Enclosures, paint and heat dissipation.

[...]
[...]

Jon,

I'm jumping into this conversation a bit late, and it sounds like you already have a workable solution for Reader #1. However, if you decide you don't want to do all that drilling and grinding and sanding and lens-polishing for Reader #2, you might consider letting a manufacturer do some of the grunt work for you with a device like this one:

Linear Sensor Array : TSL1401CS-LF Overview:

The TSL1401CS-LF (lead free) linear sensor array consists of a 128x1 array of photodiodes, associated charge amplifier circuitry, and a pixel data-hold function that provides simultaneous-integration start and stop times for all pixels. The pixels measure 63.5 um (H) by 55.5 um (W) with 63.5-um center-to-center spacing and 8-um spacing between pixels. Operation is simplified by internal control logic that requires only a serial-input (SI) signal and a clock.

Diodes, already laid out for you "like pretty maids all in a row" (but hopefully less contrary than the gardener ). Requires

3.3-5V and "less than 5mA".

You can get it mounted on a board with a lens from Parallax for $50:

TSL1401 Linescan Imaging Sensor Daughterboard

And there are others out there. Scavenging one from a $5-10 thrift store scanner is another approach, but unless what you dig out is clearly markes and an item with an available datasheet you may be in for a lot of work.

But don't let any of this spoil the fun you're having building Reader #1. It's a good lesson on the difference between creating the first, proof-of-concept prototype (I seem to build a lot of these! ) and building a production prototype.

Hope this helps...

Frank McKenney

--
  "The true rule, in determining to embrace or reject any thing, is
   not whether it has any evil in it; but whether it has more of
 Click to see the full signature
Reply to
Frnak McKenney
Loading thread data ...

On Aug 23, 2:20=A0pm, Ecnerwal

My understanding is that glossy is better than flat. Flat generally being glossy with talcum powder added.

Dan

Reply to
dcaster

Silly man. flat black paint, especially for use in hot applications, uses CARBON.

Reply to
StickThatInYourPipeAndSmokeIt

I disagree. The glossy surface is a better reflector than the flat; that reflectance works in both directions, ergo flat is better. In my not so humble opinion. ;-)

But if I was using the enclosure itself to dissipate heat, I'd use a finned heatsink, black anodized.

Cheers! Rich

Reply to
Rich Grise

Paint, oil: average of 16 colours 0.94 Paint: aluminium 0.45 Paint: oil, black, flat 0.94 Paint: oil, black, gloss 0.92 Paint: oil, grey, flat 0.97 Paint: oil, grey, gloss 0.94 Paint: oil, various colours 0.94 Paint: plastic, black 0.95 Paint: plastic, white 0.84

Ref:

formatting link

Cheers!! John S

Reply to
John S

A flat or 'matte' surface quality is more emissive. You got the reason right. The IR actually gets reflected back into the medium if the surface is polished, and if it is matte, there are millions more angles of departure for the IR, so the overall effect is MORE emissive for a given temperature of a given medium.

The anodize actually reduces emissivity. If it is a conductive cooling scenario, a thick base plate to act as a spreader for the heat being soaked into it, and a good, dense medium to attach that plate to for the heat to soak into.

All small package form factor electronic devices that dissipate a lot of heat are reliant on being attached to a good sink if they are of a conductive cooled design. If they use a finned sink, then moving air is a requisite over said fins. Otherwise it merely acts as a poorly designed conductive cooling scenario.

Any case walls being thin in nature are a bad thing for conductive cooled designs. You want at least one thick plate to sink your heat generating elements into. Trying to pass it to a thick sink outside the box through a thin wall is a VERY BAD idea. You would need to incorporate an interface plate between the inside and outside of the box.

Reply to
Pieyed Piper

=20

All valid ideas. The oiled paper for punches was about 0.01% oil for lubrication, more than that caused problems also. Oiled paper for = windows was almost 2% oil, with much more impact on transmissivity. For = comparison check wax paper for kitchen use in the grocery store.

=46or even greater fun some papers had plastic layers or even metallized plastic layers. And some tapes were flat out metallized plastic (often mylar).

I worked with/on ASR-33 type equipment in the early 1970s.

?-)

Reply to
josephkk

It sure sounds like a lot of fun.

Engineering is making what you want out of what you have.

?-)

Reply to
josephkk

I just spent last night reading in my first papertape in a long time. I'd forgotten the odor/smell of oiled paper tape. But now all the memories flooding back. Of course, also the winding and talking while winding with someone else nearby.

Interesting thing. While I've just been going through the shed uncovering these old items, it turns out that the paper tapes are still just fine. But that the old EPROMs have actually lost their memories!! I suspect these tapes, even modestly taken care of, will outsurvive many other forms of recording data -- including writable DVD and CDROM.

Like papyrus, almost.

Well, I need to build up a linear light system -- long filament followed optically by a glass or plastic rod (cylinder lens) to reimage the filament onto the detectors as a bright line.

