Reply by Paul October 30, 20122012-10-30
In article <afara8F64vU1@mid.dfncis.de>, HBBroeker@t-online.de says...
> > On 29.10.2012 20:07, rickman wrote: > > > Solder is not really all that strong if > > subjected to repeated stress. It seems to crack easily. > > Exactly. A solder joint consists of three regions of material: the > copper pad on one side, solder on the other, and an alloy of the two in > between. Now here's the problem: that alloy is _brittle_, and weak. > It'll crack under bending or shearing loads much more easily than either > of its constituent materials. It's also seriously vulnerable to Varying > stress (i.e. vibration, or repeatedly applied stress).
If you want to make the joint even more brittle, such that dry joints appear over time or temperature cycling. Add Gold ! Yes soldering onto gold plated contacts lifts gold plating into the solder making the solder even more brittle. A lot of avionics and military specs for components specify parts to have a solder finish, NO Gold. If you do double desolder on parts you have assay the solder used to remove gold to calculate the full amount of gold has been removed.
> In a nutshell the OP's design would be setting up a race between two > failure modes. The only question would be which of the two joints will > fail first: pin--solder--pad, or pad--glue--board. But it's pretty damn > certain that one of them would fail.
-- Paul Carpenter | paul@pcserviceselectronics.co.uk <http://www.pcserviceselectronics.co.uk/> PC Services <http://www.pcserviceselectronics.co.uk/fonts/> Timing Diagram Font <http://www.gnuh8.org.uk/> GNU H8 - compiler & Renesas H8/H8S/H8 Tiny <http://www.badweb.org.uk/> For those web sites you hate
Reply by Hans-Bernhard Bröker October 30, 20122012-10-30
On 29.10.2012 20:07, rickman wrote:

> Solder is not really all that strong if > subjected to repeated stress. It seems to crack easily.
Exactly. A solder joint consists of three regions of material: the copper pad on one side, solder on the other, and an alloy of the two in between. Now here's the problem: that alloy is _brittle_, and weak. It'll crack under bending or shearing loads much more easily than either of its constituent materials. It's also seriously vulnerable to Varying stress (i.e. vibration, or repeatedly applied stress). In a nutshell the OP's design would be setting up a race between two failure modes. The only question would be which of the two joints will fail first: pin--solder--pad, or pad--glue--board. But it's pretty damn certain that one of them would fail.
Reply by Rocky October 30, 20122012-10-30
On Oct 29, 8:03&#4294967295;pm, Rob Gaddi <rga...@technologyhighland.invalid>
wrote:
> On Mon, 29 Oct 2012 10:53:47 -0700 > > > > > > Jim Stewart <jstew...@jkmicro.com> wrote: > > langw...@fonz.dk wrote: > > > On 28 Okt., 22:53, Hans-Bernhard Br&#4294967295;ker<HBBroe...@t-online.de> &#4294967295;wrote: > > >> On 28.10.2012 21:34, Paul wrote: > > > >>> Personally if the switch is so close to the outside, that you need to > > >>> save height, I would be looking at touch sensors anyway, just areas of > > >>> PCB track. > > > >> But those come with their own can of worms attached that's bigger and > > >> heavier than an entire PC keyboard's worth of conventional push buttons. > > > >> SMD push buttons are routinely below 4 mm in height, and can be had down > > >> to 1.2 mm. &#4294967295;Rubber-mat-with-conductor-pads type buttons are of similar > > >> height. &#4294967295;If you can fit neither of those, then your PCB is entirely too > > >> close to the front panel ;-P > > > > my immediate guess is that he wants to use the pcb as the front panel > > > > All components on the back side, button actuators protruding through > > > holes in the PCB, add some reverse mount leds silkscreen what ever > > > text > > > needed > > > Close. &#4294967295;Using the PCB as an end panel for an extruded > > enclosure. &#4294967295;Small LCD mounted on the inside pointing > > out through a hole. &#4294967295;Switches mounted on the backside > > poking through holes, laser cut Mylar overlay on the > > front. > > > Hans concern about the switch lead fragility is something > > I haven't thought about and needs to be considered. > > I think that's less a concern about the fragility of the lead itself, > and more a concern about the connection between the pad and the FR-4. > I think you'll wind up with a perfectly sturdy solder joint, still > connected to the pad and a dangling bit of trace that's ripped clean > off.
My experience has been that the solder suffers from stress fractures.
