How to estimate the amount of gold in electronics and jewelry

How to estimate the amount of gold in jewelry and electronics scrap.

Karat Gold jewelry:

9 karat gold is 9/24 = 0.375; 37.5% pure gold

10 karat gold is 10/24= 0.417; 41.7% pure gold

12 karat gold is 12/25= 0.500; 50% pure gold

14 karat gold is 14/24= 0.583; 58.3% pure gold

18 karat gold is 18/24= 0.750; 75% pure gold

22 karat gold is 22/24= 0.917; 91.7% pure gold

24 karat gold is 24/24= 1.00; 100% pure gold



Gold filled jewelry and decorative items.

1/10 12 karat = 0.10 time’s total weight of object times the karat percentage.

1/20 14 karat = 0.05 time’s total weight of object times the karat percentage

Example = 1/20 gold filled 12 karat ring that weighs 8 grams.

0.05 x 8 x 0.5 = 0.2 grams of (theoretical) pure gold in ring.

Filled Value x Weight x Karat = Weight pure 24kt gold


Gold rolled (heavy plated) jewelry and decorative items

1/100 to 1/40 total weigh times the karat rating of item.

Example = 1/100 gold rolled (plated) 14 Karat ring that weighs 3 grams.

0.01 x 3 x 0.583 = 0.0175 grams of (theoretical) pure gold in ring.

Rolled Value x Weight x Karat = Weight pure 24kt Gold






Gold plated jewelry and electronics items.

One must first determine the class of gold thickness of the item in question. Once the thickness is obtained or estimated, then determine the square surface area of the item and then there are two methods to figure out approximately how much theoretical pure gold is in the item.

Decorative Applications:

Gold flash 7 micro inches
Gold electroplate 10 micro inches
Gold plate 20 micro inches
Heavy gold plate 100 micro inches

Engineering Applications

Class 00 20 micro inches
Class 0 30 micro inches
Class 1 50 micro inches
Class 2 100 micro inches
Class 3 200 micro inches

Military specifications:

Additional classes

Class 4 300 micro inches
Class 5 500 micro inches
Class 6 1500 micro inches

The formula for determining content of gold in decorative is the same as electronics scrap.

Electronics Circuit Board Fingers

Silverpro’s method:

Daily price of gold times the square inch surface area times the thickness.

Example: Gold prices at $655.00 an ounce. And you have a total of 40 square inches of surface in 50 ea. DIMM 100 pin memory sticks. Estimating the class as class 0, or 30 micro inches thick plating.

Base value figure is 1/1000 Spot price per micro in. X surface area in.

0.655 x (gold price) 40 sq in. x 0.30 (thickness) = $7.86 of gold at today’s prices.

Catfish’s method:

Measure the square surface area of the item and total the aggregate amount of square surface area of all the items and then determine the cubic volume of theoretical gold. This final result will be in cubic inches. Then convert cubic inches to cubic centimeters. Take the result of total cubic centimeters times 19.3 and this will give you the amount of gold in grams.

One cubic centimeter of gold weighs 19.3 grams.
$655.00/31.1 grams/ounce = $21.06 per gram.

Example: Gold prices at $655 an ounce. Same square surface area of 40 inches.

40 square inches times 0.000030 inches =0.0012 cubic inches
0.0012 cubic inches = 0.0197 cc.
0.0197 x 19.3 = 0.380 grams
0.380 x 21.06 = $8.00 worth of gold at today’s prices.

Actual test case on 50 sticks of memory

Test conducted on 7/5/2007 on 50 sticks of DIMM 100 pin memory.

50 sticks of DIMM memory weighs 2.03 lbs
Trimmed fingers weigh 68 grams

The gold plated surface on each finger measures

0.04 in. by 0.10 in. = total 0.004 sq. in. per pad on each finger.

There are 100 pins with both sides gold plated.

0.04 x 0.10 x 100 x 2 =0.8 sq. in. per board x 50 ea = 40 sq in. total surface.

The above mathematical derived results come out to approximately 0.38 grams of pure gold in the 50 circuit boards.

The fingers were refined in Aqua Regia and the result gold was 0.4 grams. It was refined the second time to obtain maximum purity and it came out to 0.3 grams of pure gold.

The actual test results came out very close to the mathematical results. The accuracy of the scale is in tenths of a gram, unable to get a 100th measurement.

Conclusion: The mathematical results matched the actual test results and were well within acceptable limits for the determination of gold cold content of electronics circuit boards.


