Enclosure advice please - glueing metal

[Dave Marsh]

I've used double-sided adhesive tape (the 1mm thick foamy type).
It is extremely strong, taking it off if you make a mistake is a real pain in the butt.

It also provides a very small but usefull bit of flexibility (any hard glue will crack,
epoxy will crack when glueing metal to metal - its not shock proof).

The front panel side is easy to glue this way, you need a convenient matching flat surface
on the PCB, edges of the LCD display (5mm or so on each side) are more than enough
for a small PCB.

In fact I prefer this solution to screws for mounting LCD boards.

SioL

Hi SioL. Thanks for your recommendation. I've just got to work out how to
use the tape in our case, since the PCB needs to be about 0.25" (6.5mm) back
from the front panel. I have no flat surfaces near enough to the panel (I'm
using 7-seg LED displays). I'm thinking I could use the tape to stick to the
back of flat headed machine screws and then bolt the PCB to those.

Just wondering if you or anyone else has experience with how long tape stays
stuck? Are there any particular brands you'd recommend or avoid? Thanks.

Dave

 

[Dave Marsh]

If the PCB has mains voltage on it, you may wish to wrap the entire
board with several layers of black electrical tape.

Best regards,
Spehro Pefhany

http://www.speff.com

In our case it's all running off a 9V DC wall wart, but thanks for the
advice!

Dave

 

[Dave Marsh]

That's a decent solution. You can press flush fasteners *through* the
inner panel to make an almost indestructable assembly. You do have to
machine or punch both panels similarly, though.

Best regards,
Spehro Pefhany

http://www.speff.com

Thanks Alan and Spehro. It does sound very good, both in terms of robustness
and tidiness. I'm getting the panels made outside (FrontPanelExpress) so I
guess I could get them to do inner panels at the same time. Just need to
make sure I line them up accurately when it comes to glueing the two panels
together - the inner panel will have to be smaller than the outer one due to
the requirements of the enclosure. I could make some kind of jig I guess.

Dave

 

[SioL]

On Wed, 9 Jun 2004 17:41:18 +0200, the renowned "SioL"


If the PCB has mains voltage on it, you may wish to wrap the entire
board with several layers of black electrical tape.

Best regards,
Spehro Pefhany

Perhaps I did not describe this setup well enough, I'm sure if you'd see it you'd appreciate
it for both simplicity, ease of installation, incredible stability, professional look when done
properly plus you get a free shock absorber for the PCB/display since the tape feels a bit like
rubber.

Its just nasty if you want to take if off afterwards. Its easy to bend/destroy the LCD since
it adheres so well.

Another disadvantage is that you need something flat like the edge of the LCD to place the tape.

SioL

 

[Frank Bemelman]

Those self-adhesive spacers look very good. Unfortunately the 0.5" square
base won't fit on our panel due to holes for LEDs etc. The second aluminum
panel does look a strong candidate now - thanks for the advice Frank.

Yes, they do take a lot of space. If you have not yet done the layout
for the PCB, you can place the spacers where there is space on the panel,
and match the holes on the PCB. Or cut off a piece and make it smaller
than a 0.5" x 0.5" footprint. But I also prefer a second panel.

 

[Activ8]

Thanks for your sensible advice Mike. You are pretty much right that I
screwed myself with this design somewhat! It's my first one and am learning
the hard way.

The panel is just 4" x 1" and made by FrontPanelExpress.com (I can heartily
recommend them btw). The "printing" is actually done by them as dyed
engraving and it looks great on the brushed aluminum, so I really don't want
to go to a sticky plastic overlay (it's a niche product for consumer audio
so needs to look pretty decent alongside people's amps etc.)

You know, a mathing (or even other color) socket head cap screw
(that's a head that takes an allen wrech) or torx screw doesn't look
bad at all IMO.

The screw thread idea sounds good, but the lack of thickness of the panel
may well be a problem as you say.

On the up side, there are just three 4.5mm diameter tactile buttons that the
user *should* press, and I think the strength of the adhesive is pretty good
in this direction, i.e. parallel with the panel. My leverage test was trying
to simulate a worst case and it succeeded!

Idea! You'll get better adhesion with more surface area so mount the
board to a subpanel/chassis with the tacts sticking out and glue the
sub panel to the back of the panel.

As you say, epoxy is no alternative to proper mechanical fastening, but I'm
hoping it is sufficient in my case. So I wonder whether sticking nylon
spacers to the aluminum is going to produce a tougher bond than using
nickel-plated brass ones? And also what is the best type of adhesive to use
in either case?

