Sealed/semi-hermetic packaging

Hi,

Exploring more options...

I'm trying to enumerate the issues that would accompany packaging a device (1-2 "unit-fists" volume) in a "sealed" enclosure. By sealed, assume "watertight" (though not to any significant depth) and "connector-less".

Of course, you have to deal with I/O's -- that's easy so forget about them.

Beyond that, the real issues seem to be:

- getting power *into* it

- dissipating power consumed within it

- replacing consumables (e.g., *all* batteries will die)

- servicing

- discouraging tampering (because that makes it non-watertight)

- mechanical robustness (so a drop doesn't render it useless)

I wonder if a sealed "black box" is likely to also cause troubles traveling?

Any pointers to consumer products that might fit these criteria?

Thx,

--don

Reply to
Don Y
Loading thread data ...

A NEMA 4 or 6 enclosure with MIL 38999 connectors and environmental backshells?

Oops... never mind.

--
Rich Webb     Norfolk, VA
Reply to
Rich Webb

I suspect some of these are generally contradictory - e.g. if something is intended to be tamper proof that tends to limit serviceabilty (assuming the tamper-proofing is not illusory to begin with, e.g. "security" screws). Potting the circuit in resin will result in spectacularly robust, well sealed and tamper-proff enclosures but if something goes wrong inside you're basically throwing the device away.

--
Andrew Smallshaw
andrews@sdf.lonestar.org
Reply to
Andrew Smallshaw

Not at all! E.g., you could solvent weld the case. Or, certain "snap together" designs would defy tampering (and make attempts to do so very obvious) yet still result in a serviceable product (you simply replace the case as part of the service procedure... tens of cents of "discarded plastic".

Any sort of visible fastener (e.g., "security screws") will

*invite* tampering. And, chances are, you'll end up servicing devices where the screws are now no longer "removable" (because the user mangled the hell out of them trying to figure out a way of working around their "unique nature") [assuming, of course, that the user doesn't simply *purchase* a compatible "wrench" to do so!]

It also complicates the thermal problem.

Reply to
Don Y

No, it's not easy. Having I/Os is in obvious violent contradiction to the thing being "connector-less". Even in "quotes".

Impossible. You wanted connector-less.

Easy enough, as long as you can make sure nobody touches the thing, so it doesn't matter how hot it becomes.

Having batteries somewhat invalidates requirement #1 about getting power in there.

These two directly contradict each other. Servicing _is_ tampering. It'll just be carried out by slight less clueless people --- if you are so lucky.

Easy enough, since you forgot to put a constraint on weight. 1 inch of steel with 1 inch of rubber inlay should do the trick.

The 4 D-cell MagLite. ;->

Reply to
Hans-Bernhard Bröker

Any "wire-less" technology will do. Radio, Ir, etc.

Couple magnetic fields through the case. E.g., essentially make two halves of a transformer -- one inside the case, the other outside.

Get power in to *charge* the batteries.

Yes. But when "servicing", you have available to you whatever parts you are likely to need in performing that service! E.g., your service procedure can *intentionally* destroy the "tamper proof case" *knowing* that you have boxes of "factory fresh" replacements. A "user" doesn't have that option.

(sure, the user could smash it open with a hammer... but, it would be hard for him to claim he didn't "destroy" the product prior to shipping it to you for repair!)

Hardly qualifies as "tamper proof" since you can easily access its internals. *All* of its internals!

Reply to
Don Y

Hi, Don.

In comp.arch.embedded, you wrote: [..]

More fun!

One way is inductive power transfer, something my rechargeable electric toothbrush makes good use of.

The toothbrush charger uses (I assume) simple sinusoidal AC. A more complex waveform (going in) could still be used to derive power while containing information, just as you can use the RF energy from an AM broadcast to create a gnat's-hiccup power supply or to create sound waves (some of which may be even worth listening to ).

As for getting data _out_ of the box, it would need (for my example) to emit an inductive signal. You could separate input and output by physical distance (two coils at different spots) or use some sort of half-duplex protocol. "CDMA" collision-detection is presumably possibl in this situation, but it feels like it would be a bit more complex.

Avoid creating it in the first place if at all possible.