Reply by lang...@fonz.dk October 29, 20122012-10-29
On 29 Okt., 18:54, Jim Stewart <jstew...@jkmicro.com> wrote:
> langw...@fonz.dk wrote: > > On 28 Okt., 22:53, Hans-Bernhard Br&#4294967295;ker<HBBroe...@t-online.de> &#4294967295;wrote: > >> On 28.10.2012 21:34, Paul wrote: > > >>> Personally if the switch is so close to the outside, that you need to > >>> save height, I would be looking at touch sensors anyway, just areas of > >>> PCB track. > > >> But those come with their own can of worms attached that's bigger and > >> heavier than an entire PC keyboard's worth of conventional push buttons. > > >> SMD push buttons are routinely below 4 mm in height, and can be had down > >> to 1.2 mm. &#4294967295;Rubber-mat-with-conductor-pads type buttons are of similar > >> height. &#4294967295;If you can fit neither of those, then your PCB is entirely too > >> close to the front panel ;-P > > > my immediate guess is that he wants to use the pcb as the front panel > > > All components on the back side, button actuators protruding through > > holes in the PCB, add some reverse mount leds silkscreen what ever > > text > > needed > > Close. &#4294967295;Using the PCB as an end panel for an extruded > enclosure. &#4294967295;Small LCD mounted on the inside pointing > out through a hole. &#4294967295;Switches mounted on the backside > poking through holes, laser cut Mylar overlay on the > front. > > Hans concern about the switch lead fragility is something > I haven't thought about and needs to be considered.
if you add an overlay anyway maybe just some pads on the pcb and these? http://www.clikdomes.com/11.html -Lasse
Reply by rickman October 29, 20122012-10-29
On 10/29/2012 2:03 PM, Rob Gaddi wrote:
> On Mon, 29 Oct 2012 10:53:47 -0700 > Jim Stewart<jstewart@jkmicro.com> wrote: > >> langwadt@fonz.dk wrote: >>> On 28 Okt., 22:53, Hans-Bernhard Br&ouml;ker<HBBroe...@t-online.de> wrote: >>>> On 28.10.2012 21:34, Paul wrote: >>>> >>>>> Personally if the switch is so close to the outside, that you need to >>>>> save height, I would be looking at touch sensors anyway, just areas of >>>>> PCB track. >>>> >>>> But those come with their own can of worms attached that's bigger and >>>> heavier than an entire PC keyboard's worth of conventional push buttons. >>>> >>>> SMD push buttons are routinely below 4 mm in height, and can be had down >>>> to 1.2 mm. Rubber-mat-with-conductor-pads type buttons are of similar >>>> height. If you can fit neither of those, then your PCB is entirely too >>>> close to the front panel ;-P >>> >>> my immediate guess is that he wants to use the pcb as the front panel >>> >>> All components on the back side, button actuators protruding through >>> holes in the PCB, add some reverse mount leds silkscreen what ever >>> text >>> needed >> >> Close. Using the PCB as an end panel for an extruded >> enclosure. Small LCD mounted on the inside pointing >> out through a hole. Switches mounted on the backside >> poking through holes, laser cut Mylar overlay on the >> front. >> >> Hans concern about the switch lead fragility is something >> I haven't thought about and needs to be considered. >> > > I think that's less a concern about the fragility of the lead itself, > and more a concern about the connection between the pad and the FR-4. > I think you'll wind up with a perfectly sturdy solder joint, still > connected to the pad and a dangling bit of trace that's ripped clean > off.
I've seen problems with surface mount connectors coming loose from PCBs when used frequently. I've seen both the solder joint come loose and the pad pull off the PCB. Solder is not really all that strong if subjected to repeated stress. It seems to crack easily. On my designs *all* connectors are secured by at least two through hole pins. Rick
Reply by Rob Gaddi October 29, 20122012-10-29
On Mon, 29 Oct 2012 10:53:47 -0700
Jim Stewart <jstewart@jkmicro.com> wrote:

> langwadt@fonz.dk wrote: > > On 28 Okt., 22:53, Hans-Bernhard Br&ouml;ker<HBBroe...@t-online.de> wrote: > >> On 28.10.2012 21:34, Paul wrote: > >> > >>> Personally if the switch is so close to the outside, that you need to > >>> save height, I would be looking at touch sensors anyway, just areas of > >>> PCB track. > >> > >> But those come with their own can of worms attached that's bigger and > >> heavier than an entire PC keyboard's worth of conventional push buttons. > >> > >> SMD push buttons are routinely below 4 mm in height, and can be had down > >> to 1.2 mm. Rubber-mat-with-conductor-pads type buttons are of similar > >> height. If you can fit neither of those, then your PCB is entirely too > >> close to the front panel ;-P > > > > my immediate guess is that he wants to use the pcb as the front panel > > > > All components on the back side, button actuators protruding through > > holes in the PCB, add some reverse mount leds silkscreen what ever > > text > > needed > > Close. Using the PCB as an end panel for an extruded > enclosure. Small LCD mounted on the inside pointing > out through a hole. Switches mounted on the backside > poking through holes, laser cut Mylar overlay on the > front. > > Hans concern about the switch lead fragility is something > I haven't thought about and needs to be considered. >
I think that's less a concern about the fragility of the lead itself, and more a concern about the connection between the pad and the FR-4. I think you'll wind up with a perfectly sturdy solder joint, still connected to the pad and a dangling bit of trace that's ripped clean off. What about tactile dome switches? Some of those can be had at only about a half millimeter. -- Rob Gaddi, Highland Technology -- www.highlandtechnology.com Email address domain is currently out of order. See above to fix.
