Sheet metal enclosure design and manufacturing?

[forrestc]

I have something I still struggle with after 20 years of designing, manufacturing and selling my own electronics products and that is actually being successful in getting a custom enclosure manufactured for what seems like a reasonable cost.

I've been using a mixture of modified (in house or by the manufacturer) off the shelf enclosures and also enclosures from a company which folds plastic to make enclosures.  They all work, but they're either off the shelf and look like it or they're very expensive.

I'd like to move to a custom sheet metal enclosure but I haven't ever been successful in either designing or having a design done, and then having it manufactured.  I could tell a lot of stories about things I've tried, but so far, every attempt has failed either without even getting a design done, or by going through the design work (and often paying an engineering fee), only to have the manufacturing piece go down in flames.   Just to be clear - these are generally just a rectangular metal box with some way to mechanically hold the board in place and close the enclosure.

I'm hoping that someone has experience with getting smallish (think small ethernet switch sized) sheet metal enclosures done from concept to metal showing up at your assembly facility, and can tell me how this works best.  Quantities around 500-1000 pcs/yr.

I will add this:  I think the problem is likely that I haven't broke the code about what a sheet metal enclosure manufacturer is expecting from me as a potential customer.  I don't know if I should take a mockup, or do the design work all myself, or let the enclosure manufacturer do it, or what.   So finding out the workflow that works well would probably be a good start.

[coppercone2]

usually the trick you need to do is figure out how to make the enclosure work for you, so you get rid of mounting hardware. Things like inserts (may have a machine to do it), bending metal creatively, etc.

Sometimes its not that the enclosure is that cheap, its that you saved the money elsewhere.

It's all a trade off. Do you need masking? Do the vent punch out need to be flat? can you have a seam?

Unless you know their place well, they basically have to figure out how to make your box with the equipment they have. It means there is a guy there that figures stuff out for money.
Its a hard subject because unless you know what machinery they have and what their employees are programmed to do, its hard for you to make a cheap design that suits them.

It's not like CNC where you can expect some router will follow a path, you need to 'know' things like what kind of bends whatever inserts in their bendomatic5000 works good enough for what you want.

I think its a massive headache.


Might not save money on just the box outside, but if you can figure out how they can make some superstructure inside the box to mount all your parts on with a minimum of screws, you might be in business.

Things that seem obvious/simple to you might make them go WTF because of whatever weirdo way they figured out how to make the parts themselves fast on their equipment (often highly modified/customized for specific big customers).


I absolutely hate dealing with that shit. It is similar to understanding foreign law. 


If you don't try to conform to whatever process they made up, you get hit with massive financial penalties. God forbid the guy feels like running the meter instead of helping you optimize.


I feel like its video game levels when there is a company thats been doing it for a while, they start to understand what the shop is doing, then they suspect MAYBE you can save money there and then you end up with some guy going insane trying to figure out some stupid shape they might be able to use.


Is it appropriate to shake someones hand after dinner, in ancient japan? 

[coppercone2]

also if you google sheet metal engineering you will see there is books on it just like machining parts, it would help skimming some with the basics of understanding how they might do things and what they look at in the prints to see if they will work.

I would start with a glossary
https://leadrp.net/blog/sheet-metal-enclosure-design-basics-and-its-tips/

[forrestc]

Because of all of the weird crap that is manufacturer specific, the process I've been trying to use for the sheet metal stuff is to find a manufacturer which is willing to do the finish design work.

For example, when mounting a PCB, I really don't care how it gets held in the case.  Some shops seem to like PEM inserts.   Some shops like things they can do on their punch.   And, what they can do on their punch is going to depend on what tooling they have, what type of punch it is, and so on.  I've always felt that getting the manufacturer involved to be able to have a manufacturable enclosure was the correct thing to do.   They know what works well and quickly and cheaply on their line, and if they have an internal design team they should be able to knock something out (after all this is just a 6 sided box) fairly quickly.

But... I have yet to succeed with this method.  I've paid several engineering fees, but have never been successful.  So the driver behind my message was that "maybe it's time for me to try something different".

The one metal case I'm using is actually a stock enclosure that the manufacturer cuts extra holes in for me and puts pem inserts in the bottom.  Works great since they have their own tooling for that enclosure and the modifications are straightforward.  Problem is - that company doesn't do anything but their own enclosures, modified, and they're kind of flaky.   I also haven't found anyone else which makes enclosures that I'd be happy with.   

[coppercone2]

Well I only saw it done but it looks like its a major project for basically a ME to do the other part of engineering along with shop visits etc.

If you want them to implement what you want right from cad without the BS, I think you can probobly get most things laser cut and welded, but its premium price.

