Specific Gravity Test Variance

[RealNick]

I just familiarized myself with conducting specific gravity test for my gold and silver buying. I'm looking to see what level of variance is acceptable in test results in comparison to predicted results. I do the tests with an AWS .01 scale and use tap water. I use a thin cotton string to submerge the subject. I acid test everything and all have passed at specified levels. No stones

Here's some of the results:

Test subject 1
14K Yellow gold scrap ring (no marking)
5.25 grams
0.42 displacement
Specific Gravity = 12.50

Test subject 2 (marked)
14K YG pendant
0.95 grams
0.08 displacement
SG = 11.875

Test subject 3
14K YG Pendant (marked on the bale)
1.50 grams
0.28 displacement
SG = 5.357

Test subject 4 (no markings)
18K YG ring
16.49 grams
1.23 displacement
SG = 13.406

With predicted levels for 14K yellow gold being somewhere around 13 and 18K yellow gold in the 15's I'm a little dismayed at the results. Obviously, test subject 3 can't be gold by any stretch, it looks like gold and passes 14K acid but with that result, no way, right?

Test subject 4 seems pretty off too. I acid tested when I purchased and brought it by a jewelers who also concluded it was 18K gold (and offered on the piece).

Additionally, I did tests on a bunch of silver morgans and an AGE and a Kurgerrand. All of these test came in very close to predicted results.

So what gives here, anyone else running into these kind of results? What level of variance can I expect on these test?

Thanks in advance!

 

[FrugalRefiner]

RealNick,

Welcome into the light. Two years as a member before your first post is rare.

Specific gravity can be very accurate on items of known, consistent composition. The Morgans, Eagles and Krugerrand are consistent because the alloy is always the same.

Now, if we take 10 grams of fine gold and alloy it with 10 grams of platinum, the specific gravity will be higher than pure gold. If we take another 10 grams and alloy it with 10 grams of aluminum, the specific gravity will be much lower. That's an extreme example, but it shows that the choice of metals used to create an alloy can affect its specific gravity. Since you don't know which metals were used in the pieces you test, the results will vary, but I don't know how much.

Cotton string absorbs water, so as you use it the string will gain weight.

It's possible that the bail was the only thing on the pendant that was 14K. I've seen it done on chains, where a karat gold tag or spring ring was attached to a crap chain.

While an item stamped as 14K could, in the past, be as low as 13K, the minimum legal level for 18K would have been 17K. An allowance of 1/2K was made for the alloy, and another 1/2K if there was solder used on the item. 15K would not have been legal for an item stamped 18K. That's not to say it was never done, just that it wasn't legal. I've seen plenty of gold plated junk that was stamped 14K. :evil:

Dave

 

[goldsilverpro]

Can you detail as to how you did this, especially the weighings. That scale is not sensitive enough to give good accuracy, but you shouldn't be off that far.

 

[RealNick]

I did the test like you see on the videos on youtube. The scale, plastic cup filled about 1/2 way with plain city tap water, and sewing thread. First weighed subject, then suspended subject in water with thread. Then divided the mass by the weight of the water I displaced, basically D=M/V. For predictions i used formulas and pre calculated results found here:

http://www.hauserandmiller.com/reference/melting.html

I also redid the test using waxed dental floss because i figured the cotton tread was absorbing water each time. I zeroed the scale out between all weighing. The results were almost identical between cotton thread and the dental floss.

On test subject 3 I didn't test the bale but on the pendant itself. With a SG of 5.35 I don't know what alloy would actually bring the SG that low but still pass acid. A 12.5 result...ok, sure, maybe...but 5.3? Both test subject 3 and 4 have small voids, not hollow but little spaces between the metal as part of the design. I'm wondering if perhaps the gold acquired a slight static charge and therefore possibly repelled water out of those spots thereby increasing the surface area of the subject...this is just a guess, not like I saw any evidence of this during testing. otherwise, how can a pendant with a SG of 5 or 6 pass 14K acid?

For 18K yellow gold I predicted results being somewhere 15.37 to 15.77 depending on the alloy % with copper and silver. I used the method at the bottom of the page of the link above. I came up with a 13.4 which indicates a clean 14K pass...but not 18K

Again, pretty new to SG testing and I know the scale is not ideal but I'm pretty sure I'm conducting a clean test. Looking to get an idea as to what level of variance is expected and at what point I should start to be very suspicious.

Thanks!

 

[goldsilverpro]

I did the test like you see on the videos on youtube. The scale, plastic cup filled about 1/2 way with plain city tap water, and sewing thread. First weighed subject, then suspended subject in water with thread. Then divided the mass by the weight of the water I displaced, basically D=M/V. For predictions i used formulas and pre calculated results found here:

http://www.hauserandmiller.com/reference/melting.html

I also redid the test using waxed dental floss because i figured the cotton tread was absorbing water each time. I zeroed the scale out between all weighing. The results were almost identical between cotton thread and the dental floss.