Reply by Jim Stewart October 29, 20122012-10-29
langwadt@fonz.dk wrote:
> On 28 Okt., 22:53, Hans-Bernhard Br&#4294967295;ker<HBBroe...@t-online.de> wrote: >> On 28.10.2012 21:34, Paul wrote: >> >>> Personally if the switch is so close to the outside, that you need to >>> save height, I would be looking at touch sensors anyway, just areas of >>> PCB track. >> >> But those come with their own can of worms attached that's bigger and >> heavier than an entire PC keyboard's worth of conventional push buttons. >> >> SMD push buttons are routinely below 4 mm in height, and can be had down >> to 1.2 mm. Rubber-mat-with-conductor-pads type buttons are of similar >> height. If you can fit neither of those, then your PCB is entirely too >> close to the front panel ;-P > > my immediate guess is that he wants to use the pcb as the front panel > > All components on the back side, button actuators protruding through > holes in the PCB, add some reverse mount leds silkscreen what ever > text > needed
Close. Using the PCB as an end panel for an extruded enclosure. Small LCD mounted on the inside pointing out through a hole. Switches mounted on the backside poking through holes, laser cut Mylar overlay on the front. Hans concern about the switch lead fragility is something I haven't thought about and needs to be considered.
Reply by October 28, 20122012-10-28
On Saturday, October 27, 2012 1:33:36 PM UTC-5, Jim Stewart wrote:
> Does anyone know of a pushbutton switch > > that can be surface mounted to the back > > side of a PCB with the actuator button > > protruding through a hole in the PCB?
I saw a patent for a tact switch like this, but it was not listed as being assigned to anyone. You didn't mention if you needed push-push or momentary. For momentary, how about a snap dome, they are pretty low profile. I've used a few from snaptron on prototypes and I think Omron also makes them.
Reply by lang...@fonz.dk October 28, 20122012-10-28
On 28 Okt., 22:53, Hans-Bernhard Br&#4294967295;ker <HBBroe...@t-online.de> wrote:
> On 28.10.2012 21:34, Paul wrote: > > > Personally if the switch is so close to the outside, that you need to > > save height, I would be looking at touch sensors anyway, just areas of > > PCB track. > > But those come with their own can of worms attached that's bigger and > heavier than an entire PC keyboard's worth of conventional push buttons. > > SMD push buttons are routinely below 4 mm in height, and can be had down > to 1.2 mm. &#4294967295;Rubber-mat-with-conductor-pads type buttons are of similar > height. &#4294967295;If you can fit neither of those, then your PCB is entirely too > close to the front panel ;-P
my immediate guess is that he wants to use the pcb as the front panel All components on the back side, button actuators protruding through holes in the PCB, add some reverse mount leds silkscreen what ever text needed -Lasse
Reply by rickman October 28, 20122012-10-28
On 10/28/2012 5:53 PM, Hans-Bernhard Br&#4294967295;ker wrote:
> On 28.10.2012 21:34, Paul wrote: > >> Personally if the switch is so close to the outside, that you need to >> save height, I would be looking at touch sensors anyway, just areas of >> PCB track. > > But those come with their own can of worms attached that's bigger and > heavier than an entire PC keyboard's worth of conventional push buttons. > > SMD push buttons are routinely below 4 mm in height, and can be had down > to 1.2 mm. Rubber-mat-with-conductor-pads type buttons are of similar > height. If you can fit neither of those, then your PCB is entirely too > close to the front panel ;-P >
You could use the Forth methodology. If the problem is too hard to solve, solve a different problem. Rick