Very cheap molded enclosures?

[Carl Ijames]

This is the third or fourth post of Rich's that used "punch" for

"drill". It took that long to seep through my poor slow brain:

If you start with an enclosure with the right materials and design,
and plan on putting a stick-on top panel then you can have the holes
punched in one operation with a custom die. A set of dies with round
holes should be relatively cheap. If you punch toward the inside of
the box the holes can even be a bit ratty as long as they're clean on
top (or at least ugly but concave) where you need to stick on the
label. You could even have a few oblong holes (like for the ribbon
cable from the panel) at a bit higher cost for the die.

I had a die made by http://www.acesteelruledies.com/ for work and was
very pleased. Roughly 3"x11" rectangle with 16 holes about 1" in
diameter, laser cut into hardwood with steel rules for about $100, in
about a week after I emailed them an Autocad drawing. This is to punch
foam rubber but it should work on thin acrylic or petg or ABS or ...
Call them; they were very helpful on the phone. Oh, if you send them
the material they will even do the punching for you if you don't happen
to have a press . That was $2.50 a piece for 50 pieces, and I'm sure
that would get better with more volume but that was all the gaskets I
needed for now and I was really tired of hand punching them .

Punching is always cheaper than drilling, even one hole at a time.

 

[Joerg]

Hello Carl,

I had a die made by http://www.acesteelruledies.com/ for work and was
very pleased. Roughly 3"x11" rectangle with 16 holes about 1" in
diameter, laser cut into hardwood with steel rules for about $100, in
about a week after I emailed them an Autocad drawing. This is to punch
foam rubber but it should work on thin acrylic or petg or ABS or ...
Call them; they were very helpful on the phone. Oh, if you send them
the material they will even do the punching for you if you don't happen
to have a press . That was $2.50 a piece for 50 pieces, and I'm sure
that would get better with more volume but that was all the gaskets I
needed for now and I was really tired of hand punching them .

Punching is always cheaper than drilling, even one hole at a time.

Thanks! That seem to be a good place to get started. I like their flame
logo on the home page. $2.50 a pop would blow the budget, even half that
would. But maybe this can be done somewhere on an automated press,
especially since precision isn't such a big concern for this job.

But I wouldn't discard drilling. Considering how cheap that is on
circuit boards there has got to be a way. Worst case one could have PCBs
made with no traces on there, just the holes, if that is cheaper.

Regards, Joerg

 

[[email protected]]

Hello Ken,

I thought about clear boxes but those aren't very cheap.


No 9V blocks, they are a pain. Ripped off snap contacts and all that, it
would create too many 'warranty' returns. Having to open the box
wouldn't be an issue though. I think people have gotten used to that
since this is what you have to do to change the batteries on many
outdoor remote temperature sensors. My non-EE wife has no problems doing
that.


Super brights blow the budget and would slurp up batteries too fast.
Spread would be a problem but that could be mitigated by sliding tubing
pieces over them. Drip irrigation tube snippets or the like.

Regards, Joerg

Another unsure idea: one punched square hole, with all your (square)
LEDs huddled up together in the middle of the panel, and legends
written using pointer lines to enable 2 columns of text to point to
your LED cluster. Or you could even divide your panel like a pie chart,
with white and light grey segments, to make LED ID visually clearer.

A block of 9 square LEDs with eg a power one as the middle one. 2 such
clusters if need be.

Less likely, but might work wth a few cases, guillotine the edge of the
front panel off, and that missing edge strip is where your line of
sqaure LEDs sits.


NT

 

[Carl Ijames]

Thanks! That seem to be a good place to get started. I like their

flame logo on the home page. $2.50 a pop would blow the budget, even
half that would. But maybe this can be done somewhere on an automated
press, especially since precision isn't such a big concern for this
job.

