-P-28809 (1975) and MIL-P-28809A (1981) -- the
> latter permitting commercial instruments with totally artificial fudge
> factors -- were the standards behind the 10 µg/in² eq. NaCl or 1.56/1.5
> µg/cm² eq. NaCl but were hardly worth the paper they were printed on, as
> far as ionic contamination was  concerned.
>
> As has been mentioned on this forum many times, the only way of validly
> using ionic contamination levels in modern electronics is:
> 1. qualify a flux/soldering method
> 2. qualify a cleaning method
> 3. determine if the reliability of the finished article is sufficient,
> using SIR, ECMR, vesication etc. testing and possibly field life testing
> 4. do ionic contamination testing in the instrument of your choice of,
> say, 20 samples of each and every assembly and determine the mean and
> standard deviation.
> 5. do production run testing several times/day in the same instrument
> with statistical analyses of the results. If they fit in the histogram
> of the qualification testing statistics, then you know your process is
> running correctly and you may sleep at night on both ears.
>
> All the above was written in essence on pp 249-257 in a well-known book
> which was published 22 years ago. AFAIK, no one has ever refuted this
> commentary and analysis of the many weaknesses of the reference method
> nor the need to rely on statistical analyses, rather than spot results,
> which I advocated since 1979, just one year after the cited Navy Report.
> In fact, I think it was published in the "Handbook of Contamination"
> (1980): long since out of print.
>
> The 1986 UK Def.Stan 00-10/3 took many of these factors (but not all)
> into account and was, overall, the first ionic contamination test
> standard that was based on more or less scientific principles and it set
> a general limit value of 1.5 µg/cm² eq. NaCl, based on 4 years research
> done at HMS Aquila, in Bromley, Kent. But remember that this was before
> surface mounting became current and, in those days, the average PCB was
> 2 sided PTH with through-hole components and average track
> widths/spacings of about 0.6 mm, 0.3 mm being hi-tech state of the art,
> with relatively low yields. In fact, on p. 260 ibid., the notion of
> different values for different types of assemblies was raised, with 1.0
> µg/cm² eq. NaCl for surface mounting and 0.5 µg/cm² eq. NaCl for
> critical assemblies, and this was over 22 years ago.
>
> I maintain that anyone who blindly follows a limit of 1.5 µg/cm² eq.
> NaCl or any other fixed value for spot checks, based on 39-year old
> methods, at the best doubtful, needs to retire from industry before he
> causes his employers the risk of untold harm.
>
> Sorry for  being blunt! :-)
>
> Brian
>
> [log in to unmask] wrote:
>> The levels were not set arbitrarily. They were set on basis of ionic 
>> residues causing vesiccation & subsequent failure of military assemblies 
>> in  high humidity environments (Vietnam). Specifically, ionic residues 
>> from RMA  fluxes on conformally coated assemblies- most often radios. 
>> (See Navy report  MRR-3-72). Mechanization of the original manual ionic 
>> testing method led to the  different nbrs used for the various automated 
>> test instruments, as published in  Navy report MRR-3-78. I wrote a column 
>> for SMT magazine on this  issue; if you'd like a copy of the column, 
>> email me separately at [log in to unmask] (mailto:[log in to unmask])  as 
>> it would be stripped  off if I tried to email it to TechNet as an 
>> attachment.
>> Bill  Kenyon
>> Global Centre Consulting
>> 3336 Birmingham Drive
>> Fort Collins, CO  80526
>> Tel: 970.207.9586   Cell:  970.980.6373
>>
>>
>>
>>
>> **************It's Tax Time! Get tips, forms and advice on AOL Money & 
>> Finance.      (http://money.aol.com/tax?NCID=aolcmp00300000002850)
>>
>> ---------------------------------------------------
>> Technet Mail List provided as a service by IPC using LISTSERV 15.0
>> 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 or (re-start) delivery of Technet send e-mail to 
>> [log in to unmask]: SET Technet NOMAIL or (MAIL)
>> To receive ONE mailing per day of all the posts: send e-mail to 
>> [log in to unmask]: SET Technet Digest
>> Search the archives of previous posts at: 
>> http://listserv.ipc.org/archives
>> Please visit IPC web site 
>> http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional 
>> information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 
>> ext.2815
>> -----------------------------------------------------
>>
>
> ---------------------------------------------------
> Technet Mail List provided as a service by IPC using LISTSERV 15.0
> 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 or (re-start) delivery of Technet send e-mail to 
> [log in to unmask]: SET Technet NOMAIL or (MAIL)
> To receive ONE mailing per day of all the posts: send e-mail to 
> [log in to unmask]: SET Technet Digest
> Search the archives of previous posts at: http://listserv.ipc.org/archives
> Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 
> for additional information, or contact Keach Sasamori at [log in to unmask] or 
> 847-615-7100 ext.2815
> -----------------------------------------------------
>
> ---------------------------------------------------
> Technet Mail List provided as a service by IPC using LISTSERV 15.0
> 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 or (re-start) delivery of Technet send e-mail to 
> [log in to unmask]: SET Technet NOMAIL or (MAIL)
> To receive ONE mailing per day of all the posts: send e-mail to 
> [log in to unmask]: SET Technet Digest
> Search the archives of previous posts at: http://listserv.ipc.org/archives
> Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 
> for additional information, or contact Keach Sasamori at [log in to unmask] or 
> 847-615-7100 ext.2815
> -----------------------------------------------------
>

---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 15.0
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 or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-----------------------------------------------------