Lead-free only? failure mode ?

[N_Cook]

Mackie ProFX12 , a couple of years old. OK at soundcheck, but failed in
first use proper. Owner retained one XLR feed and used a 1/4 inch line
for the other channel and continued.
With me checks out fine. Now the double XLR boards has that style of
soldering , pin and cup?, where the end of the pin is enveloped in
solder. Then as 4 closely spaced pins for the XLR difficult to check by
flexing. Will try tomorrow with proper continuity check on all pins.
I've seen this failure before , but I don't think with proper solder.
Does PbF expand on cooling and so the cup helps to break contact, more
than just a ring of solder, or just cold-solder joint failuredue to
thick + long pins so substantial heatsinking.
At the moment no sign of any solder sweated thru the presumably plated
holes to the component side of the pcb, "nice" clean gold finish,
suggestive of inadequate heating at soldering.
Hopefully a pbf problem on that rear board and not balanced line feed
buffer amp failure , requires removing all the top panel hardware.

[Paul Drahn]

Lead-free solder does not flex. It fractures. Lead based solder will
flex because of the lead.

Paul

[N_Cook]

a pic of the "pin and cups"
http://www.diverse.4mg.com/mackie_pbf.jpg

[N_Cook]

I wonder if there is a significance to the sharp deliniation between the
mirror finish of the cup part and the maculate finish of the sleeve
part. Indicative of differential cold-shocking while cooling ?

[Gareth Magennis]

"N_Cook" <div...@tcp.co.uk> wrote in message
news:m4sp8e$ds7$1...@dont-email.me...

Was this XLR output being fed to a balanced input (power amp)? If so then
you would probably have to have both pins 2 and 3 solder joints fail at the
same time to lose signal altogether. If only one failed you would get a 3dB
loss of signal.

A possible culprit is the Insert jack socket on that output channel, which
may have had a dirty normally closed switch contact. Plugging a jack into
the nearby line out jack socket could have mechanically jolted it back into
life again.




Gareth.

[N_Cook]

I'd not thought about that, will have to test the bypass breaking force
etc on the Inserts. So fixated on trying to prove a PbF touching
contact. I'd liked to have seen movement or measured 10 ohm or more ,
consistently somewhere
The 2 break situation is consistent with the following sorry saga,
perhaps, but I've concentrated on one "solder" joint, could be one
regularly bad and sometimes its mate goes bad. A regular 3db could
easily be masked by unmatched sliders or something.
Picked up the board this morning and first contact I went to measured 40
to 60 Kohms, tried the rest and .5R or so. Returned to the first and
similar .5r now, so has the contact remade or was it oxidised/dirty
contact related to my first DVM reading.
Felt-tipped the exposed pin/standoffs and flexed the pcb and no shifting
seen .
Soldered a good contact away from the first suspect joint and fitted a
proper XLR into the socket , to make good DVM monitoring points.
All measured .5R or so and tended to reduce a bit on heating the pcb to
50 deg C or so and increase to 1.5 ohm or so on freezer spraying the
extender parts of the pins.
So suspicious but not conclusive . Tried the first contact with passing
3 amp through it , as no active stuff on this board. Dropped just 20mV
at the contact. Returning to DVM-R and now it was .1R or so , but
increasing to .5R or so on pcb low-level heating.
Decided to grind off the cups of the solder joints , bite by bite , to
avoid melting solder. Again tried twisting the pcb and deflecting and no
movement seen between the brass and the solder .

[N_Cook]

I desoldered the 2 ground pins. Then tried just finger pressure to pull
out the remaining pins. I assume from my previous twisting, or maybe
from being bashed in use at some time. The brass pins failed.
These are 1.6mm diameter and the pcb holes are for that size. But these
Neutrik XLR would seem to be made for old-style hand-wiring then solder,
not pcb use. There is a turned recess ring on each pin, down to 1mm
diameter, to take a loop or twist of wire before soldering.
This recess is the thickness of the pcb. With the pins sheared off now
and the remnant stubs still in the pcb , you can see that no solder at
all has passed into the .3mm gap around the pins. So in effect the 3
main pins (not the frame ground pin) are only "surface mount" soldered ,
not through-board soldered, so mechanically weak joints regardless of
use of PbF.
The shoulder length of the frame ground pin is right for pcb mounting,
so are these recessed pins intended for pcb with eyelets , but not used
here, or for handwiring?

[N_Cook]

I cannot find an exact image of them
http://www.mondospettacoli.it/shop/index.php?main_page=popup_image&pID=6854
is the nearest long standoffs, is the distant ends of the pins, in that
image, eyelets near the ends

[Phil Allison]

Paul Drahn wrote:

> Lead-free solder does not flex.

** The term "lead free" incudes a huge variety of solders nearly all of which are quite flexible.

As wire from a reel, 99C (tin/copper) solder differs little from 60/40 in terms of flexibility.

> It fractures. Lead based solder will
> flex because of the lead.

** Strange how cracked solder joints on PCBs long predatse the use of Pb free solders.

IME - the main reason for bad joints with Pb free solders is failure of the base metal to alloy with the solder - partly due to the use of almost pure tin and partly to do with the use of low activity, water soluble fluxes.


... Phil

[N_Cook]

no bypass contacts on insert mode. They've skimped on having no MUTE
reminder LEDs per channel and one under the bargraph that mutes the
whole lot, unLEDed, just where you would unknowingly push it on handling
the mixer, 10 minutes wasted there