Your welcome. I know $2.50 is too much for you, but that was for 50
pieces and punching isn't their real business, diemaking is. I let them
do the punching for the convenience. A local rubber company does
punching for us on other stuff for about 1/2 that, on much larger
gaskets, so I'm sure you could get it well under $1 with much less than
your 10K per year (make the die big enough to do 4-8 lids at a time, for
starters, whatever the press will stand). Now that I have the die I'm
going to our local company for the next run. We also do some boxes with
LED displays where the LEDs are on a PC board behind a hole in a metal
box, with a stick on label on the outside. The label has solid color
regions, all the lettering, and clear windows over the LEDs. We punch
holes for switches to stick through. Sounds great, but, in 100's for a
4"x8" label they are several dollars each which kills your budget. We
are okay with it because we want the rugged box and the quantity is so
low the cost would be about the same no matter how we made them.
Anyway, just a long-winded way of saying that a label is also going to
be expensive. Maybe punch a panel and then silk-screen it? That is
pretty cheap to do, and easy to set up so long as you only want one
color. I've been reading up on building a vacuum forming machine for
fun - really doesn't seem too hard and it would be neat to form your own
boxes . Form the box, punch the lid from sheet, silkscreen it, and
glue it on since you don't plan on repairs anyway. Hire some cheap
labor and do them yourself until the volume takes off and you can afford
to go offshore. Then start offering semi-custom boxes for other people
who have the same needs as you do now .

 

[Clifford Heath]

Joerg,

I've recently used these electrical junction boxes for an electronic
project:

http://alfred.clipsal.com/scripts/apeweb.dll?PrintCatNo&CatNo=240/16/1D

This was for devices that will be left out in the rain and chained
around street posts for days at a time, and must survive casual
attempts at deliberate destruction, such as being jumped on by
neighbourhood kids etc.

I fitted a short PVC pipe in the input port (as a handle to attach
the chain to) and drilled the lid for an IButton socket and LED.
They come with a variety of port options. The inside is just the
right size to make a 2xAAA battery holder a tight fit in the base,
leaving plenty of height for a circuit board.

Price was $AU4 each for 30, pretty good, about half the price of
the cheapest low-volume pricing on Jiffy boxes and the like. You'd
get a much better price in volume from Clipsal. Their molds are all
made by CNC I suspect, at least they provide full 3D CAD drawings,
so for a sufficient volume could perhaps turn a special, such as a
version with no ports.

Clifford Heath.

 

[[email protected]]

Hello Rich,


Yes, else you get too many returns or 'warranty' claims. It's not just
people. Then there is the new puppy dog that has the unquenchable urge
to sink its teeth into anything that resembles a toy.



That would be ok if board and Lexan panel would snap into place without
screws and the back of the box had a battery compartment with a sliding
lid. When I looked at the standard program of the major box
manufacturers I could not find anything like that. Just the usual (and
expensive) project boxes.

Units like the one I am thinking about don't need to be repairable. So
the snap-in action may well be irreversible.

Regards, Joerg

A thought slightly off track from boxes: are you using the cheaper
brown phenolic paper pcbs? Theyre.... cheaper.


Not really on topic here... I've always vaguely wondered about going to
extremes with pcb cost, and using a piece of thick card with copper
foil stuck on it. Varnish the card, etch the foil, and mount in the
case by applying glue to the PCB and pressing pcb into place. The case
provides mechanical rigidity. However thats not a serious suggestion
for your project, any more than would be using a tortilla, just a
muse-ment. I'd just like to see someone some day do some work on cheap
PCB materials and come up with something far less costly than FR4 or
phenolic thats good enough for mass production.


NT

 

[Ken Smith]

The parallax made it hard to tell which was which, so I got a couple of
soda straws and cut them up to make a little sleeve for each LED.

I've used a about 6" long rod to bring the light from a LED on the PCB out
to where it can be seen. It works quite well. The rod just has to line
up with the LED when the unit is assembled. It doesn't bring ESD into the
PCB.

 

[Rich Grise]

Hello Jeff,


Yes, it is. Coming up with simple and cheap tricks to solve a problem is
what makes engineers great. I used drip irrigation hose for this job
before. But come to think of it, that stuff cost about $4 for 100ft at
the hardware store. I bet Rich's straw solution costs less per 100ft.
Dang, I have wasted money...

Probably not as much as you think - they still all have to be cut to
length by somebody, or by an automatic machine that somebody would
have to invent and set up.

And black irrigation tube would be a lot more opaque than a straw.

<anecdote>
One of my buddies in high school had an old player piano at his house,
and all of the little rubber tubes were cracked and falling apart, so
we replaced them all with plastic aquarium hose - it was also fun
cutting 88x 1.5" tubes to reconnect the read head to the manifold.