On test subject 3 I didn't test the bale but on the pendant itself. With a SG of 5.35 I don't know what alloy would actually bring the SG that low but still pass acid. A 12.5 result...ok, sure, maybe...but 5.3? Both test subject 3 and 4 have small voids, not hollow but little spaces between the metal as part of the design. I'm wondering if perhaps the gold acquired a slight static charge and therefore possibly repelled water out of those spots thereby increasing the surface area of the subject...this is just a guess, not like I saw any evidence of this during testing. otherwise, how can a pendant with a SG of 5 or 6 pass 14K acid?

For 18K yellow gold I predicted results being somewhere 15.37 to 15.77 depending on the alloy % with copper and silver. I used the method at the bottom of the page of the link above. I came up with a 13.4 which indicates a clean 14K pass.

Again, pretty new to SG testing and I know the scale is not ideal but I'm pretty sure I'm conducting a clean test. Looking to get an idea as to what level of variance is expected and at what point I should start to be very suspicious.

Thanks!

First of all, you should be using distilled water.

Test something that you know the SG of. Try a piece of pure copper (8.9) or a pure silver coin (10.49). Maybe even a clean, rust-free piece of mild steel (~7.86) if you don't have the others. If you have all 3, test them all and see what you get. Note: I just noticed that you already tested some known objects and that they tested correctly. Either the unknown objects you tested are all low or your technique is poor. While in water, all the air bubbles clinging to the object or trapped in cavities must be eliminated or you'll get errors. A small drop of liquid dishwashing soap gently mixed in the water can help this. Also, while submerged in water, I tap on the object with a stir rod and this releases most of the bubbles. Unlike coins, jewelry often has a lot of cavities that might trap air.

There are 2 basic ways of making SG measurements with a scale. When weighing in water, (1) you can either suspend the object from the scale itself or (2) you can put the water container on the scale, zero it, and hang the object so it's submerged in the water. The calculations are a bit different. For the first (1), SG = Weight in air divided by the difference of the weight in air and the weight in water. For the 2nd (2), SG = weight in air divided by the weight of the suspended object in water. The 2nd method is simpler to set up. There is a good discussion of the 2nd method somewhere on the forum.

The 1st link shows method (1). The 2nd link method (2). If using (2), I would hang the object from a stand and let it stabilize. Some of the techniques used in the videos aren't great but they do show the general idea.

https://www.youtube.com/watch?v=86DNkoQBk9w

https://www.youtube.com/watch?v=xYdSEAm-7uI

 

[RealNick]

Thanks for the helpful responses!

As I mentioned before, I learned this test off of a youtube video. I had an opportunity to buy some gold coins and needed another test that wouldn't damage the numismatic value of the coin for resale. I didn't want to invest in a Fischer tester, although they look great, just too pricey for me. Turns out I passed on the coins as I couldn't get them at a price I was willing to pay.

I’m doing test method 2. I needed another, good, “rubber meets the road” type of test that's non destructive to protect my interest when buying. Test method 1 is just too involved a setup to do when I'm meeting a guy at a starbucks to buy gold coins or picking up some Morgans at an estate sale.

I believe my initial test technique must have been somewhat poor, and lead to some inconsistent results. I still can't account for the low results of test subject 3 but using the tiny drop of soap and tapping on subject has helped, I pulled a 7.5 from test subject 3. still way low for 14K gold, but higher. The 18K gold ring (TS4) pulled a high 14's with the refined technique.

I have a question related to this test: we are just taking measurements to arrive at density (D=M/V) and the “weight in water”, the “V” part of the equation where you are submerging the subject in the water, isn't that just weighing the weight of the water that the subject displaces? If so, why does test procedure 1 work? The weight (mass) of the subject doesn't change in water, does it? Hopefully some of you scientists here will clarify this for me...

I'm going to search for the other thread about specific gravity test on this site. I must say that I am very impressed with the wealth of information found at this site. Thanks again!

 

[FrugalRefiner]

I am far from a scientist, but I'll try to explain.

I have a question related to this test: we are just taking measurements to arrive at density (D=M/V) and the “weight in water”, the “V” part of the equation where you are submerging the subject in the water, isn't that just weighing the weight of the water that the subject displaces?

As you said, you are weighing the weight of the water displaced. But all densities are measured as they compare to the weight of water, and 1 milliliter of water weighs one gram, so a volume of 1 ml (V) has a mass of 1 gram (M). The density of water (D) = 1/1 = 1.

If so, why does test procedure 1 work? The weight (mass) of the subject doesn't chance in water, does it?

The mass of the object remains the same, but the weight as measured by the scale is less by the weight of the water it displaces. If you lift a boat on a crane fitted with a scale, you'll find the mass. If you lower the boat onto the water, the mass doesn't change, but as long as your boat floats, the weight borne by the crane will go to zero.

1 cubic centimeter of gold has a mass of 19.3 grams. It will displace 1 ml. of water. The weight in air will be 19.3 grams. The weight in water will be 18.3 grams. Weight in air minus weight in water is 19.3 - 18.3 = 1. The weight in air divided by the difference is 19.3 / 1 = 19.3

I hope that makes sense

Dave