In order to leave the device, heat has to be transferred to the housing, and you probably will need some way of mounting it so as to absorb as much heat as you can from the housing to the mounting.

a) Feed power constantly, and use only non-volatile memory.

b) If you assume you can feed in power on a frequent-relative-to- consumption basis, a fraction-Farad "super capacitor" such as the $4 Panasonic EEC-S5R5V105 (1.0F, 5.5V).

Or, to put it another way, how to permit servicing that won't make it "non-water-tight". I can see two directions to take on this:

a) Discourage undetected tampering/servicing (e.g. "Not covered under warranty because someone _else_ messed it up." This just requires a "seal" of some kind.

b) Really prevent foolish/malicious tampering. This is difficult, sort of like computer system security; how much do you want to spend making it how difficult? Even "one-way" bathroom stall mounting screws can be extracted, as I discovered in high school. One option is to _really_ seal the unit, perform service via "swaps", then use a hacksaw (or whatever) when the swapped-out unit returns to the shop.

Will it be sensitive to airport scanning? One quick test for possible damage would be to put the "innards" into your carry-on luggage _without_ the housing, the reasoning being that, if the components made it through unprotected, the assembled unit should be okay when housed. Then you actually test this reasoning with a couple of assembled units.

As for avoiding hassles with TSA when you want to "check" or "carry on" (say) demo units, much would (I'm guessing) depend on whether the box were opaque to whatever scanning methods you run into. Beyond that, that is, as far as convincing an arbitrary TSA security person that a complex conglomeration of electronics, "batteries", motors, and assorted lumps, I have nothing to offer except to suggest FedEx and friends.

Hope something in all this helps.

Frank

--
    Southern DOS: Y'all reckon? (Yep/Nope)
Reply to
Frnak McKenney

What Don said later. Also, lately, people have been experimenting with ultrasonic sound to run power and signals through submersible hulls. Still experimental, I think.

Mel.

Reply to
Mel

Or use the same method used by RFID. IIRC RFID shorts the pickup coil to send information back to the reader device.

Reply to
Dombo

Yes, I exist solely to entertain you, Frank! ;-)

Yes, I had thought of that. But, its hard to couple much power across that "divide". I suspect you might be able to locate some ferrites (and, opt to a higher switching frequency) to help improve the coupling. Not sure how mechanically robust that would be, though (e.g., ferrites like to shatter when subjected to stress... "drops")

I'd also considered a simpler, more direct approach -- make the two halves of the case be two "contacts" (i.e., make the case itself *be* a connector).

The problems I saw here are:

- if those surfaces form mechanical connections with their "mates", over the long run, you end up with the case looking very deliberately "scuffed" from the abrasion of those "meetings and partings" (I don't know if there are any materials that would be conductive and "impervious" to that sort of abrasion)

- a metal shell limits I/O choices (e.g., RF gets ruled out or forced to "cheat")

- a metal shell would probably greatly upset the TSA folks

(otherwise, a nice *polished* chrome-like case would be wicked cool!)

This would only address the I/O ("I") requirement when the device was "in its charger". You'd still need a way of getting data in/out while "deployed"

That goes without saying. Every microwatt counts. This also has repercussions regarding other technologies used *in* the box (i.e., your communications subsystem can't be wasteful of power)

This was another advantage of the metal shell. You could judiciously distribute the heat pattern around the case so that there was no *one* "hot-spot". If it feels slightly warm to the touch, people will get used to it. If there's a hot spot, people will complain each time they *forget* its location and are rudely reminded!

I think that only works at much lower power levels than I'm talking about. Figure a couple to many watts, average, in my case.

Yeah, but that is easily ignored. You also have to address the folks who deliberately ignore that -- and keep them from rendering *you* subsequent servicing "difficult".

E.g., the metal case idea couldn't rely on fasteners as some bozo *will* try to hack together something to try to remove them (usually buggering them in the process).

Every fastener has to be removable *somehow*. Keep that in mind and you know there will be the temptation to attempt it.

Exactly. If the case is considered disposable (a *plastic* case costs less than the postage to mail a letter informing the user that he has voided his warranty) and you *plan* for its "removal with extreme prejudice" during servicing (i.e., locate things internally so that the means by which you remove the case doesn't put them in jeopardy), then you get the assets you seek with very little associated cost (time or money)

I'm more interested in "customers" having theirs "confiscated" while boarding aircraft, etc. It's too high a value item to forfeit at the checkpoint (unlike nail clippers, etc.).