Maybe. I'd have to experiment, but Liquid Nail is good and JB weld
is good. The LN is probably better for larger surface areas. I just
had a wooden chair repair fail with JB Weld - lasted a year or so -
because I should've drilled out the seat and leg to accept a dowel -
then carpenters glue would've sufficed.

 

[Activ8]

Nylon is almost impossible to attach to anything with glue of any kind
I have ever tried.

Yeah me too, now that I think of it. Never could repair broken
nylon, anyway.

Epoxy has exceptional adhesion to anodized
aluminum, better than to plated or unplated brass, I think. It also
has good stick to thermally oxidized aluminum. You can make good
clean thermal oxide by baking aluminum parts in a 450 degree oven for
an hour or so. But any clean aluminum has a thin film of oxide on it.

Most epoxies don't develop their full strength for a long time, or
until after they are thermally post cured. Can your control panel
tolerate a bake at 120 to 150 degrees F?

Another rule of thumb is that the slower the cure , the higher the
strength, so quick cure epoxies are useless for high strength
applications.

Yup, accelerators mess it up, but "normal" curing time is the way to
go. I isn't the "slow" that helps. Curing retarders also affect the
strength. That doesn't neccessarily mean that using less hardner to
extend the working time isn't ok, within limits (and the
documentation on that is non existant, IME)

Filled epoxy is no stronger in adhesion that plain epoxy. That said,
some of the highest strength epoxies I have used (especially if
properly post cured) are Hysol filled epoxies like 9430 (white) or
9460 gray). These both contain silica, so they cannot be machined by
steel tools, but this doesn't sound like it would be a problem in your
application.

Hmm... Hysol. Noted.

 

[John Fields]

Hi,

I have a small front panel made of anodized aluminum (0.06 inches thick). I
need to space off the PCB (buttons, LEDs etc.) behind it by approx 0.25
inches.

The PCB has 4 drilled holes for M3 machine screws. My problem is that I
can't put holes through the front panel. My idea at the moment is to glue 4
hex spacers to the back of the front panel and then bolt the PCB to it.

So far, I've tried nickel-plated brass spacers attached with epoxy ("plastic
steel" by Draper) on to the sanded aluminum. This came apart after some
pretty gentle leverage testing.

Just wondering (a) if I'd be better using nylon spacers - would these glue
better to the aluminum? And (b) what type of glue should I use for metal
spacers (or nylon spacers) on to anodized aluminum?

---
If you haven't had the panels made yet, you could use PEM CHA-440-XX
concealed-head studs from Penn Engineering, the 'XX' part being
whatever length you'd need so you could slip 1/4" spacers over the
studs and have enough left sticking up past the PCB mounting holes to
let you use whatever hardware you wanted to to secure the PCB to the
panel.

http://www.pennfast.com/fastening_products/pdf/chdata.pdf

They're made out of aluminum so they could be pressed into the panels
and then the visible part of the panel prepped (grained, jitterbugged,
etched, whatever...) before anodizing and dyeing so that the stress
marks from the pressing wouldn't show. Also, since they're aluminum,
they could be anodized after being pressed in, right along with the
panel, no problem.

If you've already had them made, then get some anodized aluminum
threaded standoffs, screw them onto a fixture with the same mounting
hole dimensions as the PCB (and a way to index them to the panel
properly) then use a decent heat-curing epoxy to affix the other ends
of the standoffs to the rear of the panel, put the whole thing in the
oven and cure it according to the epoxy manufacturer's schedule.

I've used some of Master Bond's high-strength heat-curing epoxies and
they've worked well, but I'm sure there are lots of folks out there
with equally good stuff. RTV or hot glue might even work, especially
RTV; since it sure won't crack and it ought to be tenacious as hell on
nice porous surfaces like anodized aluminum!

 

[Activ8]

Hi John. It's a preproduction run of 100 units to test customer feedback.
From the trouble this has caused, we may well change the way the front panel
is made! I will look into the Pem blind spacers - FrontPanelExpress who are
probably going to make our panels offer a similar kind of service but say
the impressions will be visible with only a 1.5mm panel.

The velcro idea is not bad at all, even for 100 units, as long as the
backing adhesive lasts a few years. Then I guess it would match the
longevity of the self-adhesive PCB spacers that have also been mentioned in
this thread.

Many thanks,

Dave

Hey. What's wrong with mounting the board to the back of the case so
it stands out far enough to... you know? What's that? No case?

 

[Activ8]

Those self-adhesive spacers look very good. Unfortunately the 0.5" square
base won't fit on our panel due to holes for LEDs etc. The second aluminum
panel does look a strong candidate now - thanks for the advice Frank.

