low volume custom pcb enclosure

[stern0m1]

Im looking for leads for affordable low volume  custom plastic pcb enclosures.  I designed the CAD files. Im looking for some way to have my CAD files manufactured.

Injection molding- seems to start at least $1500.
3d printing- seems for my files something like $10 a piece.


Thanks

[ChemE]

I would think the cheapest way to make this happen would be if Felix could justify the mold costs and have a bunch injection molded.  Not sure if he could reasonably expect to sell a few thousand of them to help defray the mold costs.  The molds are so expensive b/c they cannot be made from Al.  They have to be milled out of high-strength steel to handle the pressures exerted by the molten plastic (not that I've explored this before ).

[Felix]

There's the possibility to make silicone molds for less than huge volume production. They can still produce a very nice case at a fraction of the cost of metal molds.
I am making a case right now for another project, silicone molds, and I'm very eager to see the outcome. The design and tweaks and proto prints/adjustments has been running for a few months now, not exactly very fast but it's a somewhat complex case. I will talk about it and details when/if it will be ready

Anyway for amateur level casing, you can use makercase to make some very cheap and easy lasercut-able cases. I used that for all cases I offer with MotionMote, SonarMote etc. They look very decent and from a little distance they look quite professional.

I have several Moteino type/sizes now so which one would it make sense to make a case for?
If I make one for the small Moteino then people will ask for a bigger one for MoteinoMEGA. If I make one that accomodates both then they will complain it's unnecessarily big for use with regular Moteinos.... the endless debates of usability vs manufacturing cost
Then I'd have to sell thousands before I get my upfront money back, not really a model I like to run by

[ChemE]

I would think if you only offered one size (which I would) it would be for a 2xAAA battery shaped mote.  I'm beginning to go down the road of making my own in case you don't ever get to it and my PTH AAA battery clips will arrive later this week.  To my mind, for all the different projects we use this platform for, 2xAAA LiFeS2 batteries clipped to a board with a 328p and RFM69CW/HCW and if it can fit a PCB antenna in a waterproof/dustproof little enclosure would be incredible.

[WhiteHare]

3d printing- seems for my files something like $10 a piece.

I'm assuming you mean $10 each in material cost alone (i.e. the cost of the plastic filament)?  I hadn't realized that 3D printing a custom  enclosure would be so expensive!  What are the dimensions of your enclosure for which that is the estimated cost?

[executivul]

ABS is about 20USD/kg, PETG is also a good option.
A cheap 3D printer (200mm/200mm/200mm~8"/8"/8") is 200-300USD.
If you have the time to fiddle with 3D printing yourself it's a great opportunity 

I have one and I'm printing the panels for custom lightswitches in my home.

[WhiteHare]

Thanks for sharing the photos, executivul.

So, about how much does one of those enclosures (an empty one) weigh?

[executivul]

If you have the 3d model of the box you can use a free slicer (Slic3r, Cura) to slice it for 3d printing setting a 0.4mm nozzle and 0.2mm layer height (most used ones) or 0.8mm/0.4mm LH, and see how much material is used and the time it takes to print one. The machine draws about 200-250W/h so you can estimate energy bill too.
Just watch a few yt videos about slicing for 3d printing

[WhiteHare]

That's a useful answer.  Thanks!

By the way, one quick thing I've always wondered about that relates to 3D printing cost of material: suppose the unused filament at the end of a roll is not enough to complete an entire 3D print.  In actual practice, what is typically done with it?  Is it thrown in the trash as leftover waste or instead somehow spliced to the start of a new roll so that all of it gets used? 

The OP never gave his dimensions but says printing an enclosure costs $10.  So, ballpark, that would be 500g of material.  By inference, that sounds like a pretty large enclosure.... much larger than what would be needed for most Moteino motes.

[executivul]

ABS density is 1.04g/cm^3, so 500g ~= 500cm^3, at 3mm wall thickness it means 1667cm^2 that is a cube with 165mm(6.5") size.
Maybe the OP said $10 including manufacturing costs, 3D Printers tend to be slow and printing is a hit or miss job, so prices for the customers are a bit high, that's why I advised him to get his own printer if he has the time, skill and nerves 

LE. the math is for 3mm thick 100% infill walls, if you use 3 inner/outer layers and 30% infill you could double the size.

[WhiteHare]

As a point of comparison, I recently purchased 5 of these black plastic project boxes for $4.32:  http://www.ebay.com/itm/201633631645?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Ignoring shipping costs, that makes them about $0.87 each.  For many, if not most Moteino sensor projects, they're plenty big enough.

