Non-Chemical To grind or not to grind....

[youngleo]

Act three, scene one...

Is it better to grind your ceramic ICs, SPUs, CPU, etc. or is it better to not grind them? Does the work involved yield more for the effort? Does it yield less than keeping them together. What is the real deal here?

I have figured out a way to do it very easily in minutes, I am talking powder here. Cups of ceramic purple processor/gold plated stuff/silicon powder.

Well, as I promote this discussion. I understand that it has been talked about but we are divided on it. Moderators, gurus, and newbies alike. I have read tons of threads on this issue and I have to say that I am still confused.

yours truly,

The greenest newbie of them all.

youngleo

 

[lazersteve]

Leo,

Personally I don't grind, or even break apart for that matter, my ceramic cpus.

Mainly because I don't have a 'quick and easy' way to grind them to a fine powder.

Secondly, because filtering the ground material is a real pain.

As for the affect grinding has on yields I would guess it all has to do with which type of cpu you are talking about. For early ceramic cpus I'm on the fence still as to how much of an increase, if any, will be realized. I'm fairly certain you won't see and increase for ceramic K-6 series cpus.

For early black MMX and Celeron, PII's and PIII's (the slotted and fiber types), I'm 100% certain grinding will increase the yield because these types have the black epoxy area in the bottom center of the cpu has a myriad of hair fine gold wires in it

For all the newer ones (FCPGA, P4, and Athlon) I'm also fairly certain grinding won't increase the yields.

If you have a 'quick and easy' home brew method of grinding ceramic cpus, I for one would be very interested in how you are doing it. If you choose to disclose your methods to the forum, I'll gladly use it to grind several cores I have on hand of various cpu types and let everyone know what kind of yields I get.

I've ground some 486 cores in the past using a pipe crusher (too much work) and I was not able to produce gold from the cores. This did not discourage me as the pieces were fairly large (1/16 to 1/8") and I feel you must get below 150 mesh to get to the fine gold wires if they indeed exist.

One thing I can honestly say is that I have looked at many fractured ceramic cpu cores and I have never found a single gold wire, at any magnification, protruding or embedded in the ceramic substrate. The opposite is true of the epoxy cores. I have found wires in them.

I think some people have ground cpus that had the cpu die (the shiny mosaic piece in the middle with the circuit on it) intact and gotten gold from the braze under this die, possibly mistakenly attributing the additional yield to 'invisible wire' in the ceramic substrate. For this reason I always remove the die from the ceramic housing to get at this 'hidden gold'.

Steve

 

[Barren Realms 007]

Steve could it be that you need a bigger pipe to hold onto, something with some weight to it that you can wrap your hands around? 8)

 

[lazersteve]

Steve could it be that you need a bigger pipe to hold onto, something with some weight to it that you can wrap your hands around? 8)

It's not the size of the pipe, it's the fact that it wears my arms out after only a few cycles. A bigger pipe seems like it would just wear me out quicker.

Steve

 

[Barren Realms 007]

Steve could it be that you need a bigger pipe to hold onto, something with some weight to it that you can wrap your hands around? 8)

It's not the size of the pipe, it's the fact that it wears my arms out after only a few cycles. A bigger pipe seems like it would just wear me out quicker.

Steve

You are correct you will wear out quicker but you get more reaction from your actions with less work in the long run. Put a little lead in your pipe to help out.

 

[Anonymous]

Steve, I have a pipe set up also, I use my air chisel to drive the smaller crusher pipe, it works pretty good.

I baught a spare round punch type chisel and welded it to the hammer pipe.
Run at 95 psi, and it crushes stuff pretty fast.

edit,
I also use a larger bore outer pipe 2 inches with a 3/4 sized pipe and cap for the hammer.
I think 3 inch would be even better, but, have not tried it yet.

Jim

 

[goldsilverpro]

On the old ceramic cpus (depends on what you define as old - I'm thinking old side-braze), From my experience, it's only necessary to break the chip in several pieces when using AR. If you just remove the lid and leave the part whole, the AR will eventually leach the gold under the 4 sides of the chip, but it will be slow. Even with a lot of heat, it can take 6-8 hours. When you can slide the chip around, you know all of the gold braze is dissolved. If you break them up in 4 or 5 pieces, which opens the lid also (and doesn't take much effort), the chip has more edge area for the AR to penetrate and it will take much less time. In any case, with enough time, all the gold will dissolve.

The same thing applies to some of the older large square purple CPUs with gold lids. At one place I worked, in about 2003, they leached these whole, with lids, about a half drum at a time, in about a 65% AR solution (52HCL/13HNO3/35H2O). I am thinking that they heated the solution a bit (but, I may be wrong - it may have started at room temp.) with quartz immersion heaters. It took about 2 days before you could slide the chip around. With gold braze under the chip, the key is to leach until you can freely slide the chip around. The chip is the last thing to go, after all the metal has dissolved. The gold was often dropped with SO2 but sometimes they used SMB.

