DOes the IC Manufacturer make a difference in Yield

[TheDragonWins]

Just curious but do the different brand of IC chips make much difference in the total yield of PM?

Say you process a SN74 from Motorola vs SN74 from AMD?

 

[resabed01]

Probably not. There are too many variables to say yes or no. Each manufacturer can make the same chip in different packaging. The most simple example of this is the same chip in a commercial package vs military packaging made by the same company. Ceramic vs plastic, etc. Age plays into this as well, 74 series logic from the 70's may contain more PMs than it's modern equivalent.
I don't think different manufacturers plays a big role in determining PM content as they all use similar manufacturing techniques. Maybe the exception to this would be for chips made in Russia.

 

[goldsilverpro]

The PM value is mainly determined by the thickness of the gold plating, which can vary from manufacturer to manufacturer. Maybe not much, but every manufacturer has it's own ideas on what is the thickness that will give the best balance between plating costs and the desired characteristics of the plating. This is determined by the designers and engineers working for the particular company.

 

[macfixer01]

I have to say yes, especially so with 7400 series TTL chips. I've actually commented on this a few times in previous threads. With RCA chips for example identically part numbered, plastic packaged, and labeled chips except having different date codes, may have visible gold plating inside one chip but not in the other. With any manufacturer, chips made at plants in different countries, or in different years, or with different part number variants, may or may not have visible gold inside. You can generally be pretty certain though that identical brand/part#/date-coded chips on the same board will all be made the same way. A second possible source of gold is the tiny bonding wires which could also be just aluminum. As resabed01 mentioned, another type of IC's are TTL chips in ceramic packages made for high temperature or military applications. Those may or may not have a spot of gold on the bottom ceramic layer which the chip die itself is bonded to. If not gold then the spot will be silver-colored and I believe is likely just aluminum. There is a link to a picture inside one of the threads linked below that shows this.

Between different manufacturers there is no consistency. Except to say in my experience certain brands like Texas Instruments are very unlikely to ever have any visible gold inside if you crack one open. Certain other brands like OKI, Motorola, AMD, RCA are more likely to have visible gold inside but it's never a guarantee. I couldn't easily find what I was looking for, but here are links to a couple other threads you may be interested in:

http://goldrefiningforum.com/phpBB3/viewtopic.php?f=60&t=20411

http://goldrefiningforum.com/phpBB3/viewtopic.php?f=34&t=12902

Macfixer01

 

[Findm-Keepm]

I have to say yes, especially so with 7400 series TTL chips. I've actually commented on this a few times in previous threads. With RCA chips for example identically part numbered, plastic packaged, and labeled chips except having different date codes, may have visible gold plating inside one chip but not in the other. With any manufacturer, chips made at plants in different countries, or in different years, or with different part number variants, may or may not have visible gold inside. You can generally be pretty certain though that identical brand/part#/date-coded chips on the same board will all be made the same way. A second possible source of gold is the tiny bonding wires which could also be just aluminum. As resabed01 mentioned, another type of IC's are TTL chips in ceramic packages made for high temperature or military applications. Those may or may not have a spot of gold on the bottom ceramic layer which the chip die itself is bonded to. If not gold then the spot will be silver-colored and I believe is likely just aluminum. There is a link to a picture inside one of the threads linked below that shows this.

Between different manufacturers there is no consistency. Except to say in my experience certain brands like Texas Instruments are very unlikely to ever have any visible gold inside if you crack one open. Certain other brands like OKI, Motorola, AMD, RCA are more likely to have visible gold inside but it's never a guarantee. I couldn't easily find what I was looking for, but here are links to a couple other threads you may be interested in:

http://goldrefiningforum.com/phpBB3/viewtopic.php?f=60&t=20411

http://goldrefiningforum.com/phpBB3/viewtopic.php?f=34&t=12902

Macfixer01

With RCA CMOS (they called the process COS/MOS) and Linear chips, always try to get the epoxy 14/16-pin DIP packages for processing - they used gold bonding wires exclusively in their plastic packages - all the ceramic packages contained aluminum bonding wire. I'll post an excerpt from an RCA app note detailing their encapsulation process later today - sizes, max bonding wire length and type weld, if I remember correctly. RCA later became GE/RCA, then Harris, and now part of Intersil.

I collect the databooks and app notes for just about all manufacturers through the 90s, just to determine the bond wire choices and any other PM-bearing info. Sylvania used some seriously large bond wires in their process - 57uM was their largest, in some RTL/DTL chips. Intersil (pre-Harris) also used Au exclusively for all their IC_#### series stuff. Burr-Brown and Datel used mostly Au as well in the larger (2x2) module-style ADCs and Converters.