Each box weighs in at 34g on my kitchen scale.  That's about 4 grams more than the advertised 30 grams, so for the sake of being conservative, let's go with 34 grams.

Unfortunately, I'm missing the key information as to whether or not it's ABS plastic, but if it is ABS, then based on weight alone the printing cost (assuming solid walls) would be $20*0.034= $0.68.

Anyhow, it seems like roughly a wash either way.  With 3D printing, though, at least you can get exactly what you want.

[executivul]

With 3D printing, though, at least you can get exactly what you want.

That's the key factor, you get standoffs, customs cutouts, vents, whatever you want and you get it in a few hours not weeks 

[TomWS]

The cost of plastic in 3D printing is virtually nil, the real costs are the design and print times.  If you do your own design, you control those costs, and, if you have your own printer, the printing runs unattended so, if you have an UPS to keep it running during brief power outages and can leave it running for as long as it takes, you can make literally anything.

The attached photos are of a prototype of a drip valve, printed in PLA, at a total plastic cost of $0.33 (approximately 1.2Meters from a 1KG spool that cost $25 and contains 880 Meters of filament).  I don't recall exactly how long it took to print, but I believe all 5 parts took a total of about 3 hours and can be printed in a single run (actually there is room on my printer bed for probably 4-6 sets of parts).

You don't need ABS unless the piece is going to be exposed to a harsh environment (temperature or high humidity) and I haven't used ABS yet since I don't have an effective means to filter/vent the styrene fumes - yet.  I can print PLA in my office unvented...

Tom

[Felix]

@TomWS - nice print, looks really good compared to other 3d prints I see online.
What printer do you have? Any customizations?
What's the project you are showing, looks very interesting (if you can share)?

[TomWS]

@TomWS - nice print, looks really good compared to other 3d prints I see online.
What printer do you have? Any customizations?

Prusa I3 MK2, no mods to the HW or SW, minor tweaks to the filament and print parameters to improve adhesion to the base.  These were done on the mid-quality setting (0.2mm layer thickness).

What's the project you are showing, looks very interesting (if you can share)?

This is an inline drip valve that's intended to irrigate small regions like window boxes, herb gardens, etc.  It requires an upstream valve and pressure reducer if the source is residential water supply, but should work well for rain barrel setups.  Naturally it's battery operated with an expected battery life of 2 or more years.  I'll share the project when I've deployed it.  I have the second version of the prototype running, but waiting for warmer temps to start setting them out.  I 'may' include the 3D print files, not sure yet.

Just for fun, I'm attaching another project I just finished.  This is an antenna rotator for an HD TV antenna mounted in our attic.  It uses ESP8266 so probably won't post here.  It has 4 different printed parts (2 PCB mounts, drive gear, antenna post gear, and, to support the world famous LowPowerLabs SOT-23 breakout board, the 3 axis mag sensor holder that positions the center of the mag sensor directly over the drive gear).

Tom

[joelucid]

I can only warn everybody here to exercise extreme restraint as you look at these pictures. You might otherwise quickly end up purchasing a Prusa i3 even if you never intended to own a printer. That's what happened to me anyway ... 

[perky]

Getting a 3D printer is on my list of things to do at some point, any feedback on them is very useful to me -)
Mark.

[Felix]

Getting a 3D printer is on my list of things to do at some point, any feedback on them is very useful to me -)
Mark.

Me too, thats why i asked!! 
Just looked at their page, the prusa i3 mkii looks great, and has good score on 3dhubs, not sure how important it is to have dual extruder (it doesn't, maybe a nice to have for printing toys?).

[ChemE]

Damn you all, now I really want one too!

[Felix]

Sorry this is turning into a Prusa i3 mkii review
But never mind about multi extruder, looks like prusa i3 is light years ahead of multi extruder printers. Check out 4 color printing with 1 extruder!!!
Now I really want one 

[perky]

Nice! This almosts warrants a separate enclosures and 3D printing section on the forum, I think quite a few will be interested in this stuff.
Mark.

[ChemE]

Nice! This almosts warrants a separate enclosures and 3D printing section on the forum, I think quite a few will be interested in this stuff.
Mark.

This is a very good idea actually.  I second the motion.

[Felix]

This is a very good idea actually.  I second the motion.

Ok, not against that but how should this be called?
There is the Projects forum already, an inclusive bucket for all Moteino and generic/3rd party based projects that anyone can share.