Actual test case on 152 SCSI Controller circuit board fingers

Test conducted July 7, 2007 on 152 SCSI circuit board fingers.

The gold plated area on each of the fingers was

0.04 in. by 0.18 in. = 0.0072 sq in of surface.

0.0072 x 54 (foils) x 2 (sides) = 0.7776 sq. in. per x 152 fingers = 118.2 sq. in. gold plated surface.

Silverpro’s method

0.655 x 118.2 x 0.30 = $23.23 of gold in fingers

Catfish’s method
118.2 sq. in. x 0.000030 (class 0) =0.0003546 cu. In.

0.003546 cu. In. = 0.0581 cubic centimeters

0.0581 X 19.3 = 1.12 grams of pure gold

1.12 grams at $21.6 = $23.58 worth of gold.

The actual fingers refined in Aqua Regia yielded 1.2 grams the first process. Ran it again through Aqua Regia 2nd time for purity and it yielded 1.1 grams of very pure gold.

The test results were very close to the mathematical results. Well within acceptable limits for determining the gold content in Electronics scrap.

On both test cases:
1. 50 ea. memory sticks.
Silverpro’s estimate was 0.37 grams
Catfish estimate was 0.38 grams
Actual gold was 0.4 grams #1 refining process
Actual gold was 0.3 grams #2 refining process

2. 152 ea. SCSI boards.
Silverpro’s estimate was 1.10 grams
Catfish’s estimate was 1.12 grams
Actual gold was 1.2 grams #1 refining process
Actual gold was 1.1 grams #2 refining process

You can see that the end results are well with in acceptable limits in calculating the gold content in Circuit board fingers or edge connectors.

Conclusion:

After evaluating the two methods of determining the approximate amount of gold in electronic scrap (circuit board edge connectors or fingers) , I recommend that the method that Silver pro uses is much faster, less complex and certainly accurate enough for all practical purposes for estimating the quantity and value of electronics scrap.



Catfish

Acknowledgements’

Many thanks for Lazersteve and Goldsilverpro’s contributions to this docum

Very Informative , I recommend this gets posted to the tutorials.

Ian:

I appreciate your comments, but Lazersteve is working on a tutoral on how to determine the amount of gold in different items, Which I am sure will be much more informative and very professional.

Thanks,

Catfish

I've said it before, and i'll say it again.

That's one smart fish.

Bet you can't get that over at shor. :shock:

Great piece of work Tom. :wink: :wink: :wink:

My method and Catfish's method, of calculating the gold value of plated objects, only vary by about 1.5%. Pretty close.

I hope that I don't confuse everyone but, I would like to alter my method a little bit. I think that my new method will make it easier to understand how it works. Here's the whole thing. The only change is in (3), below.

Why would you want to go through this rigmarole? I use it mainly for buying and selling, when I don't have an assay. I've used it a lot to estimate the value of stuff on Ebay. When dealing scrap, knowledge is king. Although the method is only ballpark, it's better than having no idea of the value.

(1) Make measurements and calculate the total gold plated surface area. You want to know how many square inches or, how many square centimeters of surface you have. You need to break the plated part down into it's geometric shapes.

Fingers are easy because they are rectangles. You just measure the width and length of one finger, multiply these together, and multiply this answer times the total number of fingers. If you have 50 sets of identical fingers, multiply that answer times 50. All this gives the total surface area on these particular 50 sets of identical fingers. For a round thin pin, you multiply the diameter times pi (3.1416) times the length. For a thick round pin, you also have to calculate the area of the end. For a square thin pin, multiply the width of one side times 4 times the length.

Type in - surface area formulas - in Google. Some sites will give drawings of the different geometric shapes and their surface area formulas. I will look for a good one and post it in the Glossary section.

You can measure in several different ways. The cheapest way is to use a triangle shaped plastic architect's ruler. You can get plenty of accuracy using one of these rulers. Get the one that measures in tenths of an inch and learn to estimate reading it in between the lines (marks). Don't ever measure in 1/16ths, 1/8ths, etc. This would make the calculations much more complicated. For small measurements, you can place the ruler on the object and use something like a 10X eye loupe to read it.

My favorite is a cheap set of calipers that measures in a decimal part of an inch and, not in 1/16", 1/32", etc. If you can pay more, you can get one that has a dial for the last digit. The cheaper ones have a vernier and that is a little more difficult to get used to.

Probably, the best is a comparator. This is an eye loupe that has a scale built into it. You can get a variety of screw-on scales for it.

You could use a micrometer but, they are cumbersome.