And wow I gotta say in general alot of the professional engineering services.. if they make it, don't expect them to be reasonable with designing it lol. Almost felt like the right way to do things would be to find a shop that I like, then hire someone that used them before to implement a design. But if they work for them forget about it 

Like I saw simple boxes take literarly forever to get manufactured because of whatever shop specific problem they having. And its not only that, its sometimes the subcontractor, like the paint, plating, welding, measuring, finishing (stainless) in some cases drilling or deburring companies! circus . It might even get sent out to be masked and then sent out to be painted!

Sometimes I feel like someone that said they are a enclosure manufacturer is in fact part of a enclosure 'manufacturing ring' that they seem to do like B2B for.  You will think they do everything themselves, until they blame a problem you find on whoever actually works for them. And they might have insane schedules for 'batch work' etc too.



Its literary shit like some buddy of some guy having some high speed drill cnc big ass machine that happens to have some down time sometimes that he will randomly send out to get work done if he feels like it maybe if you can squeeze money out of it but from febuary to march it will be done by hand drill using a temp worker. and masking might be done by the owners son or sometimes sent out to honduras or mexico to get taped up.


I think its a weird ass world of hook-up's and esoteric information. 


I think when there is some heavy equipment involved and its precise, EE can grasp it because its like monolithicish and professional. You might have like one external thing that comes up, heat treat or plating (even that can get crazy), but Chassis seem to be what 'industrial engineers' are for, because conceivably it can be made with like neolithic labor, and that opens up a giant cost savings area that people try to exploit and it turns into a mess.

[floobydust]

CNC has improved sheet metal fab by removing a lot of the human factors.
It used to be union labour, mostly stoners doing tin bashing. Low quality and expensive.
I had to do a lot of shopping to find a modern shop with decent prices. It's really about their investment in equipment. One shop is all automated, they load up the machines and let them run all night. Consulting with the shop for them to decide the bends and how close PEMs can be etc. instead of doing all the CAD myself.
Although, one shop's policies were to keep those drawings as their own - this caused problems you couldn't go anywhere else.

For fun check out Protocase
The chinese fabs I've never tried and expect the language problem, shipping fees etc. to outweigh their dumping-level pricing.

[temperance]

CAD tools with tools for the development of sheet metal enclosures can be pretty expensive. FS360, also known as confusion 360, and BricsCAD are affordable. The remaining free or low cost CAD packages can't do the job properly without a lot of manual work unless freecad has been improved.

Getting the enclosures made is not difficult. Over the years I've developed four custom enclosures and manufacturing is very competitive in comparison with off the shelf sheet metal enclosures while the custom enclosures has all nuts and bolts pressed in all ready.

As stated by some you must find a good shop. Laser cutting is fine. But if you use plated steel it is better to use mechanical cutting (I think it is called mill cutting which is done with a blade) because the "plating" will be smeared onto the edges of the sheet metal instead of being burned. This prevents corrosion later on. But not many shops offer this kind of machinery and there are some disadvantages in terms of what those machines can cut.

Other things the shop must have:
-An automatic deburring machine. Specially if you go for laser cutting.
-Tools to properly mount press in nuts and bolts. Some shops use outdated hand tool and they are not accurate enough resulting in stand offs "looking" into all directions and damaged nuts and bolts.
-Of course, a CNC bending machine.

I'm hoping that someone has experience with getting smallish (think small ethernet switch sized) sheet metal enclosures done from concept to metal showing up at your assembly facility, and can tell me how this works best.

I've done a box in that size and made 400 pieces last year without any problem.

[coppercone2]

I think that CNC things that are not the classic mill/lathe are sometimes hard to get your hands on cheap. They might not want to deal with you unless you have volume, or it may be charged heavily.

But some people are just lucky I guess.


I thought a decade ago that it would already be 'star trek' but it was star wreck.

Especially for bigger stuff.

I heard alot of people try to setup shops 'for all' and then they end up getting a big contracts by some things nearby that occupies 89% of their capacity, and like.. its just not worth trying to be accessible when that happens, because the work is so much more steady and becomes essentially easy compared to dealing with random people. Then its like, are you gonna give up this steady nice pay to maintain the original business plan... usually not. But then if something expires, they are suddenly very eager to please, but they grow to be more linear then you would like since they ended up working for a small group of repeat customers.. half the settings and capabilities stopped being used, and no one remembers how to do them economically anymore. That means that suddenly you need to specify every single thing, when they were smaller or more diverse, they would come up with 'your question' so often that they basically can read the mind of people submitting things so you don't need to work so hard on blue prints.. .but when that skill is forgotten your gonna end up in blue print hell. Part of a reason why I think that companies often benefit having special shops for R&D because you never know how things are going to change when someone needs the service alot more then you do.

