Going to need a custom plastic enclosure.

[viper]

Reality has set in that for our commercial project, we will have to source custom plastic enslosures.  I don't really want to go down this road but at least hoping to find a company that does a little of this both in and out of country?  I know the actual molding process is very economical, but if we use say injection molding, those molds are NOT cheap!  Ironically we are actually a manufacturer and could machine the molds, but I also know companies want to be in control of their own molds, which I can understand.  We get people all the time thinking they will just "program our machines, and we will run them".  Ain't happn....lol 

But I do need to poke around with other technologies.  I am not sure if 3D printing is really there yet that can give true value at production volumes, and a pro finish. 

[squadchannel]

JLC3DP's nylon or resin have already evolved to the point where they can be used in products. completely difference the FDM.

but mass production, injection molding is the way to go.

[viper]

I didn't think about a resin mold.  Sort of wondering if that could be an appropriate commercial "step" as we might be able to run that in lower volumes inhouse? 

Are these silicone molds?  It could be a wise step to prove out the design as changes in an injection mold can be dramatic. 

[Coordonnée_chromatique]

 Selective Laser Sintering (SLS) 3D Printing should be appropriate for a small volume because a mold is very expensive and the FDM quality is too low for a commercial product.
https://www.sculpteo.com/en/materiaux/materiaux-jet-fusion/tpu01/

[tszaboo]

If your numbers are not high enough, try vacuum casting.
I don't think it's appropriate to sell 3D printed parts unless it's some sort of vanity project.

[viper]

Agree, I think a pro molded box will be required unless I can find another technology.  So that brings me to a couple questions.  In this R/D stage, we will have to just  find a 'close enough' project box and mod it for our needs.  Are there some preferred vendors you guys like?  I was on polycase but nothing there seems close enough. 

As we get closer to production, does anyone here have any contacts for molded cases?  I sort of need to at least get some prelim price info so I can figure out how to optimize for mfg. 

[squadchannel]

in Japan, Takachi is well known, but US?
https://www.takachi-enclosure.com/

[Psi]

Depending on volumes and other factors you can sometimes just find an existing case on the market that you can 'make work'.
Then buy enough of them so you have plenty of time to source your own case in the future should the case disappear from the market.
If you go on alibaba there are so many enclosure companies with so many different plastic cases they will sell you in any color and with any printing you want on them.
And since the mold already exists they are pretty cheap.

Also consider if it actually needs to be plastic, can it be metal aluminum extrusion cut to a length with front and back panels, as that is much easier.
I get one of my products made that way using a Hammond extrusion which I just get them to cut to the length I want.

For lower volume stuff you can design a case yourself, 3D print it etc, then once you are happy send the design files to PCBWay and get them to make it out of plastic/metal etc..
It wont be as cheap as you can get from proper injection molding places but for lower volumes it can make sense.
PCBway also have an injection molding service but i've never used it. But at least dealing with PCBway is easier than dealing with a china factory that may not speak English that well. They buffer you from those issues.

[edpalmer42]

Before you go too far down the custom route, talk to companies that make enclosures.  They will probably have various cases that have never made it to their catalogs.  Give them your requirements and approximate quantities and see what happens.  If they don't have something available off the shelf, they may be able to easily/quickly/cheaply modify an existing design.

Many years ago a friend was looking for tiny cases.  He couldn't find anything appropriate, but when he talked to a manufacturer he found exactly what he wanted.

[Psi]

Yes talking to them usually helps a lot, which means either you find one that speaks good English or you have someone at your company who speaks their language.
The enclosures you see on listed on alibaba are only a fraction of what's available so explaining what you need is always good.

[CatalinaWOW]

When you do you your discussions you will need to have numbers for volume.  High volume means different things to different people.  Are you talking 1000, 100,000 or a million.  Numbers below that really are 3D print or standard case volume.  The molding guys will have use different technologies and have different up front and unit costs at various points on the volume range.