It was also fascinating working with the centering mech - there was a
cat whisker at either edge, connected to proportional vacuum valves,
that turned a motor that physically moved the whole carriage (as needed)
to keep the roll centered on the head.
</anecdote>

Cheers!
Rich

 

[Rich The Newsgroup Wacko]

Hello Carl,


Thanks! That seem to be a good place to get started. I like their flame
logo on the home page. $2.50 a pop would blow the budget, even half that
would. But maybe this can be done somewhere on an automated press,
especially since precision isn't such a big concern for this job.

But I wouldn't discard drilling. Considering how cheap that is on
circuit boards there has got to be a way. Worst case one could have PCBs
made with no traces on there, just the holes, if that is cheaper.

Maybe you should check with Guy Macon's group. ;-P
--
Cheers!
Rich
------
"There was a young man from Dundee
Who buggered an ape in a tree.
The results were quite horrid:
All ass and no forehead,
Three balls and a purple goatee."

 

[Joerg]

Hello NT,

Another unsure idea: one punched square hole, with all your (square)
LEDs huddled up together in the middle of the panel, and legends
written using pointer lines to enable 2 columns of text to point to
your LED cluster. Or you could even divide your panel like a pie chart,
with white and light grey segments, to make LED ID visually clearer.

A block of 9 square LEDs with eg a power one as the middle one. 2 such
clusters if need be.

This would make identification of the lit LED tricky from a distance and
for people with slightly impaired vision (folks won't always have their
glasses on while using it).

Less likely, but might work wth a few cases, guillotine the edge of the
front panel off, and that missing edge strip is where your line of
sqaure LEDs sits.

I did think about a larger version of a bar graph row. But then I
realized that these kind of fell from grace in the audio gear world and
became expensive. There it's mostly LCD now. I could go LCD but that is
totally cost prohibitive unless you are doing a million units a year.

Regards, Joerg

 

[Joerg]

Hello Rich,

<anecdote>
One of my buddies in high school had an old player piano at his house,
and all of the little rubber tubes were cracked and falling apart, so
we replaced them all with plastic aquarium hose - it was also fun
cutting 88x 1.5" tubes to reconnect the read head to the manifold.

It was also fascinating working with the centering mech - there was a
cat whisker at either edge, connected to proportional vacuum valves,
that turned a motor that physically moved the whole carriage (as needed)
to keep the roll centered on the head.
</anecdote>

Ah, the old days. I remember repairing telex machines in the 70s. These
were rugged but when something broke on an old Lorenz we often could not
get a spare without paying some exorbitant price. So we had to re-invent
things. Lots of fun.

Then we bought a very old and neglected piano. Pre-Edison, so it had
'candle illumination'. Everything was wood, leather and cloth. That was
also a fun restoration project. Afterwards it truly sounded like one of
those uprights in the old saloons.

Regards, Joerg

 

[Joerg]

Hello Carl,

... We also do some boxes with
LED displays where the LEDs are on a PC board behind a hole in a metal
box, with a stick on label on the outside. The label has solid color
regions, all the lettering, and clear windows over the LEDs. We punch
holes for switches to stick through. Sounds great, but, in 100's for a
4"x8" label they are several dollars each which kills your budget. ...

Sounds familiar. I priced out translucent labels 15 years ago and still
remember that sticker shock. We then used the silk screening route.

... Then start offering semi-custom boxes for other people
who have the same needs as you do now .

I strongly believe there is a market and that many products never come
to fruition because the enclosure costs are breaking the deal. So there
should be a healthy market. I really don't know why box mfgs don't offer
boxes with some standard size holes in them for switches, potmeters,
LEDs. Then they could offer plugs for those holes that you don't want
and it should be the often cited win-win situation.

Regards, Joerg

 

[Joerg]

Hello Clifford,

http://alfred.clipsal.com/scripts/apeweb.dll?PrintCatNo&CatNo=240/16/1D

Those are a bit too utilitarian for this project.

I have scoured our hardware stores for anything that might be useful
since electrical boxes are very cheap in the US. At least those blue
plastic ones which you can buy in bulk as a 'contractor pack'. But there
wasn't anything suitable. Another option would be to look in the toy
department of stores like Walmart for something suitable, still have to
do that.

Regards, Joerg

 

[Joerg]

Hello NT,

A thought slightly off track from boxes: are you using the cheaper
brown phenolic paper pcbs? Theyre.... cheaper.

