processed ceramic processors

[Sucho]

i have few kilos of AR processed ceramic processors ( only ceramic matrix, without DIEs, no visible metal)

is it worth to mill ceramics to very fine mesh and process it ?
are there fine gold wires inside the ceramics ?
how much gold should i expect ?

 

[samuel-a]

i have few kilos of AR processed ceramic processors ( only ceramic matrix, without DIEs, no visible metal)

is it worth to mill ceramics to very fine mesh and process it ?
are there fine gold wires inside the ceramics ?
how much gold should i expect ?

I suggest you grab one of each type in that batch, brake it to 4-5 pieces and have a look inside the ceramics.
in black square IC's it looks like that:


but this is a sintered matrix which allows that kind of configuration of leads to the core.

on this kind of solid ceramic CPU's, i personally didn't find any gold wires inside, but i always check anyway:



sorry for the poor quality of the pictures, taken by cell cam.

Even if you do find, i can't say that it will really be worth your while pulverizing them for the fine wire, unless you add them to other batch that is already on it way to the mill...

GOOD LUCK
Sam

 

[goldsilverpro]

Sam,

Are you sure that the diameter of those wires is 20-30 microinches? At 30 microinches (.75 microns), it would take about 8,000 ft to equal a dollar's worth of gold, assuming my math is right. Also, I don't think the technology is available to draw gold wire that fine. Where did you get those numbers?

Is the wire round or flat?

Chris

 

[Sucho]

its OK.
i know that in black square IC`s are fine gold wires, i have processed them

i have crushed a lot of ceramic CPUs and was looking for wires everytime but with no sucess

there were no visible ways that connects DIE with pins throw the ceramics

what kind of metal is used to connect DIE with pins and why are there no visible connections ?

 

[samuel-a]

Chris,

This is a manufacturers data, at numerous sources they state 30 μin of round wire 24K.
They also state the length of the wires, resistance etc'...
Remember that a gold wire need to be soldered as well, long time now that I'm pondering myself about that, most likely half of the yield of those IC's are from the solder points.

Another thing, and this is only my opinion, regarding your comment about pulling such thin wires, i believe that they are not fully disclosed about the makeup of the gold and it's most likely that the gold is alloyed at the vicinity of 0.2-05% with other strengthening elements. This may very well considered secret of the trade. Again my opinion only.

A good night walk down the Google lane will turn up some interesting products spreadsheets, that's what i do. (and never remember where it was in case I'll need to refer to it again, like now)

 

[lazersteve]

This topic has is already been discussed several times before, search the forum for 'ceramic wires'.

Steve

 

[Sucho]

i am asking, because i want to test it with a hot AR continual pressed throw that fine ceramic mesh but i dont have time because of exams now

was it tested on a large batch ?

 

[Barren Realms 007]

i am asking, because i want to test it with a hot AR continual pressed throw that fine ceramic mesh but i dont have time because of exams now

was it tested on a large batch ?

You will need to crush them into powder and run the powder in AR to do your test.

 

[Sucho]

i know how to do that, but i want to know if its worth (how much gold in kilo)

 

[Barren Realms 007]

That is something you will need to do because what you come up with and what others come up with might be different on yield's.

 

[patnor1011]

That question cant be answered due to zillions of types produced. You can not put together representative batch as nobody else may get what you have in your own one.

 

[Sucho]

OK, i will test it after exams...i will post my results for various types of ceramic processors

 

[Barren Realms 007]

OK, i will test it after exams...i will post my results for various types of ceramic processors

A test of a single chip will not represent what you might get on average. You should pick a larger number with different date codes spread over a period of time of mfg. I wouldn't do less then 100 for a test.

 

[Sucho]

i know

i would test 100 pieces of processed ceramic processors of each type(486s,386s, P1s etc)

 

[Barren Realms 007]

Yes, it just takes time an patience.

 

[goldsilverpro]

That question cant be answered due to zillions of types produced. You can not put together representative batch as nobody else may get what you have in your own one.

You're right. Every batch is a different deal. Different manufacturers. Different people. Different controls. What you get, you get. A pure gold bar, that you can weigh, is the only absolute proof of the value of the material. What you get, you get. Then, go on to the next project. Don't try to squeeze that last little bit. That last few percentage bit can cost more to get than it's worth. That last 3% can cost more to get than the previous 97%. Sometimes, it pays to only go for the bulk. You'll never get 100%, no matter how good you are. No one has ever gotten 100%, even our illustrious moderators. Close, but no cigar.

 

[Sucho]

i know goldsilverpro

but i am very curious young chemist still thinking about how i can manage that , how i can improve that process , need i another preprocess, which chemicals will be the best etc etc..

 

[Anonymous]

i know goldsilverpro

but i am very curious young chemist still thinking about how i can manage that , how i can improve that process , need i another preprocess, which chemicals will be the best etc etc..

I understand that you are curious and probably very excited about what it is you are trying to accomplish,however I need to give you some really good advice....
When moderators,such as GoldSilverPro and Lazersteve,take the time to try to help you understand something,that I do not believe you understand,you should not come back with a response like "I know".No offense but if you knew,you would not have asked in the first place.And what they,and everyone else on this thread so far,is trying to explain to you,is that it is virtually impossible to give you an accurate answer to your question.As Barrenrealms stated,something as insignificant as a different date code can give you a very different yield.
I have yield data from several loads of chips I have ran through my ball mill and processed,but I only use them as a reference when I am purchasing chips so I have an idea of how much to spend for the chips and still make a little profit.There are too many chips and variables for me to go into my yields on here.
If you have a mass amount of one(or more) particular type of chip,then grind up about 100 of them to 100+mesh minimum and run them in a very weak AR.Then make a decision on what to do with the rest of them based on your results.

 

[texan]

Sam,

Are you sure that the diameter of those wires is 20-30 microinches? At 30 microinches (.75 microns), it would take about 8,000 ft to equal a dollar's worth of gold, assuming my math is right. Also, I don't think the technology is available to draw gold wire that fine. Where did you get those numbers?

Is the wire round or flat?

Chris

An article I just read about bonding wires says the gold wire is 12.5 microns and up. Another article says that some in the ic packaging industry are trying to develop techniques to use copper wires due to the rise in the price of gold.

Texan

 

[samuel-a]

I can't find those spread sheets i was talking about...

This thread kinda "forced" me to do further research about gold wire bonding and the more i look into it, the more convinced I'm that my initial statements should have been in microns rather than micro inches... I'd hate to spread disinformation.



I've done the calculation in microns, one foot of 30 microns thick gold wire equals to about 1.8 $ at today's prices.
20 micron will be 0.8 $ a foot (give or take)
15 micron will be 0.46$ a foot (give or take)