[executivul]

Take care guys before taking a leap of faith and ordering a 3D Printer
Give your self a few weeks to watch YT videos and reviews, decide on a model, shall it be ready to print or a kit, a generic Chinese kit is cheaper but you have to print upgrades as soon as you buy it, some ready to print models go in the 3 zero price range and spare parts cost an arm and a leg.
I own a Prusa I3 clone which I hate, now I want a larger print size, Tevo BlackWidow is a good option with larger print size but still has the moving bed with it's inherent problems and limitations, but for $500... Or you could go for a Delta, but get a good one with linear rails. Or you could wait a month or two until the Chinese start mass producing corexy designs

After the printer you'll want a laser cutter, so prepare to get a K40, an air-assist head, a small compressor, a Smoothieboard and upgraded exhaust fan. To cut the brushed steel looking plastics and acrylic cases and windows...

The last tool of course will be the CNC, a 3040 or 6040 but the one with ball screws, doesn't matter it's for old parallel port, you can rig up a grbl controller in no time using any 328p and start milling your own pcbs, for low production it beats up etching, mill and drill at the same time, from design to produced board in less than 30 minutes, great, but you'll need some vacuum cleaner to take care of the carcinogenic fiberglass dust, and that vacuum cleaner should vent outside, but it will get clogged with dust, so you need a cyclone dust separator and the Oneida Dust Deputy is too damn hard and expensive to get in your part of the world and you decide to print one instead BUT the damn printer is too small and you're back to square one.

PS. any resemblance to my life story is purely intentional 

LE. made a few days ago a serial gateway for BananaPi: ProMini+Level shifter+3.3V reg module+RFM69HW

[TomWS]

Fascinating!  @executivul, I went in the exact opposite direction!  Being a woodworker geek, I built a CNC router which worked well, but didn't have the resolution I needed for gears and hubs and such. And it wouldn't do aluminum

I then went to an 'inexpensive' laser cutter/engraver and spent another 75% and a lot of time getting it 'useful' and have actually made some useful, attractive, and interesting things.

Finally, this past holiday, I was somehow drawn into 3D printing and 'gave' a 3D printer to my family.  All were sort of, uh, interested/grateful, until I showed them some stuff we could do with it.  So far, I and two daughters are coming up with a host of things to make.  My wife is surprised by what we've done, and that, alone, is a major accomplishment! 

Net: I've done interesting things with all three of the devices you've mentioned, but I have never felt so 'liberated' as I feel with a 3D printer.  Literally, if I can envision it, I can now make it.  Just the other day, as I was struggling to try to decide what torque a motor I needed to operate at to control a shutoff valve, I realized that I didn't have an adapter that allowed my small torque wrench to operate the large nut on the valve.  Easy peasy as we say in 3D printing world - I simply printed one and, an hour later (including design, conversion, print time) I had my adapter and measured the torque.  It would have taken me that long to drive to the nearest store that wouldn't have even had the adapter I needed!

Call me a zealot, I'll 'fess up.  But I'm astounded at how much I've used this little device...  But I will also admit to being the 'drug pusher' that Joe has fallen prey to...

Tom
PS: to Felix, a 'fabrication' kind of category might be a good thing!

[ChemE]

Ok, not against that but how should this be called?
There is the Projects forum already, an inclusive bucket for all Moteino and generic/3rd party based projects that anyone can share.

The projects subforum has so many stickys now that things already get buried there.

PS: to Felix, a 'fabrication' kind of category might be a good thing!

I think 'Fabrication' or '3D Printing' would be a good name.  Seems like this would be different from projects in that what will live here is more like code than showing off a project since anyone could download a design/gcode and make one on their printer.

[ChemE]

...but you'll need some vacuum cleaner to take care of the carcinogenic fiberglass dust, and that vacuum cleaner should vent outside, but it will get clogged with dust, so you need a cyclone dust separator and the Oneida Dust Deputy is too damn hard and expensive to get in your part of the world and you decide to print one instead BUT the damn printer is too small and you're back to square one.

http://www.jpthien.com/cy.htm

Try making a Thien separator and run it on a shop vac.  A well made one will prevent all but perhaps 1 cup of fine "flour" from getting to a shop vac even when sucking up 55 gallons of coarse/fine chips and dust.

[executivul]

I guess a subforum "Additional tools" with main topics: printers, cnc, lasers, soldering, measuring would be nice.

@ChemE endless battle of cyclon vs thien. Cyclon is easier to buy or print. Thien is easier to build in a large wood shop. I live in a flat unfortunately. Besides the visible fine dust (flour) they miss ALL invisible (deadly if <3um) dust. So a single air circuit vacuum is better than double circuit which almost all shop vacs have and cross contaminate. And it must vent outside! I don't trust hepa.