(2) Estimate the gold plating thickness. This may be the most important factor. You have to have an idea of how much is needed for different types of parts. Gold plating is measured in millionths of an inch. Another name for this is microinches. I will refer to them as micro". First of all, you must understand that this is a pure educated guess. You must understand that the thickness of gold plating has changed over the years. Also, the plating on such things as fingers or pins can run all over the map. I have seen fingers that ran from 15 micro" to 80 micro". It depends on the usage requirements and which company made the parts.

On the average, Catfish has been using 30 micro", with very good success, for things that plug into other things, whether male or female, or for those things that rub against other things, such as those tiny gold plated balls in small multi-switches. This includes such things as gold plated connector pins and many fingers. This has also been pretty well confirmed by lazersteve's yields.

For things that have to be heated, in order to solder an integrated circuit (chip, die) to a package, you can figure 50 (40 to 60) micro". This includes many CPU's, sidebraze IC's, etc. It also includes hybrid packages (microwave, etc) that often contain hybrid circuits. The gold thickness on all these types of packages have run quite consistent over the years, unlike fingers.

For everthing else, you are on your own. There are hints, however. For things that just sit there, on run-of-the-mill electronic equipment, figure 15-20 micro". The gold is there only for mild corrosion protection. It doesn't need wear resistance. If you run across some rare, high tech, military part that has to withstand high current flow or a highly corrosive environment, the thickness could be out of sight. Most common modern items run between 12 micro" to 35 micro".

Two ways to really know the real gold thickness, on specific parts, is to assay them or to refine them (the whole batch or, just a few of them, as a sample). You can then measure and calculate the surface area. From this, you can calculate the gold thickness. There are specific types of equipment that measure plating thickness. The Microderm and Betascope utilize radioactive isotopes to do this. The Kocour machine actually dissolves a small spot of the plating. There is also X-ray thickness measuring equipment. Most all of these need standards in order to set them up. You can also plate backup nickel on the gold plate, make a metallurgical mount, polish and etch it, and measure the thickness using a metallurgical microscope. I have lots of experience using all of the methods and equipment (except for x-ray) covered in this paragraph. If you want more info on any of these, let me know.

A special category is gold brazes. They probably run from 500 to 1000 micro" thick. A 80/20, gold/tin braze is used around the edge on gold plated lids used to seal IC packages, such as some CPU's, side braze packages, or all-gold plated hybrid packages. Also, on many of these same parts, a 96/4 (I think), gold/silicon braze is used to attach the chip.

The mil specs and other official thickness charts, concerning electronic parts, aren't that helpful unless you know what thickness Class Number was called out when that exact part was manufactured. If you assayed parts, you could probably get some correlation between the data, however. The official decorative gold plating charts are a little more helpful, since some plated jewelry is marked. If you see a marking of H.G.E., e.g., you know it should be 100 micro" thick, when new.

(3) Calculate the gold plating value.

First, for us in the US, who have measured things in square inches.
(a) Get the spot market price of gold, in dollars per troy ounce. Divide this number by 100,000 (ten thousand) on the calculator (Note: This is the only change I made in my method). This gives the dollar value of one square inch of gold plating, one micro" thick. Let this sink in your brain. For example, if the gold price were $660 per tr.oz., the value of one square inch of one micro" thick gold, would be 660 divided by 100,000 or, $.066. Please note that this was a little discovery of mine and you won't find it anywhere else. I just noticed that, when I calculated the value of one square inch of one micro" thick gold, using the density of gold, etc., it just happened to be within 1.5% of dividing the spot price by 100,000. Pure coincidence.

(b) Multiply this times the gold thickness in micro" and multiply this times the number of square inches you have.

Dollar value of gold plating = Spot price divided by 100,000 X thickness in micro" X area in square inches.

Example: The gold spot is $650. You have 9.58 square inches of gold plate that you estimate to be 30 micro" thick.
650 divided by 100,000 X 9.58 X 30 = $1.87.

For those measuring in centimeters and calculating area in square centimeters:

Spot price in US dollars divided by 100,000 X thickness in micro" X area in square centimeters divided by 6.452.

You'll have to convert other currency values on your own.

Play with this. It's easier than I've made it sound. As you probably know by now, I'm a detail guy. Just remember it's limitations. At best, it's a estimate, which will often (but, not always) be a bit on the low side.

Chris

NOTE: I originally made an error in this post. Instead of dividing the spot by 100,000, I divided by 10,000. The correct number is 100,000. I have edited this post to the correct value.