 

[Scott2357]

For those who wish to go the crushing/grinding route, I think you'll see better results using thermal shock first. Heat the ceramic to a minimum 350-400' C and then quench in cold water. You can use a torch for singles or the BBQ grill for batches. Just make sure you have fresh cold water at regular intervals as it tends warm as you process more and more chips. Even though there is usually not any visable difference, this micro fractures the ceramic frit making it brittle, therefore easier to break. It becomes more like glued sand instead of glass so when broken it makes more crumbled pieces instead of shards. Thick packages may require higher heat or additional shocks to get the job done.

 

[Palladium]

It's not the size of the pipe, it's the fact that it wears my arms out after only a few cycles. A bigger pipe seems like it would just wear me out quicker.

Steve

Astroglide ! :shock: :shock:

 

[goldsilverpro]

For those who wish to go the crushing/grinding route, I think you'll see better results using thermal shock first. Heat the ceramic to a minimum 350-400' C and then quench in cold water. You can use a torch for singles or the BBQ grill for batches. Just make sure you have fresh cold water at regular intervals as it tends warm as you process more and more chips. Even though there is usually not any visable difference, this micro fractures the ceramic frit making it brittle, therefore easier to break. It becomes more like glued sand instead of glass so when broken it makes more crumbled pieces instead of shards. Thick packages may require higher heat or additional shocks to get the job done

.

I've mentioned this a couple of times on the forum. The primary ceramic used is powdered alumina. The alumina is held together by adding 5% glass powder (frit) to it. This greenware is formed and put through a furnace and the glass melts and holds everything together. When you heat to about 700-800F and quench in water, the glass frit fractures and weakens the ceramic.

 

[Scott2357]

For those who wish to go the crushing/grinding route, I think you'll see better results using thermal shock first. Heat the ceramic to a minimum 350-400' C and then quench in cold water. You can use a torch for singles or the BBQ grill for batches. Just make sure you have fresh cold water at regular intervals as it tends warm as you process more and more chips. Even though there is usually not any visable difference, this micro fractures the ceramic frit making it brittle, therefore easier to break. It becomes more like glued sand instead of glass so when broken it makes more crumbled pieces instead of shards. Thick packages may require higher heat or additional shocks to get the job done

I've mentioned this a couple of times on the forum. The primary ceramic used is powdered alumina. The alumina is held together by adding 5% glass powder (frit) to it. This greenware is formed and put through a furnace and the glass melts and holds everything together. When you heat to about 700-800F and quench in water, the glass frit fractures and weakens the ceramic.

I do see mentions of this now that I searched for it but don't remember seeing it before. It must have been posted when I wasn't keeping current with the forum. While looking through the search results, I noticed your mention of thick film resistors, etc. on ceramic substrates which I had not seen either. This is good info, I have a few little cards like that. I'll try to post pics soon.

 

[lazersteve]

I remove the bottom lids with a MAPP torch and a good tap. Sometime the ceramic will fracture just from the heat of the torch, especially if the weather is a little cooler out.

This could mean that if you cool the ceramic first, then heat you may get a good fracture.

I'm still not convinced the ceramic substrate holds any hidden gold aside from the side braze and the gold hidden under the core.

Steve

 

[hemicuda]

Hello all,
I have seen a few postings here on the forum from Gustavus in regards to crushing,actually more along the lines of ball milling.
He has done a large amount of material in his mill ( all of which I am assuming had been ran through an AR bath prior to an incineration )
and he was rewarded with a decent return from the mill.
So if one does indeed have enough material to do, in my eyes, I would feel it would be worthwhile to do.

For more info on his returns search for either Gustavus or the Ball Mill in the search section.
Cheers all,and have a good day!
Keith.

 

[goldsilverpro]

I have more experience with all gold side-braze packages than with any other type. They are pictured in the bottom photo on this link.
http://www.chelseatech.com/packages.htm

These are made in 3 ceramic layers. The middle layer contains all of the traces that connect the legs to the inside bonding fingers. These traces are made by silk-screening a moly-manganese paste and then firing it. The paste and the ceramic both contain a small amount of glass frit which melts and bonds the traces to the ceramic. The traces extend to moly-manganese pads on the sides where the legs are attached with silver solder. The inside bonding fingers are also moly-manganese. Only when the 3 layers have been bonded together by firing, is the part nickel plated and then gold plated. In the photo, 2 tie bars are still attached to the legs. Contact for plating is made to all portions of the package, inside and out, though the tie bars. These are trimmed off once the chip and gold lid are attached with various gold brazes. Also, the notch on one end of the package is ground away, which severs the connection to the lid ring. The part is then tested.