I've learned, when traveling, not to wrap things in aluminum foil (e.g., baked goods). These folks have enough problems to deal with. I don't want to complicate their jobs needlessly (e.g., use plastic wrap, instead, and just plan on consuming those goodies *sooner* rather than *later*! :> )

There's never an easy answer. I really like the metal shell solution but it seems unworkable for too many reasons. But, maybe a plastic shell with some flush (Au) contacts might work. Not sure how those would fare when exposed to water... :-(

Reply to
Don Y

[...]

Forget temptation. Think about guarantee. As in: some border control agent somewhere _will_ open that device, just to make sure it's not something else.

So how will you distinguish units opened for inspection by border control agents, from those maliciously tampered with by the customer themselves?

That may be your ultimate show-stopper. Even after you've essentially given up servicing (_crushing_ devices once they've been sent back to you, isn't), there's still one type of tampering/inspection by someone other than yourself that there's really no way get around: customs.

And you already know that making it harder than necessary to those guys will not help anyone in any way. You might as well put a bright red padlock on it, labelled "TSA won't open _this_!"

Well, Don, I'll try to put this nicely: that may be because sometimes you do appear hell-bent on _never_ accepting any of those.

Reply to
Hans-Bernhard Bröker

That would be my main concern. Although I've no direct experience (I hate travelling and avoid it as much as possible) I have heard many stories about anything that looks incomplete or amateurish being labelled "suspicious" and while the equipment isn't exactly confiscated its made very clear that things will be a lot easier if you'll just dispose of that here... Usually after your flight has already been delayed an hour and a half.

--
Andrew Smallshaw
andrews@sdf.lonestar.org
Reply to
Andrew Smallshaw

Ah! A sign that the Universe is operating properly.

Enough to match the stated specification.

How much do you need, and how often?

Or (equivalently) embed insulator-surrounded contact points of an appropriate size in the case.

Perhaps. I don't know what kinds of metals block their scanners. All? Ferrous only? Does anyone else out there know?

I see that my example inclued an untended limitation. Whether you charge or use the inductive power as your primary power depends on what you wish to accomplish. You (or conceivably the customer) could charge batteries quickly (heat!) or slowly, or you could surround the box in situ with a power+I/O coil that provided constant power.

I realize that it's often difficult for someone to completely spec-out a new project, but anyone commenting back to you can only go with their best understanding of what you've described to date. The result is usually an iterative process, with much guesswork involved, and occasionally frustrating to either or both parties.

As far as the I+O, I don't think you've said whether you're thinking of a data logger (e.g. data out-loaded only infrequently) or somthing which reports data constantly. If the latter, there's no reason the customer's "data-extraction box2" can't charge any batteries and update configuraiton information (e.g. box input ) while downloading the captured data.

[...]

Can the power your "box" will need be provided by batteries?

Is the "box" usable (sellable) if a dead battery requires (from the customer's viewpoint) complete replacement of the unit?

"Seals", at least as far as the local power company's metal tags and those manufacturers who slap labels over vital screws, are not intended to prevent unauthorized access. They're there to make sure such access can't be done without leaving a clear trace ("doing this will void your warranty").

But... are you merely responding in the abstrace ("someone could do it that way"), or it is this a "usable scenario" for you?

[...]
[...]

The picture I have of the process is that much is left to the discretion of the TSA staff who happen to be present. It sounds likely that your "box" would fit into the "we don't know what it is" category ("it's not a camera and it's not a laptop and its not..."). That's likely to present problems, even if you plaster it with red-barred "NO ..." circles and NOT A BOMB! labels drawn by Antonio Prohias.

Making the case out of plastic -- perhaps even clear plastic might make a difference, but this is pure speculation. I suppose you could try calling TSA for advice.

[...]

Good luck.

Frank

-- "So many ideas that sound crazy are indeed crazy, and if a busy man paid equal attention to all of them he would never get anything done. The test of a truly first-rate mind is its readiness to correct mistakes and even to change course completely -- when the facts merit it." -- Arthur C. Clarke

-- Frank McKenney, McKenney Associates Richmond, Virginia / (804) 320-4887 Munged E-mail: frank uscore mckenney aatt mindspring ddoott com

Reply to
Frnak McKenney

Wow! That's got to be *incredibly* inefficient (?)