Probably will but PCB costs are quite low. It definitely must be no more
than two-layer, maybe even single layer.

Not really on topic here... I've always vaguely wondered about going to
extremes with pcb cost, and using a piece of thick card with copper
foil stuck on it. Varnish the card, etch the foil, and mount in the
case by applying glue to the PCB and pressing pcb into place. The case
provides mechanical rigidity. However thats not a serious suggestion
for your project, any more than would be using a tortilla, just a
muse-ment. I'd just like to see someone some day do some work on cheap
PCB materials and come up with something far less costly than FR4 or
phenolic thats good enough for mass production.

Usually consumer goods show the way here. There they attempt to squeeze
out every last penny. I was a bit surprised when I opened a modern TV
set and found that the main board was a rather thin phenolic. Same for
some smaller hand-held gear.

Regards, Joerg

 

[Frank Bemelman]

I strongly believe there is a market and that many products never come
to fruition because the enclosure costs are breaking the deal. So there
should be a healthy market. I really don't know why box mfgs don't offer
boxes with some standard size holes in them for switches, potmeters,
LEDs. Then they could offer plugs for those holes that you don't want
and it should be the often cited win-win situation.

That's a daft idea. The holes would always be in the wrong place, or
even worse, slightly misplaced. The plugs would cost 2 cents which you
find too expensive.

Products come to fruition because some companies have the guts to
get on with it.

If you plan to make a 10k-20k/year of these boxes, you should at least
have the guts to buy 1000 cases at *whatever* it costs to see if your
product flies. You are *making* it an impossible project, because you
want to make profit on the first box.

Did you buy a few of those cable testers? No, of course not. You're
not even prepared to buy them, rip them apart, to 'massage' your own
creativity, and to see what happens, to see if it sprouts other ideas.
You only see elephants that crush your boxes, you only see 9V batteries
that you don't like, warranty claims, stickers that were too expensive,
and what have you.

Make 1000 boxes for $15 a piece, and sell them for $5. You can save
a buck if you drill them yourself, using a steel template/rig, if
that makes you sleep better. The $10.000 loss on the first batch
is just an investment.

 

[Joerg]

Hello Frank,

That's a daft idea. The holes would always be in the wrong place, or
even worse, slightly misplaced. The plugs would cost 2 cents which you
find too expensive.

Nope. One can always design a circuit board so they line up with
whatever the enclosure demands. It's probably what we are going to do if
we can strike a deal with a company that uses a suitable enclosure for
something totally different.

Products come to fruition because some companies have the guts to
get on with it.

If you plan to make a 10k-20k/year of these boxes, you should at least
have the guts to buy 1000 cases at *whatever* it costs to see if your
product flies. You are *making* it an impossible project, because you
want to make profit on the first box.

Well, I guess you haven't yet lived through a venture capital start-up
phase. It's rough out here. Neither you nor I make such decisions, the
guys with the Armani suits do.

Did you buy a few of those cable testers? No, of course not. You're
not even prepared to buy them, rip them apart, to 'massage' your own
creativity, and to see what happens, to see if it sprouts other ideas.
You only see elephants that crush your boxes, you only see 9V batteries
that you don't like, warranty claims, stickers that were too expensive,
and what have you.

These tester didn't look too appealing except one. But how do you know I
didn't buy and rip stuff? I have ripped numerous parts that are used by
rock bands and found some that might work. Some less than $10, BTW. Plus
toys, plus...

Make 1000 boxes for $15 a piece, and sell them for $5. You can save
a buck if you drill them yourself, using a steel template/rig, if
that makes you sleep better. The $10.000 loss on the first batch
is just an investment.

Now go and tell that to the VC. There is a whole lot more than those
$10k involved. Right off the bat you can't even test market stuff
without a slew of agency approvals. And, you know, that costs. Since you
live in Europe you should know how bad bureaucratic hurdles for
start-ups can be. WEEE and things like that.

Regards, Joerg

 

[Spehro Pefhany]

That's a daft idea. The holes would always be in the wrong place, or
even worse, slightly misplaced. The plugs would cost 2 cents which you
find too expensive.

Products come to fruition because some companies have the guts to
get on with it.

If you plan to make a 10k-20k/year of these boxes, you should at least
have the guts to buy 1000 cases at *whatever* it costs to see if your
product flies. You are *making* it an impossible project, because you
want to make profit on the first box.

