Steve:
A more realistic approach is that the end test of whether or not a solder sample is good isn't whether or not it passes some presupposed test regimen. Rather, whether or not it can judiciuosly solder parts onto the PCB without impact to your customer's operations (inclusive or quality and reliability). I would think there are a couple of avenues-first, what's the intent of your test? If it is to check your process, and not the solder materials, then increasing your frequency of testing would minimize the risk. If it is to test the solder materials, then you would want to environmentally age the paste sample and test each lot of solder. This would reduce you risk as well.
Wish you luck...
Dr. Bill
> -----Original Message-----
> From: Stephen R. Gregory [SMTP:[log in to unmask]]
> Sent: Monday, June 01, 1998 10:43 AM
> To: [log in to unmask]
> Subject: Re: [TN] Solder Sample Testing?
>
> In a message dated 6/1/98 8:31:18 AM Pacific Daylight Time, [log in to unmask]
> writes:
>
> <<Ladies & Gentlemen,
> I am currently working on a schedule for solder pot sampling. The
> pots will be tested every 30 days and two samples will be taken from
> each pot. This way if one test happens to fail, we can send the other
> sample to verify the test. The problem I have ran into is what to do
> if it fails. In a recent customer audit I was asked what kind of
> containment would we have if the test failed? Since we are using the
> ANSI/J-STD-001A there are clearly defined limits that must be met. It
> is my understanding that when you test the solder you are testing for
> the last 30 days. The problem I have is this. If I test the pot on
> 1-1-98 and it passes I'm fine. But when I test again on 2-1-98 and it
> fails I have a huge problem. Since the last day I can prove the
> solder was within limits was Jan, 1 that makes everything suspect bad
> material to that date. I can't write the procedure to recall
> everything for the last 30 days, I need a different solution. If you
> have any information on how your company handles this please let me
> know. Please keep in mind, that we must meet the ANSI-STD and any
> element out of range is considered a failure to our customer,
> regardless if the solder integrity is affected or not.
>
> Thanks
> Chris Smith
> Manufacturing Engineer
>
> ############################################################# >>
>
> Chris,
>
> I can understand your problem trying to think of what to do if your pot
> were to fail in between sample intervals...I mean you're right, who wants to
> think of the logistical nightmare it would be trying to recall everything one
> has built over the last 30-days? Makes the hair on the back of my neck stand-
> up just thinkin' about it...
>
> But to look at things realistically, how would that ever happen if
> you're paying close attention to the results of the samples you're submitting,
> and have some sort of idea of the volume of PCB's that have run across the
> wave in those 30-day periods. It's been my experience that wave solder pots
> don't suddenly go wacko over a 30-day period unless something unusual happens.
> You can spot trends taking place in your analysis over time, and then have an
> idea of what happens to your pot based on the volumes you run over a thirty
> day period. You should also be able to see what happens when you run a lot of
> gold plated boards vs. HASL, OSP, etc., then be able to have some sort of idea
> what it might take to throw your pot out.
>
> That's something that you need to get a history on anyway, because the
> volume of solder that your machine holds is most likely going to be different
> than the pot I have, and will be affected differently in the percentage of
> different metal introduced than my pot for a given volume of boards waved.>
>
> So, I guess in a nutshell you need to get to know your solder pot,
> become familiar with it's make-up, and watch what it eats
> (GRIN)...establishing a baseline, and then monitoring things should keep
> everything under control.
>
> What you might want to tell the next auditor so he has a "warm fuzzy",
> is that you've established a baseline, you monitor it once a month, and then
> explain to him that it's very, very rare for a pot to go out of limits because
> you're monitoring it once a month.
>
> For the times that you find yourself waving a majority of boards that's
> different from the normal mix of boards you wave (such as all gold or all OSP
> boards), you may want to increase to testing to twice a month during those
> production conditions.
>
> One other time that I can think of that might warrant a special analysis
> is if you start having defects that you're not used to seeing. There are
> certain types of defects that can be contributed to, or caused by a
> overabundance, or lack of, certain metals. So that's another way you can
> explain how closely you monitor your wave, by watching how current product is
> running.
>
> I've never had a pot go bad on me except for one time, it got low on
> tin. The reason I had that was because the operator I had turned the chip wave
> on all the time whether there was SMT on the bottom or not (a bad habit he
> picked-up somewhere), and he didn't use the care he needed to when cleaning
> the dross out of the pot...my recycler was giving me a lot for my dross
> though! But that's what was eating the tin.
>
> I've seen the gold level get close a couple of times in the past when
> the boards were being plated with a lot thicker gold than what's used today,
> but that's all I've ever seen.
>
> I hope this helps a little...
>
> -Steve Gregory-
>
> ################################################################
> TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
> ################################################################
> To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
> To subscribe: SUBSCRIBE TechNet <your full name>
> To unsubscribe: SIGNOFF TechNet
> ################################################################
> Please visit IPC web site (http://jefry.ipc.org/forum.htm) for additional information.
> For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312
> ################################################################
>
################################################################
TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
################################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe: SUBSCRIBE TechNet <your full name>
To unsubscribe: SIGNOFF TechNet
################################################################
Please visit IPC web site (http://jefry.ipc.org/forum.htm) for additional information.
For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312
################################################################
|