[viper]

Appreciate all the input and ideas.  Qty is est at 2000/y, then ramping to around 10,000/y by Y3.  the main issue we face is the enclosure is going to be a combination of functions, and will have passages for air flow through it.  In short, it will be far different than just a box for PCBs.  For that reason alone, it is safe to assume we have few options.  1, as a cheap proto, cut/bend/glue whatever to prove it out and try to refine.  2, 3D printing.  This has come a very long way but still no match for injection mold, but I also know we are not pushing too hard on qty either.  Being in a similar biz, I might even entertain the idea of just bring in printers and dealing with the job, but I'd have to really think on that. 

3, I guess is explore full mold options.  I am told some companies will actually let us machine our own mold, in which we can certainly do the machining, but I'd have to get up to speed on a few mold things like gates, runners, draft angles, etc.  In short, not sure the first one would be gold, BUT there is probably value in that struggle. 

4, Some other process I am not thinking of just yet....

[CatalinaWOW]

I only have marginal exposure to all this but will provide comments I have drawn from this exposure.  Quite aside from quality/appearance issues 2000/year is at or above the practical/economic limit for 3D printing.  I would still endorse 3D printing as you develop you design to make prototypes that validate fit, airflow, access and general fit and appearance.  Maybe even for some beta test field units.   Just remember not to design in features that are easy to print but hard to mold.  Aluminum molds are far cheaper than steel and have production lives that can approach your initial annual rate.  Might make sense as a final test of the design, avoiding a huge dollar commitment to something that hasn't had a lot of field testing. 

[Psi]

Are these silicone molds?  It could be a wise step to prove out the design as changes in an injection mold can be dramatic.

Yes, you can get silicone tooling made, it's quite a bit cheaper, but it has a limited lifespan.
A company I used to work for did that to prove out the design before getting the costly steel tooling done.
Depending on type and how complex your design is it can last for 10 to 1000 injection cycles before failing.
If the silicone mold has lots of intricate protrusions they tend to get damaged faster with every release and eventually snap off.

If/when you do get steel tooling done, I recommend getting more than 1 cavity cut into the mold if you can, even if you don't need it.
It's good for higher volume later on, but mainly it gives you a spare in case something happens. It de-risks things a bit, if you need to change the tool you can change only one cavity first and test it before you change the other ones. If you screw up the first change you're still ok, just down to n-1 good cavities from now on. Instead of having to fork out money to weld-up the mold or create a new one.

[Poroit]

Anything suitable from Boxco?

https://www.boxco.eu/index.do?langSite=en#BOXCO/1

[Fire Doger]

Prototyping
1. Full custom: Get clear resin 3D printed cases. You can validate also the airflow with smoke. You may find someone that can do them locally or from china, i think even jlcpcb offers it.
2. Talk to a manufacturer, find something close enough and maybe talk about modifications.

Injection Mold sample
Before going to full metal mold, you can create resin based molds. They have a metal base that fits into injection machine but the cavity is made out of resin material. It's used for evaluating final results such as texture, check the snap hooks of final plastic mixture etc...
It can also be used for small productions. You can probably be ok for a couple thousands per year.
A price range example on case that had 5-6 different parts costed 5k for the first injected molded samples and it looked perfect with textures from standard and everything...
For the final mold the quote was above 50-60k

[kosine]

If you can do your own CNC machining then a blank injection mould base is actually not that expensive these days, just a few thousand depending on size. Most of the cost is in cutting the cavity plates, which you may be able to do in-house. (E.g., https://www.meusburger.com/en-gb/mould-making/mould-bases/f-p-lengthwise-mould-bases but you need to register to get prices.)

Your best bet is to find a very small firm of toolmakers in your area to talk to. Smaller the better, preferably toolmakers and moulders. The bigger companies are not as approachable in my experience, and yes, they'll quote you crazy prices. I've worked with a couple of family-owned firms in the past who haven't had a problem with me doing some of the work, and I'm currently helping a small company who've just acquired a small moulding machine and want to have a go at CNCing their own moulds. So you wouldn't be alone in going that route.

It is a learning curve, but totally doable. They've obviously not usable, but I've CNCed dummy mould plates in wood before to test things out. Take those to a local toolmaker and they'll be quite happy to tell you what you got wrong! If you can then do the job properly in steel (or even aluminium) they wouldn't have much problem finishing it off and running it for you. You'd save a heck of a lot of money and also have the ability to make changes later on.