[ChemE]

@ChemE endless battle of cyclon vs thien. Cyclon is easier to buy or print. Thien is easier to build in a large wood shop. I live in a flat unfortunately. Besides the visible fine dust (flour) they miss ALL invisible (deadly if <3um) dust. So a single air circuit vacuum is better than double circuit which almost all shop vacs have and cross contaminate. And it must vent outside! I don't trust hepa.

Decently off topic now but I went with this upgrade for my dust collector: https://wynnenv.com/products-page/product-category/35a274nano-cartridge-kit/

99.999% of dust trapped at 0.5um.

[WhiteHare]

Not to mention that if you want to collect it all at the source you need a humongous vacuum motor (probably at least 2HP), and to go with it you probably need a new 220v dedicated electric circuit.  Before you know it that modest little CNC has a gargantuan contraption to go with it. 

Still, I agree the allure is pretty strong.  If only it could produce the tiny traces needed for SMD parts....

[executivul]

If only it could produce the tiny traces needed for SMD parts....

0.1-0.15 mm wide isolation, so you can go for 0.2mm wide pad/track with 0.2mm in between, that's at least what I tested with cheap Chinese 0.1mm 10deg v-tips.

I use a Philips 1600W (2hp?) home cyclonic vacuum cleaner, with an extended intake hose, with the exhaust hepa removed and instead output is connected to a tube venting outside(3d printed part again). All it's inside a thick plastic bag, no leaking inside, exhaust is 65mm, intake is 40mm, all negative pressure. I hand vacuum following the mill, using my 60000rpm spindle the circuit you saw above took about 2 and a half minutes to isolate and 2 minutes to drill.

[TomWS]

On the topic of custom PCB enclosure, here's one I just completed for an RF Switch Control point.  This enclosure was designed to handle a generic 60x80mm perf board that I bought in bulk a while ago.  It's a handy size and I can take the basic design and add whatever mods I need for connector access, etc to a one-off design and print.  As you can see in the first picture, this took 6 hours 45 minutes to print, but I started it when I left my office last night.  This morning I've got a completed enclosure.

Material cost for this enclosure?  28.5 meters of 1.75mm PLA = approximately $0.81.

Tom
EDIT: AND it's in Carolina Blue! 

[perky]

Oh you're only going to make me buy one! 

[joelucid]

Do I see an espgw in there, Tom? 

[TomWS]

Do I see an espgw in there, Tom? 

You do see a Node MCU ESP8266 and an old Moteino I salvaged from my  breadboarding collection.  The Moteino is just being used as an OOK transmitter, the control point controls those nifty 433MHz on/off switches that I bought a year and a half ago 

Having said that, I'll probably use the same enclosure to house my espgw (once I have one running)...  with minor mods to change mounting bosses.

Tom

[msjfb]

A note of caution! The world of 3D printing can be highly addictive !!!

After about a year of thinking about it I purchased a printer (kit) late fall, with the specific intention of building my own enclosures.
I figured I could do this in a couple of weeks...  I was achieving very good test prints, but...

I quickly got caught up in all the aspects of this technology and as a retired EE with a career in software I just had to understand all that was going on...
So first came a number of add-ons: high performance heatbed, Octoprint on Raspberry Pi with web cam for remote monitoring, control and stand-alone printing, etc.

Now that I was happy with the 'physical' stuff, I started looking into the software: lots of time getting to understand Marlin (Open source, kind of a mess) and lots and lots of trials with every 'slicer' program out there.

Then came testing and choosing printing materials: there are literally hundreds of different materials available. For enclosures you are looking for accurate dimensions, and UV resistant if it is to be installed outside.

And finally I realized that I needed to master a good CAD software program if I really wanted to do my own stuff: enter the FUSION 360 (Autodesk) learning curve !

3 months into this and a major backlog of my Moteino projects, I am just about ready to start on my enclosures.

My objective is to model all the possible parts that can go into an enclosure: moteino, battery + holder, power shield, Mighty boost,etc.
This way I can play around positioning the parts, and all standoffs, cutoffs, etc. will be done automatically.

Have a look at the www.thingiverse.com site for an idea of what people are building with these low cost 3D printers.

François

[ChemE]

Sounds somewhat familiar.  When you're an engineer all the world is your oyster!

[TomWS]

Sounds somewhat familiar.  When you're an engineer all the world is your oyster!

