How accurate is a pin sample ? misc questions

[HAuCl4]

Harold got out of the refining business long before the price of gold rose to what it is today. A refining fee of 2% today will yield a dollar margin in the same range as 10% in the time frame when Harold refined.

It is unfair to judge percentage rates without figuring the base price of the metal involved.

The 10% margin I spoke of would be a damned good figure, even today, if you consider how hidden fees change their advertised fee.

And this is a perfect example of why posting and reading in forums can become confusing and ego driven, despite the best of intentions. I don't even know what we are discussing anymore. :shock:

I'm done. By the way: Did my cast iron pans win the contest?. I want that platinum!. :lol:

 

[Lobby]

REFINERS ASSAY 55.559%
INDEPENDENT ASSAY RESULTS 55.67%
MY NITON XL2 XRF RESULTS 56.31%

I'd say the refiner's and the independent's assay are the same number, statistically speaking.

I believe your concern is accuracy of your xrf.

 

[irv]

I agree except with the difference in the refiner and independent assay , in my case due to the size of the melt, equals $1,334.00 real dollars which should be in my pocket-not the refiners.

 

[Lobby]

I agree except with the difference in the refiner and independent assay , in my case due to the size of the melt, equals $1,334.00 real dollars which should be in my pocket-not the refiners.

I understand your pain. Also that the refiner is using the lower of the assay numbers.

But from a statistical point of view, there is no difference between the assays. They're the same number. Unlike your bank balance, where you know how much you've got to the penny, in the analytical industry there's no "true" number. It's more an average of many measurements.

I think what you need to address is how accurate the Niton is. Perhaps develop error estimates ("it generally reads about 1% high").

By the way, how much did the Niton cost?

 

[irv]

17k

 

[4metals]

You are using an XRF which is not capable of the degree of accuracy you need to work at the margins you are working. A $17,000 Niton is on the low low end of their equipment. They make units with alloy make-ups in their memory to tell the difference from 304 stainless and 316 stainless and also give percentages of other metals in the alloy. When it matches a alloy profile in memory it identifies it. The numbers it spits out of gold and silver content are not as accurate as the more costly but it is fine as a go or no go type machine but if you're buying for a few percentage point gain you need a better machine.

If you go to a refiner and watch the melt and take the sample and have it fire assayed, you will know the assay. That doesn't really help when you are paying based on a number your XRF spits out and it is high to begin with.

There are many gold buyers in NYC who use the stone to determine karat and they pay on 2% less than plumb at payouts as high as 99%. These guys are having trouble as well settling on XRF results.

 

[irv]

I like the xrf because I do not need to use acids anymore, however, it is not a silver bullet and you can still make a big error if you are not careful and use your past experience in visually inspecting different type items. I need to grind some items to penetrate the outside layer to get a true reading.

 

[HAuCl4]

I agree except with the difference in the refiner and independent assay , in my case due to the size of the melt, equals $1,334.00 real dollars which should be in my pocket-not the refiners.

A mulligan last post...

That .11% difference is normally TOO MUCH imho, but it is possible due to sampling and the average karat of what is in there. If you ship fine gold, those assays, if done accurately and honestly should be within .03% using a 4 decimal assay balance, and even tighter (0.01%) with a 5 decimal (or microgram) balance.

You must have an agreement with the refiner/buyer to split the difference in the assays, if they differ more than X%. That should cut your $1,334.00 in half.

I'm a bit extreme in my recommendations but anyways: Sell the XRF or recalibrate it with an array of karat sample pins, setup a fire assay and melting lab, and become a master of the touchstone/acid test (with the array of sample pins) for buying.

You have done the hardest part, which is to establish an steady flow of material to buy. It's always up to you, but the technical part is much easier to establish or acquire.

Good luck. 8)

 

[goldsilverpro]

I like the xrf because I do not need to use acids anymore, however, it is not a silver bullet and you can still make a big error if you are not careful and use your past experience in visually inspecting different type items. I need to grind some items to penetrate the outside layer to get a true reading.

It seems that what 4metals is implying is that there is no "true reading" with your xrf, at this level of accuracy, no matter what precautions you take.

 

[irv]

From my limited experience with the XRF I can say for sure-

1. you must be careful and visually inspect items and test when there is doubt

2. it is great for silver and platinum

3. it is very helpful to me when I am buying high karat gold..18k and higher

4. most important, it gives me a lot more credibility with my customers because they know I am trying my best to be fair to them and to myself at the same time, I can't emphasize this benefit enough for me. Many of my customers now come to my office just to test things which is fine because usually I buy the item later, I even give my customers a printed copy of the xrf reading on their item which they can show to their seller.

5. I am happy I do not need to use nitric and a stone as my fingers are turning white again.

As I said earlier this is not a silver bullet to correct all problems but the benefits are there.....from my experience to date.

 

[Lobby]

There are many gold buyers in NYC who use the stone to determine karat and they pay on 2% less than plumb at payouts as high as 99%. These guys are having trouble as well settling on XRF results.

One can get that accurate with a stone and acid?

 

[4metals]

With a stone and acid you can get within 1/2 a karat or 2%. The issue is the small margins of profits if you are working that close. Bottom line is the acid testing on the stone is centuries old technology, the XRF is new technology but the system is not perfect. Ironically fire assay is the oldest and most accurate and reproducible way to know for sure, it is why refiners still use it. But to buy a ring and do a fire assay is not an option so you need time and cost saving techniques for the fast purchase. Trouble is a fast buy can bite you in the a** if you are working too close.

The Niton also displays a degree of uncertainty for the metal in question. Some buyers, especially when buying odd stuff that you suspect is off karat, use that factor to lower their payout to cover themselves. You can program the software to print the result of the measurement times the degree of uncertainty if you need to give out printouts to satiate a customer.

I have found that rolling a pin sample for example gives a better reading because the beam reflects evenly off the measured surface. In this case the uncertainty should be less and it generally is. Trouble is rolling out a guy's wedding ring is rarely an option.

I've said this before an XRF is great for buying gold (if you are conservative and have a working margin to profit from) but when you go to a refiner and the scrap is melted and sampled, I'd take a 90% advance on an XRF reading but settle on the age old fire assay. That is how you get the best deal, watch the melt and do a fire assay.

 

[irv]

I sincerely respect 4metals opinion and I have incorporated his suggestions and knowledge into my modus operandi. Thank you 4metals for your help and factual answers in helping me.