This is a fair bit of fun, again. Went down to the store to buy an aluminum cooking pan and some Coleman stove pots (also aluminum) so that I had some soft metal to save my poor bits from destruction in making the thing. Forgotten what it was like to build up my own computer components, with the ease of buying things in the huge market that now exists. Back then, it was all by hook or by crook. And here, once again, it is also. Love it.

Jon

Reply to
Jon Kirwan

I just posted a little more elsewhere on this very point, before reading what you just wrote. I'll add some details here.

I started out using vector board as a drill guide. It's 0.1" spacing, nicely. I was able to make tap points that way exactly where I wanted them and very nicely aligned without the need fo fancy equipment. Just a cordless drill and no fixturing. Once I had that done, making the .06x.12 slots wasn't difficult.

By the time I'd worked this out, it was late. I went down to the store around 10PM, which meant not going to Home Depot or the like. I wanted a specialized bit for milling out each slot. A large store that included groceries but also some tools and the like had a dremel bit that was just right. But sadly, not carbide or strong -- specified only "soft metal." Okay. So I decided to get it anyway, give up on the piece of steel I was planning, and ran around looking for aluminum. In the cooking section, got a nice rectangular baking pan (without insulation compressed inside it) made entirely from aluminum. Also, picked up some thinner stuff found in the Coleman camp stove cooking stuff -- 25% sale, good luck. Took all that home, night before last, and worked.

I do need to get a linear bulb with a filament of about the right length. Incandescent. Then add, just below it and also horizontal, a glass or plastic/acrylic cylinder rod to refocus the filament onto the detector array. Like a ball lens would do for a fiber optic system and a point source like a diode, for example, but in 'array form' which means dragging out the ball lens into a cylinder instead. That should make a very nice, bright line exactly where I want it.

This whole thing is a lot of fun and reminds me of when I didn't have _any_ money (few know what I mean as very few people I knew then were anywhere near as poor -- so I mean it here) to speak of and had only just gotten a job at $525/mo. Had to build up everything, if I wanted it. I would literally scrounge dumpsters at the phone company or elsewhere for electronic parts. I'd dismantle TVs, you name it.

I could buy all this stuff. Or I could use my friend's Remex and just be done with the functional parts about getting these tapes copied into a modern computer. But it wouldn't be nearly as much fun and I can use his unit, anyway. So it's more about recovering some of that fun -- and learning experience as you go -- and teaching some kids who are watching all this, right now.

I think some of this attitude has been lost, somehow. And it is sooooo much fun and provides awareness and self-confidence and pride, too.

I'll probably do up a finished hardwood surface and bury the aluminum optical head so that the aluminum detector plate is flush with the wood. There will be a top aluminum piece placed with a gapping of about 0.02" above the detector plate to keep the papertape closeby while it goes through and I will use a thin sheet of aluminum along one side to both create that 0.02" gapping as well as an 'edge' against which the papertape gets properly aligned (I won't have sprocket guides to start.) Will likely add some kind of washer-based thing on both the incoming and departing sides to keep the paper tape pressed against that guide.

Used to be able to buy these heads, I recall (not that I could afford them, back then.) And now you can't even buy them. But you can make them from scraps and from kitchen supplies!!

This is fun.

Jon

Reply to
Jon Kirwan

Could you post something like your visible and IR comparison pics that = you once did for kapton film/tape? =20

?-)

Reply to
josephkk

Maybe you could use a cold-cathode backlight from a laptop? Fairly bright and already a 'line source'. Or, cut a slit in some aluminum to make an aperture/mask to place in front of a regular incandescent bulb. Probably still need a condenser lens as you mention. A glass stirring rod used in chemistry might work. Or if there is a stainedglass hobbysupply around, ask them if they also have 'lampworking' supplies or know where you could get some clear glass rod.

Reply to
1 Lucky Texan

This is fun:

formatting link

One other interesting thing: water, ice, and snow are black at thermal wavelengths, e around 0.98.

John

Reply to
John Larkin

Well, except maybe that the amount of paper needed to hold the content of a single DVD could be heavy enough to _crumble_ under its own weight, if doesn't crash right through the floor before that. The paper to store a non-negligible DVD collection might just have enough mass to collapse itself into a black hole ;-P

It's one of those "the internet still can't beat a 747 full of DVDs on bandwidth" kinds of argument. Yes, well-made books can outlast a lot of higher-density archive media --- but what good does that do you with data being produced at a higher bandwidth than the trees can be grown to make the paper to print them on?

A "point" source parallelized by a spherical lense/mirror and then focussed onto a line by a cylinder might be a whole lot easier ...

Reply to
Hans-Bernhard Bröker

Indeed. In our physics lectures they used raw arc lights to project experiments onto a wall several meters away for hundreds of students to see nice and big, and without having to dim the normal lights.