Is this a direct consequence of (and *exploitation* of?) the "submersible" aspect? (it seems like air as the coupling medium would make the idea "impossible"... but, that's fyziks and, therefore, BFM in my book! :> )

Any ideas as to what the end-to-end efficiency is?

Reply to
Don Y

There's nothing you can do to completely prevent this -- nor would you *want* to (as that would represent a "security hole").

I went to a local mall some time ago. Wearing a "bandana" to tie back my hair. After just a few minutes, a security guard approached me and told me I *had to* remove the bandana -- or exit the mall. On asking for clarification, I was told that they have a problem with "gangs" and (apparently) gang members identify themselves with bandanas (presumably different colors -- I wasn't "quick" enough to ask which colors to *avoid*).

I ended up leaving the mall -- not because they wouldn't let me wear my bandana but, rather, because I had no desire to put myself at risk in an environment where the mall management felt there was enough of a problem that they would risk alienating *customers* with such a policy. (i.e., there are lots of malls; if you put up any sort of impediment, I can easily take my business elsewhere!)

As I walked back to the car, I started musing over other behaviors that "gangs" could presumably adopt and their (potentially amusing) consequences:

- "I'm sorry, Sir. You'll have to remove your shoes or leave the Mall as we have a gang problem and they use their shoes to signal their affiliations..."

- "I'm sorry, Ma'am, you'll have to remove your *blouse* or leave the Mall as we have a gang problem and they use red shirts (blouses) to signal their affiliations..." etc.

I.e., there was no other (practical) solution for the Mall management.

Come up with a design that isn't likely to draw the attention of TSA (border control is a lesser problem) agents! :> E.g., if the case was *not* opaque to X-Ray, I suspect it would be less likely to attract attention. Or, if it wasn't opaque to "visible light"!

E.g., there is nothing to stop them from subjecting all cell phones to detailed inspection. Or, iPads. etc. (they already do it with *shoes* :> ).

The more important issue is to have a design that doesn't cause them to want to *confiscate* the item out of fear or ignorance.

I'm an Engineer, not a Mathematician. Very few of my problems have "closed form solutions". *Every* solution has tradeoffs involved. You *never* get a list of criteria with

*weights* associated with each one that allow you to reduce the solution process to one of linear algebra... "which solution gives me the highest numerical score".

I pride myself on being able to, at least, *imagine* aspects of solutions that could easily prove to be problems after-the-fact. (would *you* have considered the impact on air travel of the "sealed packaging" issue?) It is entirely possible that there is

*no* acceptable solution. Or, that circumstances will change after the design is complete which add new criteria that weren't reflected in the original "calculation".

The more of these issues you can raise and address in the design phase, the better, IMO, the resulting design. I am not ashamed to tell a client "This is the best I could come up with -- and here are a list of flaws I see in this, *my* design." Chances are, I've thought of more issues than he could *imagine*...

Reply to
Don Y

Ditto. It was (marginally) "fun" when I was younger. But, too many coast-to-coast flights and attendant delays, c*ck-ups, layovers, etc. have managed to squeeze every *ounce* of "fun" out of it!

I'm not worried about prototypes. Those *I* can arrange to ship "hassle-free". Rather, I am worried about production devices that aren't mainstream enough to get coverage on The Nightly News.

Reply to
Don Y

Completely forgot where I found this -- here it is:

Strange article: started out looking like spy ware, but I can only see it as communication on purpose.

Mel.

Reply to
Mel

"And, now, for your *listening* pleasure, the dulcet tones of The Kolodny Triplets singing, 'Row, Row, Row Your Boat' -- in Swahili ..."

Usage model would be similar to someone who talks on the phone *continously* (i.e., need more power stored than a typical phone has available... and, need to be able to replenish that fairly quickly -- how long do you want to be without the use of your "phone"?)

Yes. E.g., I thought of "flush" gold contacts embedded in a high-impact plastic. I think any "direct" contact has got to be an order of magnitude (or more?) more efficient than anything I do indirectly. The trick is to avoid a "real" connector, etc.