If only I could find info on Hughes Hybrids - I've got a few dozen that have some nice plating, but I'd sure like to know how thick. There were none made for commercial use, so data is like hens teeth.

Cheers,

 

[macfixer01]

I have to say yes, especially so with 7400 series TTL chips. I've actually commented on this a few times in previous threads. With RCA chips for example identically part numbered, plastic packaged, and labeled chips except having different date codes, may have visible gold plating inside one chip but not in the other. With any manufacturer, chips made at plants in different countries, or in different years, or with different part number variants, may or may not have visible gold inside. You can generally be pretty certain though that identical brand/part#/date-coded chips on the same board will all be made the same way. A second possible source of gold is the tiny bonding wires which could also be just aluminum. As resabed01 mentioned, another type of IC's are TTL chips in ceramic packages made for high temperature or military applications. Those may or may not have a spot of gold on the bottom ceramic layer which the chip die itself is bonded to. If not gold then the spot will be silver-colored and I believe is likely just aluminum. There is a link to a picture inside one of the threads linked below that shows this.

Between different manufacturers there is no consistency. Except to say in my experience certain brands like Texas Instruments are very unlikely to ever have any visible gold inside if you crack one open. Certain other brands like OKI, Motorola, AMD, RCA are more likely to have visible gold inside but it's never a guarantee. I couldn't easily find what I was looking for, but here are links to a couple other threads you may be interested in:

http://goldrefiningforum.com/phpBB3/viewtopic.php?f=60&t=20411

http://goldrefiningforum.com/phpBB3/viewtopic.php?f=34&t=12902

Macfixer01

With RCA CMOS (they called the process COS/MOS) and Linear chips, always try to get the epoxy 14/16-pin DIP packages for processing - they used gold bonding wires exclusively in their plastic packages - all the ceramic packages contained aluminum bonding wire. I'll post an excerpt from an RCA app note detailing their encapsulation process later today - sizes, max bonding wire length and type weld, if I remember correctly. RCA later became GE/RCA, then Harris, and now part of Intersil.

I collect the databooks and app notes for just about all manufacturers through the 90s, just to determine the bond wire choices and any other PM-bearing info. Sylvania used some seriously large bond wires in their process - 57uM was their largest, in some RTL/DTL chips. Intersil (pre-Harris) also used Au exclusively for all their IC_#### series stuff. Burr-Brown and Datel used mostly Au as well in the larger (2x2) module-style ADCs and Converters.

If only I could find info on Hughes Hybrids - I've got a few dozen that have some nice plating, but I'd sure like to know how thick. There were none made for commercial use, so data is like hens teeth.

Cheers,

Yes both DAC's and ADC's quite often are gold rich, outside and inside. It's probably something to do with the need for speed, stability, and accuracy of conversion? As far as the TTL series chips I was referring mainly to visible gold plating on the internal pin ends and on the metal strip the silicon die is bonded to. It's quite likely from that point the bonding wires are gold also, but if you crack the package open the wires aren't often visible and are likely left embedded in the epoxy. I crack one of each chip type on a board open to separate them and keep the better quality chips together. Once I see one is gold inside I also grab the others which are the identical brand, part#, and date code. I should have mentioned you'll occasionally see TTL chips in an identically shaped but grey-colored epoxy packaging that (unlike the black epoxy) usually seems crumbly and crushes fairly easily. These can be different brands, but quite often have visible gold inside. I don't know why, maybe it has something to do with manufacturing standards during the timeframe when that grey epoxy was still used?

macfixer01

 

[Ahmed Lotfy]

What I know from refiners here in Egypt for Example the North , South bridges IC ceramic chips which made by Intel have 7:8 Grams gold content per Kilo

while the avarage of other types of many diffrant manifacturers have 5:6 grams per Kilo

this is such example

 

[macfixer01]

What I know from refiners here in Egypt for Example the North , South bridges IC ceramic chips which made by Intel have 7:8 Grams gold content per Kilo

while the avarage of other types of many diffrant manifacturers have 5:6 grams per Kilo

this is such example

I don't specifically recall if your numbers are correct, but they could be? You have to realize though, they weigh so little that it takes a lot of them to make up a kilo. Those north/south bridge chips have been discussed extensively on the forum before so you could search here for more info. In any case, they are in a whole different class of scrap. The original question asked about 7400 series TTL chips which are generally going to be fairly low yield.

Macfixer01

 

[TheDragonWins]

Thanks for the answer to my questions. I have about 50 lbs of boards with literally 4 different manufacturers of the chips on each board so was just curious. Dont know what to do with them yet.