What to do with loose BGA/LGA IC's?

[redshift]

I have a whole bunch of loose BGA and LGA IC's that are old(2003-2008 maybe), not in trays, and have not been stored in a proper environment in terms of static and moisture. What should I do with them?

I have at least a hundred of each of these:

TMS320C64I2GNZ
MPC5554MZP132
XPC860ENCZP50C1

and a bunch of other little BGA's.

They were being thrown away and I "saved" them but now I'm wondering if they have any value. There is a high cost for prototyping with any of them(surely each requires a 4+ layer PCB) and why would anyone take the risk? Could I sell them as a lot? Make key chains out of them?

I welcome anyone's opinion on what I should do with them.

[alsetalokin4017]

Embed the whole lot in clear casting plastic, for a nice paperweight.

Or just toss them in the hazardous waste bin and forget about them, move on with your life.


Of course some obsolete chip sources list these for outrageous prices per each, like 48 dollars each -- so I can see why you might want to sell them. You could always list them on Ebay and see if you get any hits, or contact some of the surplus resellers and see if they want to buy. But since yours have not been stored properly... chances are slim.

[calexanian]

+1 for casting them in clear epoxy and making some cool drink coasters.

[Ian.M]

If you are feeling really brave, 'dead bug' the TI TMS320C6412 Fixed-Point Digital Signal Processor.  Of course its going to be one stone bitch to run the power rails it needs with 'tombstone' decoupling caps linked up by a ground grid as its core can be run at 500MHz or faster depending on speed grade, with an external clock input of up to 33.3MHz.

There are some hints that there may be GCC support for it - check out toolchain possibilities *BEFORE* spending a week slaving over it with a hot soldering iron.

[bitslice]

I've got a TMS320 dev board (somewhere) that I was going to have a play with, and yesterday I was looking around to see how much a tray full of TMS32010's were these days.
So given that I was still looking for this kind of thing, I'd imagine there may be a niche market for the TMS320C64 if they were cheap enough.
But in Canada, unlikely to be worth your while.

[redshift]

I don't think I'm brave or interested enough to deadbug solder any of this. But if anyone here wants to try and is willing to pay shipping, I'd gladly send a pile of them. Maybe they could be useful to someone with an appropriate socket? Or maybe someone would like to try decapping them?

[NiHaoMike]

Sell them for cheap to artists wanting to make "electronic" art?

Or use them to tile (part of?) a workbench. After all, they withstand soldering temperatures quite well.

[uncle_bob]

Hi

Plant them in the garden and give them lots of sun and water. Who knows ... they may sprout new IC's 

Unless you *really* want to live dangerously or are going to spend a fortune on them, there is no practical use for the parts.

Bob

[bitslice]

Sell them for cheap to artists wanting to make "electronic" art?

Oh God, don't encourage the jobless morons to make any more crap

"Oooo, I made a pile of stuff I don't understand and added flashing LEDs, just look how it portrays a dystopian nightmare of evil technology working against humanity to destroy the trees!...."

[molotov]

Put them ebay and make a few bucks.

[bitslice]

There is a high cost for prototyping with any of them(surely each requires a 4+ layer PCB)

Build something, a 4 layer PCB in 100 off, $2.40 each.

A DSP and a microcontroller, that's a building block for a lot of nice filtering things, especially as people look at DSPs and are put off at the idea of trying to get one working.

[AndyC_772]

people look at DSPs and are put off at the idea of trying to get one working.

...with good reason.

Getting one of these up and running is a job even when you have known good parts, a reliable manufacturing process, a Code Composer Studio licence and the skills in PSU design and PCB layout to make a board that has a hope of working.

The C64 architecture - especially older parts - runs hot and doesn't perform well unless you can make good use of the multiple execution units to run a lot of things in parallel. Typically that means writing critical code in assembler.

On the plus side, if you can learn to program one of these, you can pretty much command your price. In fact, I'd quite like to hear from you myself.