Do you have military grade chips in your Amiga?

[Floid]

DonnyEMU wrote:
LOL I have an old Amiga 1000 myself and i am not sure that what you are describing constitutes millitary grade chips, from what I understand from the Amiga launch in NYC, any chips in original A1000s are there because C=MOS didn't make them.

Someone wanted pictures of the *supposed* "mil-spec" chips; well, here's the girls (and Gary?  My mind is going)... in our own archive.  http://amiga.org/gallery/photo.php?lid=908.

Looks like it might say MOS up top, though that doesn't mean much, since foundries seem to have always let the client's brand take precedence.  (As you can see in the AGA machines with the HP chips, at some point Commodore decided not to care.)

Please someone tell me if I am wrong. I can tell you this I don't believe these were millitary "grade" chips because I saw a lot of Amiga 1000's with their CIA chips go bad.. Mostly because people did stupid things like plugging things in and out when they shouldn't have.

Y'know.. it's been a long time since I've had to care, but I could swear "MIL-SPEC" is self-audited somehow, or just applies more to paperwork-shuffling than any particular capabilities (after all, the military has to certify practically every nut and bolt, which basically just means they need a degree of documentation on it)...  I'm sure there's plenty of "MIL-SPEC" hardware that can still be killed by shorting a connector improperly, though any regulations around the whole program probably try to prevent that.

http://www.dsp.dla.mil/ seems to be the group responsible for... everything related, though it's not like they have an obvious FAQ for what "MIL-SPEC" (versus any of the thousand other terms, specific certifications, etc) means.

Anyone who's been in the .mil want to hook us up?  Is it as BS/general a term as I think it is?

[Floid]

Floid wrote:
Anyone who's been in the .mil want to hook us up?  Is it as BS/general a term as I think it is?

Er, and implied from there, that Moto and Commodore could easily have had it on *any* chips they produced, whatever the packaging, radiation-hardness, whatever?  (The military has plenty of regular *offices,* I'm sure there also exist 'MIL-SPEC' pencils and staple removers...)

There are certifications for security, combat-hardness, and so forth, but those are *specific* certifications, with confusing names and numbers of their own.  And I'm sure even the stock 'consumer-grade' 680x0s must've met a few of them... Y'know -- and I'm just pulling this out of my butt here -- "DoD Standard 12345678-XY; Requirements for Complementary Metal Oxide Semiconductor Devices for Unclassified Document Processing Systems" or something.  :-D

Then things, 'certified' or not, get standard requisitioning product codes, for classification at auction and such...  From probing the .mil auction sites a while back, I know they have them for things like "Bucket, unspecified, plastic," or even "Cow."

[Floid]

levelLORD wrote:
OK, I took a quick look for the first bidder and he's claiming this: 'RARE Ceramic MOS 8361- Agnus in AMIGA Systems'

After that, I searched the web and find out that MOS 8361R3 is not an Agnus, actually.

Sheesh.  As we should all know, Agnus went through only, what, a billion revisions?  A quick Google for "MOS 8361" (not in quotes when I threw it in the engine, should you care to replicate), shows that the 8361 did ship in some 1000s -- it was doubtless one of the original, "thin" iterations.

[Floid]

DonnyEMU wrote:
The pictures you are showing me are frauds. I still own an original A1000 with a 1985 manufacturing date..This isn't millitary spec chips, these pictures you are showing me aren't even  what the real chips look like, they are advertising fakery. I have a stack of old Amiga Worlds with that same ad. As a dealer/repair center from 85-92 I saw a lot of these machines.

For further proof: anyone care to  count the number of pins on Agnes,Denise, and Paula then count the number of pins on those chips in the AD and get back to me. It sure makes a cool looking ad though.. The plate over the chips in this ad is a simple 79 cent glue on from MOS.. I have seen the innards of plenty A1000s to know what I am looking at.

http://amiga.org/gallery/images/896/1_1183.jpg
Hm, someone must've gone through some trouble, then.  48-pin DIPs in both shots.  Unless I can't count.  And you can see here that they got the 8361 (that also doesn't exist?) into a regular epoxy package.