Did you buy a few of those cable testers? No, of course not. You're
not even prepared to buy them, rip them apart, to 'massage' your own
creativity, and to see what happens, to see if it sprouts other ideas.
You only see elephants that crush your boxes, you only see 9V batteries
that you don't like, warranty claims, stickers that were too expensive,
and what have you.

Make 1000 boxes for $15 a piece, and sell them for $5. You can save
a buck if you drill them yourself, using a steel template/rig, if
that makes you sleep better. The $10.000 loss on the first batch
is just an investment.

Good advice, though I think I'd try to make 500 for $7500.

In any case, you don't want a product like that that only sells 500
per year. If it fails in the marketplace, forget it. If it succeeds,
go buy the proper tooling. Consider it "forward pricing" or a
marketing experiment.


Best regards,
Spehro Pefhany

 

[Frank Bemelman]

Hello Frank,


Nope. One can always design a circuit board so they line up with
whatever the enclosure demands. It's probably what we are going to do if
we can strike a deal with a company that uses a suitable enclosure for
something totally different.



Well, I guess you haven't yet lived through a venture capital start-up
phase. It's rough out here. Neither you nor I make such decisions, the
guys with the Armani suits do.

Well, if someone else makes the decisions for you, why post the question?

Now go and tell that to the VC. There is a whole lot more than those
$10k involved. Right off the bat you can't even test market stuff
without a slew of agency approvals. And, you know, that costs. Since you
live in Europe you should know how bad bureaucratic hurdles for
start-ups can be. WEEE and things like that.

Sure, go and tell the VC. Of course there are other costs, but if you
continue to think in obstacles rather than solutions or oppertunities,
this thing is never going to fly. Tell that to the VC as well.

Okay, make 500 then, as Speff suggested. The $7500 materials investment
allows you to *give* them away, instead of selling them for $5. Why
isn't that website up and running, with the product information and
registration form for a free product sample? You have to spend money
to make money, not the other way around. That includes wasting money.

 

[Joerg]

Hello Frank,

Well, if someone else makes the decisions for you, why post the question?

The original question I posted was for a low cost type molded enclosure.
There were some excellent replies and have found a few possibly viable
solutions. Working on it right now.

Sure, go and tell the VC. Of course there are other costs, but if you
continue to think in obstacles rather than solutions or oppertunities,
this thing is never going to fly. Tell that to the VC as well.

With investors you have to be up front not just about the solutions but
also about obstacles. It gives you a lot of credibility. I don't do this
for the first time. Chances are, at the next meeting one or more of them
will be quite smart and ask about such an obstacle. One has got to have
a good answer. Or as the scouts say, be prepared.

Believe it or not but the toughest business questions I fielded during
prior finance group or board room discussions hardly came from people
with an engineering or MBA background. They came from physicians and
attorneys. I kind of liked that and ended up developing the closest
relationships with those who had prodded me with the more difficult
questions.

If we can make it for the target price it will fly. Like other stuff
before that.

Okay, make 500 then, as Speff suggested. The $7500 materials investment
allows you to *give* them away, instead of selling them for $5. Why
isn't that website up and running, with the product information and
registration form for a free product sample? You have to spend money
to make money, not the other way around. That includes wasting money.

Setting up a web site should be done once you can actually ship finished
goods and not before. I know that even some semiconductor companies see
that differently and put products up there that they didn't yet have but
their credibility has taken a nosedive because of that. I think we all
know who those are.

Regards, Joerg

 

[Frank Bemelman]

With investors you have to be up front not just about the solutions but
also about obstacles. It gives you a lot of credibility. I don't do this
for the first time. Chances are, at the next meeting one or more of them
will be quite smart and ask about such an obstacle. One has got to have
a good answer. Or as the scouts say, be prepared.

Yes, but it doesn't make sense here. You say an initial low of 10-20K per
year, that suggests a product life of more than a year, and an increase
in numbers as well. So in the end you will be making perhaps 250K of these
things. So why worry about the costs for a test batch? If it flies, you know
you can have them custom made for a 10th of the price of a hammond box or
whatever. I really don't understand why you want to scrape off a few dimes
for your first run.

Ask the VC what his budget is. And if he has any balls left. Like a good
scout
would ask