Dave

Plastic would be cheaper.

 

[Dave Marsh]

You know, a mathing (or even other color) socket head cap screw
(that's a head that takes an allen wrech) or torx screw doesn't look
bad at all IMO.

You are right I reckon - I did consider torx screws a while back, but can't
remember exactly why I dismissed them now... In a perfect world I'd still
prefer a totally "clean" front panel though, especially on such a small
front panel as this one.

Idea! You'll get better adhesion with more surface area so mount the
board to a subpanel/chassis with the tacts sticking out and glue the
sub panel to the back of the panel.

Thanks. I think that is what I am going to try now - either a single
subpanel or maybe a plate for each screw/spacer (as I can make them up
myself and not worry about the alignment until I come to glue the two parts
together).

Maybe. I'd have to experiment, but Liquid Nail is good and JB weld
is good. The LN is probably better for larger surface areas. I just
had a wooden chair repair fail with JB Weld - lasted a year or so -
because I should've drilled out the seat and leg to accept a dowel -
then carpenters glue would've sufficed.

I'll try to give Liquid Nail and JB Weld a go in my next test. Thanks for
the recommendation Mike.

Dave

 

[Dave Marsh]

---
If you haven't had the panels made yet, you could use PEM CHA-440-XX
concealed-head studs from Penn Engineering, the 'XX' part being
whatever length you'd need so you could slip 1/4" spacers over the
studs and have enough left sticking up past the PCB mounting holes to
let you use whatever hardware you wanted to to secure the PCB to the
panel.

http://www.pennfast.com/fastening_products/pdf/chdata.pdf

They're made out of aluminum so they could be pressed into the panels
and then the visible part of the panel prepped (grained, jitterbugged,
etched, whatever...) before anodizing and dyeing so that the stress
marks from the pressing wouldn't show. Also, since they're aluminum,
they could be anodized after being pressed in, right along with the
panel, no problem.

I'll take a look at these, thanks.

If you've already had them made, then get some anodized aluminum
threaded standoffs,

TBH I didn't know anodized aluminum standoffs were available - I suspect
they will stick better than the shiny nickel plated standoffs I've tried so
far.

screw them onto a fixture with the same mounting
hole dimensions as the PCB (and a way to index them to the panel
properly) then use a decent heat-curing epoxy to affix the other ends
of the standoffs to the rear of the panel, put the whole thing in the
oven and cure it according to the epoxy manufacturer's schedule.

I've used some of Master Bond's high-strength heat-curing epoxies and
they've worked well, but I'm sure there are lots of folks out there
with equally good stuff. RTV or hot glue might even work, especially
RTV; since it sure won't crack and it ought to be tenacious as hell on
nice porous surfaces like anodized aluminum!

That's excellent! Many thanks for the advice re the heat-curing epoxies. I
want to try out RTV too soon.

Dave

 

[Dave Marsh]

Hey. What's wrong with mounting the board to the back of the case so
it stands out far enough to... you know? What's that? No case?

I'm using an extruded aluminum Hammond case - 7" front to back, 4" wide, 1"
high. With those dimensions I can't see it being very steady if I fastened
the front panel PCB to the rear panel (plus I'd need to get the right length
of "spacers" made up. I could somehow mount the PCB to the base of the case
I guess. The subpanel idea seems the most practical at the moment TBH.

Dave

 

[Michael]

Hi,

I have a small front panel made of anodized aluminum (0.06 inches thick). I
need to space off the PCB (buttons, LEDs etc.) behind it by approx 0.25
inches.

The PCB has 4 drilled holes for M3 machine screws. My problem is that I
can't put holes through the front panel. My idea at the moment is to glue 4
hex spacers to the back of the front panel and then bolt the PCB to it.

So far, I've tried nickel-plated brass spacers attached with epoxy ("plastic
steel" by Draper) on to the sanded aluminum. This came apart after some
pretty gentle leverage testing.

Just wondering (a) if I'd be better using nylon spacers - would these glue
better to the aluminum? And (b) what type of glue should I use for metal
spacers (or nylon spacers) on to anodized aluminum?

Many thanks,

Dave

There is not very much adhesion area on a standard, e.g. hex, spacer, so
what glue strength you get from it isn't much. But *really* going
against you here is the fact that the shear strength of epoxy is pretty
weak, and your applied forces - both on the flats of the spacer and on
its face (spacer/glue/panel interface) - from simply tightening screws
are shear forces. Epoxies can have very high *tensile* strength but
your application cannot take advantage of them.