Even if it wasn't opaque to the scanners, a "manual inspection" could rouse suspicions. "You say it's a computer? Well, where's all the *buttons* and *connections*...?"

E.g., I think of the advertisement re: the "dockable phone" that now offers -- and the "agents" querying the owner, suspiciously, "Well, which is it? A *phone* or a *computer*? ..."

Sorry, it's primarily a portable device (I failed to mention that). Of course, it should continue to be operable "docked" (else the charge cycle would have to be *extremely* fast -- or, some scheme of swapping battery packs...)

Of course. I've found that the *less* constraint placed on queries, the wider range of potential solutions you make available. Some will obviously not fit. Others might be ruled out before being offered -- based on preconceptions of the potential respondent. I'd rather be exposed to as wide a variety of approaches as possible as something that doesn't make sense ... *might*!

No, the usage pattern is more like "someone on the phone constantly". But, you can get that sort of I/O with wire-less (connector-less) technologies -- at the expense of complexity and power :< (a wire is a lot more efficient than a "less-wire" :> )

Yes. Though I am working hard to reduce power requirements to get prolonged use without unduly increasing battery pack size/volume/weight.

Yes. I expect to have to replace the batteries "at the factory" unless I can come up with a scheme to separate the battery from the electronics -- and still retain all of the above criteria ("water tight" becomes a problem)

Yes. But, folks will still "try". E.g., "removing this sticker voids your warranty" *obviously* placed over one of the screws to open the case is bound to be removed on some percentage of devices that you see for service. Then, you end up in a pissing contest with the owner. You *want* the owner to win that contest (so he doesn't irrationally get annoyed with you "for being a stickler").

So, if you can eliminate the temptation *completely*, then you are never faced with the issue of confronting the user *or* having to "suck it up" when you receive something with a broken seal.

I think if the user has to take a *hacksaw* to the case, it's a lot easier to hold your ground and say, "Sorry, your warranty has been voided because you cut the case open with a hacksaw. We will return your device to you for the cost of postage.

*Or*, replace/repair it for "

No, this is a viable scenario. *If* the case's cost can be kept low (i.e., plastic or plastic with some metal contact inserts), then you just treat it as a cost of doing business. So, you figure a new battery costs the price of the battery plus the price of a new case (plus labor, etc.).

IMO, this is the only (practical) way to get serviceability with "tamper proof".

I think that is the only approach that would make sense! A "list of approved devices" would be too easy to subvert. I think they would have to rely on circumstances "at the time" -- their impressions of the person carrying the item, current world conditions, etc.

You (I) don't want to give them cause to be *less* vigilant.

OTOH, if everyone's iPad was confiscated, there would be a huge outcry!

[no idea who "Antonio Prohias" is... :< ]

But, you've hit on the problem, exactly. The more "unexpected" it is, the more likely it is to receive extra attention.

E.g., many phones can "voice dial". So, imagine your cell phone was just a slab of chrome and plastic. No buttons (hey, it's got *voice* capabilities!) and no display (why do you *need* a display? can't you *ask* it questions and *listen* to its replies??). You've got a bluetooth earppiece. The phone sits in your purse/back-pocket/briefcase/etc. -- why bother carrying it when the "interface" fits conveniently in your ear??!

How is *this* received by the screeners?

"You say it's a phone, huh? Well, where's the buttons? Where's the display? Where do you plug in the charger??"

I think that's going to be the only way to get a semi-definitive idea of what to do and what to avoid. I fully expect that "opinion" to be deliberately vague... :<

Reply to
Don Y

Yikes! Too far above *my* head! Unfortunately, lots of details omitted (e.g., my efficiency request).

Fun to see "novel" approaches to problems, though... (thanks for the link)

Reply to
Don Y

Magnetic stirrer technology!

formatting link

I've actually seen this used to drive an impeller to pump a fluid without a physical connection. This was in an application where leakage would have been Very Bad (not only were two of the fluids being pumped fairly toxic, they reacted with each other instantly and

*very* exothermically). The aluminum pump housing was welded shut (and then welded to the inlet and outlet pipes), and had no seals or anything that could have leaked.

Probably driving a small generator that way is not ideal, but you can transfer a fair bit of power if your driven rotor is a big neodymium magnet.

Reply to
Robert Wessel

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.