The reason injection moulds cost so much is because you're paying for the expertise. The mould components themselves are really quite affordable and available off-the-shelf. Find yourself a local grey-beard if you can. They're a dying breed and generally quite keen to pass on their knowledge if you want to play apprentice.

[viper]

Really great replies guys!  I really do think we need to drive down the 3D print road as we perfect the assembly, but we also really need a commercial level product to put in front of people.  You know how this goes that people don't understand how it works, but they comprehend what it 'looks' like.  I also see some features that would really be so much easier in print than IM.  I have to think about releasing the mold.  Certain concave features require a rethink. 

BUT, what I glean here is I probably need to get a trusted and experienced mold guy to examine any mold we might make. I really think there is huge value there for us to learn.  We actually have some extreme talent around here (think SpaceX mfg), but mold work ain't in the stack.....yet. To be honest, I've been itching to just buy IM equipment.  As for Aluminum tooling, I think we could make that call as we do that every day between Al and steel, and which flavor of steel.  I would guess for this project, if we already have our programming and feed/speed for Al, it could be easier to just reproduce replacements.  But to be honest, steel doesn't require a total rethink, just some adjustments, different tools, and more time. 

I honestly think I need to work the beta design as 3D print, BUT plan for it going to IM.  I wish there were better techs but  I am just not in that space.  We only work stuff on a very structural level, so machining and forgings are the norm.  I at least had thoughts about WTF we are not doing complex molds as 3D metal prints?  Is that a thing?  Seems like a good blend of tech.  The expense in IM is the mold work.  Print the mold, send off for IM.....? 

[kosine]

Not up to speed on metal 3D printing, but I know there has been research on using it to produce moulds. If you want to toe-dip, good secondhand moulding machines are typically in the 5-10k range (check ebay), something around 40-60 tonne clamp force is more than adequate to get started. (They're not physically all that big to fit into a modest workshop.)

You only need ancillaries like a mould chiller if you're running for more than a couple of hours, and mould heaters if you're using high temp materials like nylon 66, so they aren't essential to begin with. (Would not suggest 66 until you have some experience, though nylon 6 is not too bad. Both need drying around 100C for a few hours before moulding, as do various other polymers, so a small oven is useful.) A hopper-loader also isn't needed, a kitchen funnel works just fine!

I've used a mould base similar to the attached image for a lot of small projects over the years. This allows very small moulds to be bolted on for short-run stuff, which is a very quick and cheap approach. There's no provision for ejection so you need to design the moulds for easy manual extraction, but it works very well for softer materials. Armed with a machine and mould base you'd be able to experiment with making your own moulds without going overboard with the prep work. Some of my moulds are not much bigger than matchboxes, and to align the two halves I just use M6 caphead bolts as guide pins. (The heads are just under 10mm and nicely fit holes drilled with a W letter drill.)

Aluminium moulds work just as well as steel, and are not that uncommon in industry for relatively short-run projects. They do wear more quickly, but I've seen them used for even for big point-of-sale displays, partly because they're quicker (and cheaper) to machine, and partly because they're much easier to polish up compared to tool-steel. If you need a cosmetic finish that may be a consideration. Hardened steel moulds only become essential if you're looking at very long production runs, I only go there for products that need making in the millions. I've had no problem running aluminium moulds for 50,000 cycles or more.

[TERRA Operative]

I have no experience with it, but PCBWay do 'small' run injection moulding.
I believe they use aluminium moulds or something to keep tooling cheaper.

For 2000 shots a year, it might be viable.

[kosine]

There's also https://www.protolabs.com/en-gb/services/injection-moulding/plastic-injection-moulding-services/

BUT, I'm not sure they actually give you the mould, just the parts. (So I was told a few years ago. May or may not be the case...)

Would be interesting to hear from anyone who has actually used these services.