I would say "When you're an Engineer, you are cursed to fall further behind with every new idea..."

The good news is, now that I'm retired, 'behind' doesn't mean much anymore, and the 'behinder' I get! 

Tom

[executivul]

@TomWS now get a Volcano with 0.8mm nozzle and get those times halved

[TomWS]

@TomWS now get a Volcano with 0.8mm nozzle and get those times halved

Gee, just when I was thinking of getting a 0.25 nozzle so I could make finer gears...  Hmmmm, SO many choices.

'Volcano' huh?  It boggles the mind.

Thanks,
Tom

[Felix]

I like what prusa did with their multi color printing option, brilliant.
I wonder if the same concept could be applied to multi nozzle, I was thinking about multi-head microscopes that just spin around conveniently. Perhaps some inner bulk printing could be done with large nozzles than the fine detail on the outside with the small ones. I guess each layer would have to be printed in both so the structure grows uniformly. It's probably just a crazy idea.

[ChemE]

I like what prusa did with their multi color printing option, brilliant.
I wonder if the same concept could be applied to multi nozzle, I was thinking about multi-head microscopes that just spin around conveniently. Perhaps some inner bulk printing could be done with large nozzles than the fine detail on the outside with the small ones. I guess each layer would have to be printed in both so the structure grows uniformly. It's probably just a crazy idea.

Makes sense to me.  Multi-tool CNC routers have spindles that operate the exact same way too.

[WhiteHare]

@TomWS now get a Volcano with 0.8mm nozzle and get those times halved

Was it the Volcano you used to print this front panel?

It really does appear that it would be blazing fast, and the printing artifacts have an almost artistic look to them.   

[executivul]

Yes. I exclusively use 0.8mm Volcano on my Sunhokey (Prusa I3 clone). I need the speed, accuracy is enough for what I need. All these panels will receive laser cut faces looking like brushed steel, the printed parts shall be hidden to the naked eye

LE. unfortunately my crappy K40 laser had some issues and I've started redesigning the gantry from scratch which is taking A LOT more time than I've expected 
Piece of advice: always get the largest machine you can afford, especially if you're going to buy cheap Chinese sh*t! The machine will not be accurate across all the working surface, so if you try to push the limits you'll be in trouble! If your working surface is twice the size you need you can remain well inside the good accuracy zone. I'm still sorry I didn't fork out for a good 30cm/50cm laser which came with air assist, linear rails, good extractor from the box. I've added all this myself and now with the new gantry I should get the same results for 1/2 of the price but the time taken to get there is priceless...
 

[stern0m1]

There's the possibility to make silicone molds for less than huge volume production. They can still produce a very nice case at a fraction of the cost of metal molds.
I am making a case right now for another project, silicone molds, and I'm very eager to see the outcome. The design and tweaks and proto prints/adjustments has been running for a few months now, not exactly very fast but it's a somewhat complex case. I will talk about it and details when/if it will be ready

Anyway for amateur level casing, you can use makercase to make some very cheap and easy lasercut-able cases. I used that for all cases I offer with MotionMote, SonarMote etc. They look very decent and from a little distance they look quite professional.

I have several Moteino type/sizes now so which one would it make sense to make a case for?
If I make one for the small Moteino then people will ask for a bigger one for MoteinoMEGA. If I make one that accomodates both then they will complain it's unnecessarily big for use with regular Moteinos.... the endless debates of usability vs manufacturing cost
Then I'd have to sell thousands before I get my upfront money back, not really a model I like to run by

Is this method commercially available? I cant find any info online. Thanks

[stern0m1]

If anyone is interested, I use the monoprice miniselect 3d printer. It costs $200. Im very happy with it.

I think the its strong points, besides the price, is that the community fully understands how it works, information is available on how to fix it, modify it. see http://mpselectmini.com/.
As mentioned before, 3d printing currently is more of an art. The technology has progressed, but not to the extent like using a laser printer. You need to be willing to tinker. You need to be a DIYer.

[Felix]

Is this method commercially available? I cant find any info online. Thanks

I think this is available at contractors who prototype and fabricate 3d models, but not really as an off-the-shelf product or service. From my limited knowledge each type of mold has certain limitations and your 3d model has to be done right so the mold can produced the desired output. If you're interested I can get someone in contact with you who will probably have more answers.

[stern0m1]

I am interested in learning more about this method. If you can get me in touch with someone that would be great. Thanks.

[Felix]

You should get an email from that person sometime soon.

[stern0m1]

Got it. Thanks.