What I'm saying is that, once the lid is removed, all of the gold contained in the part is visible to the eye, except for what is under the chip. There is no gold between the ceramic layers. Complete grinding of the part is, therefore, a big waste of effort. However, the dissolving of the gold under the chip is sped up by breaking at least the portion containing the chip into several pieces. It takes very little effort to do this. Anything more is a wasted effort.

The all ceramic packages as in the 1st photo are essentially made the same way. There is simply no technical reason for gold being in between the ceramic layers.

 

[goldsilverpro]

On several common types of all ceramic CPU packages, the most difficult part to dissolve is the gold under the chip, and, that's where a good part of the gold is located.

Say you have a 1/4" square chip in the package. The AR can only penetrate under the chip from the edges, which total 1" in edge length. In order for all of the gold under the chip to dissolve and for the chip to be easily slid around, the acid from each edge has to penetrate to the center, a distance of 1/8". That's the same as dissolving a piece of gold 1/8" thick, only worse. To maintain it's efficiency, there has to be constant exchange of acid under the chip. Spent acid has to leave and fresh acid has to enter. Otherwise, the reaction will stop. Solution exchange rate is assisted by a higher temperature, strong aqua regia, and the gassing of the reaction itself. For whole chips, this might take 1 to 2 days.

Break the chip into 4 equal pieces. Now you have 2" edge length (twice as much) and only 1/16" distance to penetrate (half the distance). So, by breaking it into 4 pieces, you cut the time by a factor of 4, theoretically. With irregular pieces, the largest one would determine the time.

Why Oh Why would they put gold between the ceramic sections in these particular parts? Besides the added difficulties and much greater costs of using gold here, the Mo/Mn works great and it's cheap. The last time I ran drum loads of CPU packages was in about 2003. At that time, I don't think that any CPU package had any gold encased in the ceramic. I never saw one, at least, and I think I saw 99% of what was out there. If there is no gold in the ceramic, explain the logic of grinding these to powder!! It would go faster only because the chip is pulverized but this wouldn't outweigh the problems (labor, costs, time) involved. You can break up the part sufficiently with one swing of a hammer.

I have seen a few postings here on the forum from Gustavus in regards to crushing,actually more along the lines of ball milling.
He has done a large amount of material in his mill ( all of which I am assuming had been ran through an AR bath prior to an incineration )
and he was rewarded with a decent return from the mill.
So if one does indeed have enough material to do, in my eyes, I would feel it would be worthwhile to do.

I haven't yet re-read Gustavus' posts on this but will. I would ask him how he knows that he gets more gold with the ball mill than without it? On the first go round, maybe he wasn't getting the AR hot enough and maybe he stopped before he could easily slide the chips (only works when 100% of the gold under the chip has been dissolved).

 

[hemicuda]

Hello all
I tried to find all of the old posts from Gustavus on his Ball Mill and I was greeted with ..... so I am stumped,again just an assumption that he has deleted the old posts.

I have sent a P.M. to Noxx in regards to finding the old posts.

I do know that he found value in milling the material, reason being stated is that he was trying to sell the material to another forum member, but was declined.
So Gustavus opted to mill the material himself and refine what he had milled with good returns.

I am not going to be a dog chasing my tail in this subject......but if there is a pin on the chip and a brain in the middle of the chip there would have to be some sort of communication between brain and pin via Gold or some other type of PM.....again it might not be a HUGE amount of pm but after a while it all does add up.

Again just my 2 bits worth...lol

 

[qst42know]

If I'm not mistaken Gustavus was grinding black epoxy chips not ceramic CPUs.

What were you grinding in your ball mill Gustavus?

 

[lazersteve]

I'm pretty certain it was black plastic ics for the most part.

The post can be found using the Forum Search tool on my webs site via my signature line below. The built in forum search engine ignores common words which excludes most of the desired results.

Steve

 

[Barren Realms 007]

http://www.goldrefiningforum.com/phpBB3/viewtopic.php?f=40&t=6467

 

[leavemealone]

Yes they were plastic and ceramic IC's.And yes he deleted all of the posts,including the pictures of the mill a couple of days ago.I have them here on my computer somewhere.Gill has been talking about leaving the forum for a while,maybe he did it this time.I will try to get him and see if he minds if I repost the pics(that is if I can find them).
Johnny
**EDIT**
I don't think he has gone anywhere since he posted a response yesterday,though he deleted those pics on the 3rd.Im sure he had his own reason for deleting all of those posts.