So if we accept epoxy's relatively weak shear strength, how about
maximizing the glue surface area and getting a threadable material at
the same time by putting a metal or plastic panel between your board and
the nice front panel? Assuming all your board mounting holes are along
the outter edge of the board, you might then get away with a plastic or
metal panel that has its center cut away ; drill and tap at appropriate
points on the resulting "border".

Or mount metal brackets inside your enclosure (it's OK to drill, e.g.
the bottom of the case, right?) and attach the PC board to those.

 

[Michael]

I'll take a look at these, thanks.


TBH I didn't know anodized aluminum standoffs were available - I suspect
they will stick better than the shiny nickel plated standoffs I've tried so
far.


That's excellent! Many thanks for the advice re the heat-curing epoxies. I
want to try out RTV too soon.

Dave

By all means give RTV a try for your adhesion/seal uses! It's full of
acetic acid though, will corrode many metals.

I can attest to its tenacious grip on anodized aluminum; sticks like
grim death. Ya gotta cut it off. I used RTV in the Air Force to cover
trim-pots that were front panel mounted. We didn't want the operators
dorking with our calibrations. We kept single-edge razorblades in out
toolboxes for slicing off RTV. I tried latching onto a wad of RTV with
pliers and tearing it off: top half came off; bottom layer stayed on the
panel.

 

[John Fields]

By all means give RTV a try for your adhesion/seal uses! It's full of
acetic acid though, will corrode many metals.

---
Shouldn't be a problem with anodized aluminum since what it'll be
sticking to will be the anodized surface, which will be aluminum
oxide. But yeah, it needs to be completely cured before the enclosure
is sealed up. Especially if there are non-hermetically sealed
mechanical swithes in there. BTW, there are acetic acid-free RTV's, I
think...

Another trick to take some the torque load off of the RTV is to put
internal tooth lockwashers between the spacers and the panel side of
the PCB so that when the spacer and the PCB both come in contact with
it as the mounting screws are being screwed down, the washers will
bite into the spacers and the rear of the panel and keep the spacer
from trying to spin.

BUT... I'd still go for the PEM press-in studs if I could. Fewer
parts, no chemistry problems, and just a lot fewer hassles and
downstream worries...

 

[Activ8]

On Wed, 9 Jun 2004 19:25:15 +0100, Dave Marsh wrote:

I'll try to give Liquid Nail and JB Weld a go in my next test. Thanks for
the recommendation Mike.

With a big surface area, I think I'd go with the liguid nail. Maybe
a building contractor can recommend a newerm better adhesive. I see
there's a wider variety at Home Depot/ Lowes these days.

 

[Activ8]

---

BTW, there are acetic acid-free RTV's, I
think...

Yes.
<snip>

 

[John Popelish]

Yup, accelerators mess it up, but "normal" curing time is the way to
go. I isn't the "slow" that helps. Curing retarders also affect the
strength. That doesn't neccessarily mean that using less hardner to
extend the working time isn't ok, within limits (and the
documentation on that is non existant, IME)

I wasn't very clear on this. What I was trying to say is that epoxy
formulas designed to cure slowly usually have a lot higher cured
strength and toughness than formulas designed to cure in a few
minutes. Those fast kinds get little tougher than dried chewing gum.
I was not so much thinking of speeding up or slowing down the cure
time of any particular kind. Epoxy with a 20 to 60 minute working
time just seems to harden to a much stronger bond, when it finally
gets around to curing. It is still stronger if its cure is
accelerated by a bit of heat. It is the chemistry that is better.

 

[Activ8]

I wasn't very clear on this. What I was trying to say is that epoxy
formulas designed to cure slowly usually have a lot higher cured
strength and toughness than formulas designed to cure in a few
minutes. Those fast kinds get little tougher than dried chewing gum.

yeah, it is pretty lame, isn't it?

I was not so much thinking of speeding up or slowing down the cure
time of any particular kind. Epoxy with a 20 to 60 minute working
time just seems to harden to a much stronger bond, when it finally
gets around to curing. It is still stronger if its cure is
accelerated by a bit of heat. It is the chemistry that is better.

Ok, I guess the really fast epoxies don't use accelerators. Just
think of what I said as additional info.

But like I said, using a little less or more hardner is ok. Like you
said, it's the chemistry. I think, however, that using too much more
or to much less would foul up the cross-linking.

One example I can think of is plasters and other gypsum products. I
think borax is a retarder - reduces strength. One easy accelerator
is to use slurry water - ground gypsum and the water used to keep
the grinder clean. Makes the stone set up so fast you gotta work
fast. Cushing strength? Crap. It doesn't change the chemistry, but
it accelerates the set up to the point that the crystals don't align
optimally, or whatever.