[tszaboo]

Aluminium moulds work just as well as steel, and are not that uncommon in industry for relatively short-run projects. They do wear more quickly, but I've seen them used for even for big point-of-sale displays, partly because they're quicker (and cheaper) to machine, and partly because they're much easier to polish up compared to tool-steel. If you need a cosmetic finish that may be a consideration. Hardened steel moulds only become essential if you're looking at very long production runs, I only go there for products that need making in the millions. I've had no problem running aluminium moulds for 50,000 cycles or more.

I've worked with aluminium molds. We got a good 30-50k plastic parts out of them without issues.
Nowadays I don't even consider it for a project. Basically alu molds only made sense if the mold was made in the EU, and then it was on par in price as a steel mold in china. Our supplier told us that they don't really make alu molds in China.
Maybe there are some companies that will do that, and then you end up with cheaper molds.
But because all the toys and everything made there, there is just much more moldmaking capacity and knowledge there than here. So even if I have to import the mold, I'll still make it there.

I have no experience with it, but PCBWay do 'small' run injection moulding.
I believe they use aluminium moulds or something to keep tooling cheaper.

For 2000 shots a year, it might be viable.

Yeah, they've ben trying to get some business from me.
I rather work with local companies, even if the part price is higher.

[peter-h]

I've been through this stuff many times over the past 40 years.

In the 1980s we did aluminium tools, in the UK. Good for 10k+ provided you didn't want really clean detail.

Today toolmaking is in China (even if the toolmaker "is in the UK") and they do steel. It is mostly spark eroded anyway; only the core is machined.

But watch the Chinese; they cannot be fully trusted and "misunderstandings" will always go to their advantage e.g. they quote for a tool which will make four cases but actually it will make four case halves Same with mouldings; I had this recently and it was farcical how in a Zoom meeting with a bunch of them I clarified everything explicitly (because I suspected they might do this) and sure enough the quote was for just a half-case. A bit here
https://www.eevblog.com/forum/chat/does-anyone-recognise-this-chinese-moulded-case-yorktron/msg3725599/#msg3725599
I recovered that tool to a UK moulder; now paying 5x more but in the context of the 30 quid product and the better sleep at night it is OK.

I would never allow tooling to remain in China. You will lose it one day and probably several times over the life span of your product. Companies there vanish all the time, with the management having stolen everything that can be stolen. Sometimes they contact you a year later offering to sell it to you

3D printing is too flimsy for most uses. Might be usable with thick sections. But the cost is pretty high and similar to CNC. And if you print metal (which you can) the cost is also similar to CNC. The box here
https://www.eevblog.com/forum/eda/autodesk-fusion-free-version-does-it-do-anything-beyond-viewing/msg6018367/#msg6018367
will cost the bigger part of $100 in CNC, 10 to 100-off. That box, in CNC, will add 200 quid to the cost of the product but it can carry that.

For 1k+ injection moulding will be the way to go. Unless you need metal of course, and zinc-sprayed plastic again works out expensive.

With injection tooling watch out for full tool cost or part tool cost. A trap for the unwary

If you use a designer to design the box make sure you get the original file (e.g. from Fusion, Solidworks etc) and not just STEP or IGES - that's another age old trick to tie you to him.

[Rat_Patrol]

I do a lot of 3D printing, and I think that it can be very viable, even in a final product. You just have to design around 3D printing; use it's strengths, avoid the weaknesses. Fun part is you can get features on the part that are impossible and/or expensive with injection molding.

Material selection is huge. Will any old material work? Carbon fiber filaments (like PETG or even the new PET-CF and PET-GF) along with a very slight "fuzzy skin" and the right build plate and first layer settings can leave you with a part that only the knowledgeable eye can spot as 3D printed.

Need ESD? There is filament for that.

Cost is interesting. If you take 1,000,000 plastic widgets, the injection molded will be cheaper per piece in those quantities, but as you know the tooling is where the $$ is. 3D printing of course has no tooling costs, just the machine, material, electricity. All those scale as production scales. You could get a print farm of say 20 printers and print thousands per year.

[peter-h]

We need to know the volumes and the cost margins

I've just had a few hundred little parts 3D printed. Normally they were injection moulded, 10k+, but this time I wanted a small change done, for a new product. They cost a few hundred to draw up and set up and about £1 each. Very expensive but ok for seeing if this thing will sell.