[Smokey]

Not sure what your requirements are, but at 1000s quantities you should also look at injection molding.  That will probably be less expensive than sheet metal and you can get features in your design that would be impossible to make out of sheet metal.  I've had housings for products made by injection molding and then screw them to a simple cheap flat sheet metal plate as a top/bottom/heatsink.  That worked really well actually.

I did an injection molding price comparison recently for a design, and PCBway had the best prices for the mold and the parts.  I was not expecting that.  I didn't actually have that part made so I can't say how PCBway was other than price quote.
I have used Protolabs before for injection molded parts in the 1000s quantities and they did a great job.  Their engineers were very helpful in working out molding issues with the design and they have an online DFM review which is nice.

[Smokey]

I've also done a bunch of all sheet metal enclosures and covers for stuff too.  Here are some ideas for making better designs if you don't have a lot of experience:
1) Look at every existing product you can for design ideas.  There are only so many ways to make a sheet metal box.  Find something close to what you are doing and copy that.
2) Find a sheet metal shop and do a facilities tour.  Ask about their machines.  Talk to them about what you are looking to do and how they would do it given their capabilities.  There are special punches for DSubs for example.  If you want that make sure they have that punch.
3) Ask what form they want your design drawings and if they want 2D or 3D.  I've worked with shops that did great work but only wanted 2D drawings.  Dimensioning 2D drawings so a machine shop understands everything is a pain.  Expect some back and forth with that.
4) If you submit a design, ask for a review when they give you the quote.  Ask what you can do to make the part easier to make and therefore less expensive.  (I still do this with just about every design).

[loki42]

I've got a lovely Amada turret punch and press brake,  pemserter,  mill turn lathe, trumpf fiber laser etc and do all my own enclosures.  I don't know how it is in other countries but it's very easy to find suppliers in China that do an excellent job and great prices.  Don't look for the cheapest stuff,  pay a little more and get great stuff.  Way easier to deal with for me than the Australian manufacturers I've delt with. I do everything myself not for any sane or sensible reason but because I enjoy Metalworking.

Knowing a little about how bending works will stop you from designing stuff that's impossible to bend or requires special tooling

[thm_w]

OPs previous thread was here where we already covered a bit of this: https://www.eevblog.com/forum/manufacture/procuring-or-designing-custom-sheet-metal-enclosures/

If it doesn't need to be metal, then I agree with Smokey can be injection mould. But sometimes metal is needed.

[Mangozac]

OPs previous thread was here where we already covered a bit of this: https://www.eevblog.com/forum/manufacture/procuring-or-designing-custom-sheet-metal-enclosures/

Yes, I thought this had already been discussed recently and wasn't sure if I was going mad!

[TimFox]

Have you tried a custom enclosure vendor such as Protocase?
https://www.protocase.com/

[forrestc]

Yes, I thought this had already been discussed recently and wasn't sure if I was going mad!

Yep, sorry for the semi-repeat.  If I had remembered that my previous post was only 6 months ago, I would have found it and referred to it.

I guess what I'm asking for is more the "business side" of getting these done.   

Since my message 6 months ago, I prepared some not-that-rough drawings in fusion 360 based on existing product enclosures, including flattened drawings.  I've then contacted about a dozen different enclosure manufacturers to see if having dimensioned drawings that they could start from would make a difference. 

For context, the enclosure I'm talking about is 90x40x160 mm.  One 140x60 side has connector holes in it and the bottom (90x160) has some pem inserts for mounting screws.  Two pieces, screwed together.  The type that comes in with every $20 China-manufactured ethernet switch - and that includes electronics and assembly.

From the manufacturers I've contacted in the last 6 months, they seem to fall into three categories:  Way too expensive.  From one vendor, I got a quote for $100/each in qty.  Others I've had to discard because they only want to do runs larger than I'm comfortable with (like 2500 pcs or an insane minimum dollar order).  And the third category wants me to jump through additional engineering hoops without being able to specify any real details - A couple said my drawings weren't good enough and to go get a sheet metal engineer to do a real design and they'll quote it.   Another quoted me a $10K engineering fee just to provide a rough quote.  This is for a drawing which apparently is good enough for a couple of the online sheet metal "prototype" suppliers to be able to quote as-is - at least the bending part of it.

I'm coming back to - this can't be this hard.  I'm doing something wrong.  I just can't for the life of me figure out what the heck it is.  I don't know if I'm not giving google the right incantation to find the right suppliers, or if I'm asking in a way which is a giant red flag for suppliers, or they want me to go old school and send them a paper RFQ, or if I have to make a sacrifice to the enclosure gods to permit me to actually get a quote.

I guess what I was hoping for was some information about what a normal, healthy, process looks like.  I.E. for those of you who buy/use custom metal enclosures what does the process look like.   Are you giving your supplier detailed drawings and not doing any "DFM" type of back and forth?  Are you giving them crayon sketches and they do the work?  Is being charged a NRE just to get a quote normal?  Etc.