Now unfiltered arc-light is seriously bad news if your experiment is heat-sensitive at all --- worse even than halogen light bulbs. So whenever that was an issue, they just parked a glass box full of water in the light's path, and that was that taken care of.

Reply to
Hans-Bernhard Bröker

That would depend on how fine the printed "data" on the "pages" would be, and how much of the page that print covered. Probably fit it all on less that a ream of paper, done right. Zeros and ones can simply be "dot" or "no dot".

Must have been where your massive brain went.

Which is stupid because the Internet DELIVERS that much data from many to many PCs. The 747 only carries a copy of any said data, and final delivery onto even a small number of those PCs would make for a bottleneck that would negate the claim.

But my gateways are the fastest in the world. 10GbE fiber all over the place. Now I understand why the monthly rates are so high. This shtuff ain't cheap. The cables alone add up to millions of dollars. We give tens of millions of dollars a year in business to Cisco, Juniper, and cable companies, and contract makers.

We probably spend $30M a year on HP computers alone.

Now, those systems move a LOT of data.

I used to watch every movie made in a year. There is just no way now, with my regular job, and everything else to do. There are just too many movies. So we BUY the movie in archive form, IF it got good reviews and sells well, on a DVD or other medium. To watch at our leisure. No, DVDs are NOT "burned". Only those you make at home are.

"Optical cubes" can hold a TByte, but they are far more costly to make, and hard drives are cheap and reliable. They are going to be around for a long time to come. Our data is safe. Sorry.

Reply to
My Name Is Tzu How Do You Do

A point, were it that I wouldn't be selective about what I wanted to last so long. But I could be. of course, I'm not planning on going out and making papertape my storage of choice. Just going out and having fun, right now.

Focal distance from the rod would be half the distance that way. But it would allow the use of an LED light source, perhaps. I'll play with it. Thanks. But it will also work with a cylinder (rod) alone, given a line filament (which is not abnormal for incandescent lamps.) So I have options, I guess.

Jon

Reply to
Jon Kirwan

Not as much as one might think. If the goal is data integrity and longevity, you can't print too densely. The usual assumption used in those typical "holds that many typed pages" is on the order of 2 KiB per page. Go much higher than that and running ink or other processes will destroy the data's readability over time just as easily as stray UV light probably did with Jon's EEPROMs

Don't tell me about it. My background is high-energy physics. My former colleagues at CERN basically get to define what "high data rate" means. There's not many other people who will use "Exabyte" in the plural form --- and feel no need to explain that it's not 8mm data tape cassettes they're talking about.

Reply to
Hans-Bernhard Bröker

My birds stream OUT (and in too) over 300Gb/s across several tens of miles bigger gaps to tens of thousands more endpoints quite remarkably, in fact.

Your hula hoops gather ten times that amount of data fed from thousands of transducers, and pass it into huge supercomputer data processors over short distances hard wired before feeding that pre-compiled data into still enormous data streams into yet more supercomputers so that you guys can sit at wonderfully crafted graphics processing monster computer 'workstations' allowing huge numbers of scientists to visualize the collisions, and work with the data that created the graphics renderings so we can make a feeble description of the BIG FART that started our little corner of the macroverse.

Good luck getting it right. It has been fun watching it all (from a distance).

Reply to
My Name Is Tzu How Do You Do

That's silly and unnecessary exaggeration, Hans-Bernard. Such taken-to-extremes language only works to undermine one's credibility than to make a non-sequitur, but true point any better. It is better said without being careless.

I have here a Cauzin Reader that was for a short time used to transmit data in magazines to their readers using easily printable marks registered in mass produced publications on cheap paper and lousy ink. 1985, or so And it provided 1k bit per square inch. 160 square inches on both sides of a US letterpaper would be 160k bit or 20k byte using simple optical means. This is an order of magnitude beyond what you are saying and even at that it probably can still improved. Certainly, using reasonable quality paper and modern methods of fused plastic ink even that could be improved along with retaining millennia lifetimes.

formatting link

(My unit still works just fine, by the way.)

Even an ancient papertape achieves 10 bytes per square inch. And this is with huge holes and I can read them, if perhaps slowly. About 1k byte for an 8.5x11 area, punching very huge holes (which could itself be readily and greatly improved upon where there some interest to do so.)

The ViewPlus Tiger embosser combines dot height adjustments (it's like Braille) in 8 heights discernable as distinct by fingers (3 bits per dot.) The dot density is 400 dot/in^2 and so may arguably deliver 1,200 bit/in^2. At that 160 sq inch size, that's 24k byte, and the reader is a human finger! Again, an order of magnitude over you figure and from another existing product already out there in the field.

But besides flogging a dead horse, it's also different horses for courses. So I'm just writing this to be argumentative. It's better not to speak in such extremes, is all. Your non-sequitur point remains.

Jon

Reply to
Jon Kirwan

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.