The 'heatspreaders' are obviously just cheap caps of some sort - I've found some other pictures where they weren't on straight.  But IIRC, with this sort of package, the chip die was otherwise exposed (I dunno why; because they encased the pins in the ceramic before they dropped the chip in and connected the cat-whiskers? -- Were chip packages coming from separate plants than chips themselves? -- Or because the same package would have a use for UV EEPROM, those would need a window there anyway, and it would be cheaper to use caps than design an all-ceramic top with a hollow for the chip?  Or because engineers would enjoy the ability to pop the cap and microscope failure modes on returned units without sanding/etching off the epoxy as they do now?)... So the cap keeps you from, say, jabbing the die with your finger, or getting WD-40 all over it when you decide to clean your board.  (Or was there epoxy under there, anyway?)  Ever dissect one of those old calculators, where the LEDs were exposed dies under a plastic magnifying lens?

One assumes the miracle of 100% plastic/epoxy packaging caught on quickly when it was discovered how much easier it was to just pour goo over everything.  (Of course, bad epoxy still strikes.  All the Fujitsu drive failures of months back were eventually traced to a bad packaging used by Cirrus Logic, to whom they'd outsourced some chips.)

Now why does everyone need to bullshit over something so simple?

[Floid]

whabang wrote:

All the Fujitsu drive failures of months back were eventually traced to a bad packaging used by Cirrus Logic, to whom they'd outsourced some chips.

By the gods, I remember that! :-o
I was working in a technical support department at the time, I must have replaced thousainds of those!!! :-D

Ouch.  It bit me in a way, too - the companies whose engineering I'm most likely to trust are IBM and Fujitsu.  (And Seagate, I suppose, but I was never 100% happy with the 2.1gb Barracudas that I 'overpaid' for right before 4-20gb drives went dirt cheap.)  Well, we know what happened to IBM, and then *bam* - Fujitsu, who were still making drives in the 'trustworthy' 20gb-40gb range (while everyone else seemed to be having growing pains pushing the envelope to 60gb+), had that hit.  In the drive capacities  that I was actually hoping to scrounge off eBay...

Ended up getting a WD 80gb retail, because, y'know, can't trust them, but, *CLICK,* nobody was *CLICK* complaining at the time, and *CLICKCLICK,* the price was right.  I can't say *CLICKTHUNK* I was *CLICKCLICKwhirrr* really happy with that *THUNK* decision *THUNK,* though.  *THUNK.*

I do think some of that was down to the power supply or drive positioning in the box that ended up in, never did get a chance to transplant it and see.

But anyhow, you can imagine *why* Fujitsu shot the food so badly on their recall; by the time they realized what was up - and perhaps that their stock of replacements had the same faulty chips? - they'd already pissed everyone off for not addressing the issue.  Sucks as much for them and their reputation as it does for the consumer.

[Floid]

@mikeymike, @karlos,

mikeymike wrote:
@ Floid

The only advice I give to people thinking about a hard disk purchase is that it's best to buy a lower capacity disk, say half the size of the largest available.  The 'latest and greatest' tend to be the most problematic IMO.  Also, remember that every disk manufacturer has had their own major PR disaster regarding dodgy batches of disks :-)

Heh, I'd never suggest that WD had a stunning track record.  I just happened to fall for the idea that, maybe, finally, they might've sorted some of it out.  Point being that Fujitsu's problem was counter to that rule of thumb - they're in the business of making reliable, conservative drives for reliable, conservative sorts of people, and unfortunately, their defective units were right in that 20-40GB 'safe zone' behind other manufacturers' problematic 60+GB drives.

Don't think we need another drive 'war,' or at least, I already hashed that out in another thread.  Seagates are great if you know what model you're getting, I've just landed some of their more boring/silly designs.  (The 351A/X, staid 40MB? IDE of times past; one ridiculously fast HH - not the normal 'slim' height we expect from 3.5" floppies and modern drives - 540MB in that era, and the 2.1GB 'cudas, which were warm-running early 7200RPM units, and had unimpressive transfer rates versus the suddenly-quarter-of-the-price larger drives with higher density platters.  Now that this Interweb thing is here, I look up the spec sheets before I buy.)