[forrestc]

Have you tried a custom enclosure vendor such as Protocase?
https://www.protocase.com/

Admittedly, not recently.

I have gotten a quote sometime in the past and due to that interaction had a mental image of them as a "prototype or very low volume" manufacturer.  Revisiting their website, it seems like they may have decided to do larger volume orders - I'll have to see if their pricing falls off at all after you get past a couple hundred copies of an enclosure.
 

[forrestc]

I've got a lovely Amada turret punch and press brake,  pemserter,  mill turn lathe, trumpf fiber laser etc and do all my own enclosures.  I don't know how it is in other countries but it's very easy to find suppliers in China that do an excellent job and great prices.  Don't look for the cheapest stuff,  pay a little more and get great stuff.  Way easier to deal with for me than the Australian manufacturers I've delt with. I do everything myself not for any sane or sensible reason but because I enjoy Metalworking.

Knowing a little about how bending works will stop you from designing stuff that's impossible to bend or requires special tooling

I have wondered if my preference for North American suppliers is part of the issue here - that is, is it time to give up on Canada/USA/Mexico and go to somewhere else.   Not that I haven't sort of been looking elsewhere as well, it's just been that the couple of quotes I've gotten from Chinese companies haven't looked that good once I added the 25% Tariff and shipping costs.  Plus, I'm not sure tying my ability to manufacture my products to a supplier in a country that our incoming president has vowed to continue to raise tariffs on is a smart thing to do.   

I have thought a few times about getting a small press brake and either a punch or a fiber laser+pemsert press.  Then I realize that's just another set of distractions I don't need, but I keep getting dragged more and more that way as I get more and more frustrated with the finding a supplier issue.   

Oh, and back 20+ years ago when I had access to a manual shear and press brake I would make my own enclosures for small electronic projects, so I'm at least somewhat familiar with how to build boxes so you can actually bend them without running into tooling (or other metal) conflicts - although some things like bend radius, how much metal you have to add/remove with certain bends, and the like is outside my current knowlege.  That said, sometime when I quit making things I have to make a thousand or so copies of I might just buy me another little manual brake/press since I really enjoyed making my own enclosures - but that is a completely different thing than running a in-house sheet metal shop.

[thm_w]

Just my same views again but:
- Detailed drawings are given to supplier
- DFM back and forth is normal with any supplier - one example is the supplier asked if we could do thinner sheet as it would be easier to manufacture, we said yes, they scaled the drawings themselves
- Paying for a quote is not normal, quote should be free. Any samples expect to pay NRE, setup fees, etc. though.

If the online sheet suppliers can quote from the drawing, it probably is good enough, but you are free to post here. Maybe they were worried you left out critical dimensions, etc. but they should know to go based off standard spec they can hit.

[forrestc]

Not sure what your requirements are, but at 1000s quantities you should also look at injection molding.  That will probably be less expensive than sheet metal and you can get features in your design that would be impossible to make out of sheet metal.  I've had housings for products made by injection molding and then screw them to a simple cheap flat sheet metal plate as a top/bottom/heatsink.  That worked really well actually.

I've got another project that I'm looking at injection molding for.  Or more accurately a series of products that only differ by connector set.  Since I already have plastic enclosure milling equipment in house, to keep costs low (and not have to play mold insert games), I plan on getting the mold done without any connector holes and then milling the holes in-house depending on the product.  Thanks for the suggestion and details, though, as I would have never thought about somewhere like PCBWay.

But, for the product I'm looking at now, I really want to stay with metal for two reasons:  The first is that these are often installed at high-rf sites so the natural shielding from a metal enclosure is a plus.  The second is that these products have enough power that I want a flameproof enclosure.  As a result, for plastic, I'd want to use UL-V0 self-extinguishing plastics, and for whatever reason anything flame resistant in the plastic world has gotten really really expensive, so in theory it should be cheaper to just go for metal.

[Smokey]

Have you tried a custom enclosure vendor such as Protocase?
https://www.protocase.com/

Nice.  They do die casting.  Might need to give them a ring and get some quotes.  Thanks.
https://www.protocase.com/products/diecast-enclosures/

[Smokey]

Have you tried a custom enclosure vendor such as Protocase?
https://www.protocase.com/

Nice.  They do die casting.  Might need to give them a ring and get some quotes.  Thanks.
https://www.protocase.com/products/diecast-enclosures/

https://www.protocasedesigner.com/
The Protocase Designer program is actually pretty slick.  Lots of metal enclosure templates with CAD models.  If nothing else, just having the cad models and built in viewer all in one place is a useful resource. 
It looks like some of their die cast enclosures, in the app at least, are Hammond boxes that they modify.