[Elecraft] K1: C67 Interferes With Mounting KNB1

Fri Mar 22 19:14:37 EDT 2013

Thanks, Don--

Tony has requested specifically that I clean his boards, so I'm doing
so, else might not bother.

Flux (connector mostly) cleaning (as taught at JPL, up to my last year
there, 2004) involved an acid brush, a small syringe, and Isopropyl
Alcohol, or for some tough cases with other contaminants, Acetone (under
a fume hood).  Safety glasses and gloves are required. 

At home, I'm using an old toothbrush and grocery-store Isopropyl 99% or
91% 'IPA91' or 'IPA99' at the kitchen sink.  Brush with a wetted brush
and the board slanted so solvent and solutes runs off the edge, squirt a
little more to keep that flow moving.  I agree it can be hard to avoid
some white residue buildup without flooding the area to flush away the
dissolved flux, and telling what type of flux you have is tricky; R
(resin), RA (resin activated). or RMA (resin Mildly Activated).  If it's
not marked, and with certs, how can you tell without sending some off to
a lab?

On my own projects, I will usually skip de-fluxing, as I'm almost frugal
with the solder.  I still have about 9 oz. left of my old roll of Ersin
Multicore Sn63 (22SWG .028 In., R, P3, QQ-S-571E) so small clean fillets
are the general rule.  When that's gone, there's Kester 282.

[A Horror Story - Out Of School (second hand, may be apocryphal, no
warranty on it.)]

Much grief has been caused in past years by cleaning with more
aggressive solvents, such as trichlorethylene (or trichlorethane) on
flight connectors at delivery and turnover of a spaceflight instrument
at the Cape (as the story goes that I heard).  When the instrument was
found not to be working, investigation showed that the solvent had
caused Solithane (a polyurethane potting material used as conformal
coating) inside, behind the connector on the board, to swell, literally
de-laminating copper traces and pushing gull-winged DIP devices off the
circuit board.  This became a major push-up for most flight hardware in
development, causing delays and schedule stress on Galileo and many
other projects, and on the ATLAS/ACRIM solar radiometer instrument where
I worked at the time.  A lot of overtime over in that lab in Bldg. 103,
while the assemblers removed parts, cleaned off every bit of Solithane
that had been under the parts, inspected and rebuilt the boards,
inspected some more, conformal-coated only over (not under) the chips,
found some few put back on backwards in the rush, cycle through again -
a nightmare!  Meanwhile, Jupiter continued in its inexorable orbit;
orbital mechanics waits for no man (or woman), and launch windows depend
on that timing, a precise choreography.

It's said in aerospace that "every cycle of rework is one step closer to
Scrap!"  That's for whatever reason, so avoiding that hazard makes
perfect sense!

On 03/21/2013 07:16 PM, Don Wilhelm wrote:
> Your "fix" for the too large C67 is likely OK.  I have not yet
> encountered that situation.
>
> However, I offer a caution about the flux 'wash' you mentioned.  I
> trust you do not mean to flood the board with flux remover  I have
> attempted repairs on too many radios that have been "flux washed"
> improperly
> In some cases, flux residue has been washed under the connectors and
> ICs and the only fix is to remove the components and further clean
> both the component and the board.  It makes one great mess and can
> result in a malfunctioning radio.  Even after such repairs, I 'cross
> my fingers' and hope the fault conditions do not reappear.
>
> For what it is worth, HP does flux cleaning of production boards, but
> I understand from reliable sources that those boards that end up at
> rework stations are *not* to be cleaned of flux.  There must be a good
> reason for that procedural directive.
>
> Solder flux is non-conductive - as long as no effort has been made to
> clean the flux.
> The combination of flux and flux cleaners can leave conductive paths
> that will react later to moisture in the air and cause malfunctions.
>
> If you *must* clean flux, do it in small areas at a time using the
> cleaner (denatured alcohol works) applied with a cotton swab (Q-tip)
> and cleaned up with a paper towel piece immediately - repeat until all
> the flux is gone.  Do not "wash" or "flood" the board or disaster may
> occur.
>
> If you switch from solder with a Highly Reactive flux (Kester 44 for
> example) to a solder with a Mildly Reactive flux, no cleaning is
> nesessary (there is very little flux residue).  Kester 285 is one
> example and the Kester 'No Clean' is even better.  Save the Kester 44
> for work on antennas and old components which have substantial
> oxidation on the leads.
>
> 73,
> Don W3FPR
>
> On 3/21/2013 9:15 PM, Phil Barnes-Roberts WA6DZS wrote:
>> Just finishing up a build for Tony, KE6JZF of a K1/KFL1-4/KAT1/KNB1
>> (#3195), and have found that the 2.2uF-50V electrolytic cap C67 (where
>> 25V is called out in the Parts List) may be a bit too tall.
>>
>> It doesn't seem to bother at C31, C54, or C10 on the NB board (all
>> 50V).  When it came time to bolt the NB to the RF board, C67 was in the
>> way.  After trying to gently re-wet the leads and lean it over U2, I
>> wound up wrapping a little wire around the (too-short) leads and
>> mounting it on its side under the board (which seems to fit OK.)
>>
>> Anybody seen a better fix for this physical interference fit?  Final
>> flux wash is the next step, unless I hear better news.
>>
>

-- 
'---O=o=O---'
73, Phil Barnes-Roberts WA6DZS  DM04we | Mailto:pbarnrob at ACM dot org 
42 is not an answer, it's an error code.  The universe is saying 
'Error 42: meaning to universe not found'  --DoctorMO, ubuntuforums.org