Chris,

As a side note to your excellent post, you can find the plating thickness for milspec equipment if you have the NSN (National Stock Number). There are numerous lookup sites on the internet that will give you all the plating data on the parts.

Steve

Excellent, Steve. Didn't know that. Do you have any links? How do you get the NSN?

Chris the NSN is typically posted on the packaging of a part if bought as new old stock. If bought from government auction sites the NSN's are provided. Components that are large enough to be individually marked, like cannon plug assemblies will have the NSN or equivalent printed right on the side.

Here's a photo of a cannon plug assembly that I've purchased:



I cross indexed the MS3112E14-19P stamped on the housing at

http://5935.iso-parts.com/nsn/

By looking up the NSN number for the housing in the lower section of the page, I found the actual NSN to obtain the data below:

Here's the NSN Technical data for that item's NSN:

MS3112E14-19P

Note the listing for the grains of gold and plating thickness near the bottom in the section.:wink:

Pretty handy eh?!

Of course you need milspec items with part numbers. When you buy these items in pallet form at the auction sites they pretty much always come with the NSN information on the way bill.

This information helps tremendously when bidding for the scrap value.


Steve

Incredible.

I found the gold weight but not the thickness. Is the gold weight for only one pin. It sure seems low for the assembly (.019 grains = $.026). I once had a partner that more than financed our fairly good sized refinery and plating shop from his several hundred drums (I'm thinking 200 drums) of gold Cannon plugs. They really ran good. Of course, they were old. We cracked them open, one at a time, with a Mini Monster shear, dumped out the male or female pins, and stripped them in cyanide. Took about a month for 2 people. Gold was only about $150 back then.

Can you trust government numbers? It's been my experience that the military are not too swift when it comes to PM scrap values and how to handle it. I read the other day that they only make $8 million per year on their PM scrap program. About 99% of it must be walking away.

GSP,

You're right the thickness isn't on this listing! :shock: Oh well, as long as you have the weight of the gold the thickness is irrelavent. I've seen it on some and not others. I agree the value must be for a single contact. I've got a nice pile of these and many other types. They are a lot of work to pluck, the shear sounds like a good way to split them open.

I discovered this information while researching some milspec part numbers I pulled off some scrap I had laying around.

Steve

Lazersteve--that sure is intersting. I tryed the math. I got 49c gold value for 19 pins at todays $666 gold pirce. Did I do it right? Thanks. Mike.

Right on. I actually got 50 cents. There are 480 grains in a troy ounce.

.019 divided by 480 X 666 X 19

GSP--thanks! You sure did it easyer than me. I never was quik at math in school. Thanks for simple exampls. Mike.

Ok, so here's what I get from this quick test:

The total weight of the cannon plug and 19 pins is :



19 Grams.

Each plug yields :



7 grams of pins.

So to get 1 pound of pins you need :

454 / 7 = 64.85 cannon plugs or rounded to 65 cannon plugs.

From all this the gold yield of the pound of the pins would be :

0.019 x 19 x 65 = 23.45 grains of gold per pound of pins from 65 complete plugs weighing

65 x 19 = 1235 grams = 1235 / 454 =~ 2 3/4 pounds of plugs.

And finally the value of the gold:

$0.50 x 65 plugs = $32.50 gold value per pound of pins .


Cool!

Steve

You did great, Mike. You got the right answer. This PM math can get confusing. There are always several different ways to work things and none of them are the "right" way. On important deals, I work the math in 3 or 4 totally different ways, just to check myself.

Hello People, I believe that am at the right place. I've tried to learn this trade the hard way and lost a lot of money, not to mention burned my feet and fingers. I'm excited to be here. I won't bother you guys with silly questions. I'm going to read and read and read , and eventually will come up with a question.
Sincerely,
Tutorus

Welcome to the forum. :wink:

LazerSteve,

I have acquired ~17lbs of those cannon plugs connectors but without the metal collar. Mine have basically two sizes 32 & 65 pin versions. The plastic shell holding the pins is very hard. What is the best/easiest way to remove the pins?

Thanks

For the larger plugs a heavy pair of cutters used to grip the pin and pry it out is the only way I've been able to do it. Most of the cannon plug housings are cadmium so try to avoid using power cutting tools on them unless you wear the proper safety gear (breather). You will still have to contend with the neoprene type grommet that holds the pins in place, it's really tough stuff.

Steve

Thanks Steve. Mine only have the grommet. First I'm counting pins and trying to calculate how much gold I have to work with then I'll remove them. I'm going to document this process well for future reference.