How many coins are minted from a pair of dies?

TopographicOceans · March 4, 2015 10:22AM

For business strikes and proofs?

coindeuce · March 4, 2015 10:28AM

Never been any standard set in stone (or steel). you could probably estimate the number of coins that any singular die would last for if you had LOTS of time to analyze the metallurgy of a specific die, and a super computer to conduct a statistical forecast. If you are familiar with the craziness that ensued at Philadelphia in 1878, you know that some dies for Morgan dollars lasted for less than a few thousand coins. Not a very economical outcome, considering how labor intensive and costly it was to make durable dies in that time period under the circumstances. Conversely, there are dies in modern use that could conceivably last for tens of millions of coins without significant erosion or concern about spontaneous failure.
For reasons that are mostly off topic to this thread, I'll state for the record that I can foresee a time when coins for circulation will no longer be struck in hydraulic presses - they'll be designed to be cut using laser technology. The Royal Canadian Mint already employs a laser cut privy/security mark on Gold 1 ounce Maple Leafs.

291fifth · March 4, 2015 10:32AM

It has varied widely over time. I have read the average die life for early shield nickels was only about 15,000 strikes.

messydesk · March 4, 2015 10:37AM

Depends largely on the coin size and experience coining it.

Cameonut · March 4, 2015 10:43AM

You asked a very broad question. Obviously this will vary by denomination and what year coins were struck. Since you were not specific, I'll answer the question for one year. This is data from the US Mint records in Philly. Only applies to the Philly mint - not Denver.

For 1954
Business strikes
50c 600,827 pcs/die for obverse, same for reverse
25c 451,410 obverse, 350,132 reverse
10c 427,495 obverse, 360,279 reverse
5c 314,123 obverse, 308,043 reverse
1c 567,257 obverse, 500,289 reverse

Proofs
50c proof 4,408 pcs/die for obverse, 4,747 for reverse
25c proof 5,020 obverse, 5,344 reverse
10c proof 4,605 obverse, 5,756 reverse
5c proof 5,278 obverse, 5,370 reverse
1c proof 6,135 obverse, 6,032 reverse

So you can see that you get roughly 100 times the business strikes off a die vs proofs.
Back in these days they were not chrome plating dies to my knowledge. It would be interesting to see some more recent data knowing the die technology has improved and the relief is much less..

Hope this is helpful.

19Lyds · March 4, 2015 10:55AM

Russell Rulau from Coin World asked this very question in July of 1972 and the reply from the US Mint was as follows:

This was pulled from the Authoritative Reference on Eisenhower Dollars - Appendix A

As you can see, die life is directly related to planchet size and planchet hardness with the Copper Cents having the longest die life.

P.S. Also remember that the number for "proofs" is for "strikes", not coins as each proof coin receives at least two strikes to bring up the relief.

coindeuce · March 4, 2015 10:56AM

To expand on what CameoNut shared: the sub five figure totals of production by individual die for each denomination of Proof coin in 1954 does not necessarily accurately portray how many coins could have been made before individual Proof dies had to be retired because of erosion or fatigue. It was more a matter of choice by Mint officials who had already done market forecasting for sales of Proof sets or singles, and the general standard they wanted to maintain for quality throughout production of the 200,000+ sets issued. The numbers could very well have varied in subsequent years beyond 1954.

TopographicOceans · March 4, 2015 11:08AM

Thanks for the numbers. I really had no clue. Very interesting.

I wonder how 19th century dies performed?

mustangmanbob · March 4, 2015 12:00PM

Slightly off topic, but I would guess the Mint has some very good metallurgists.

I was with the M1 Abrams tank. We had sensors on the main barrel that registered the number of rounds fired, the type of each round (SABOT, HE, HEP, etc.), the air temperature, temperature of the ammunition, the barrel temperature, etc. that fed all the data to a shot computer to compensate for all these factors. Each round caused a little wear on the barrel. If the sun was high the in the sky, the barrels would heat more on the top, causing a slight bow in the metal. The barrels will sag with gravity. The metal of the barrel is different after firing 7 shots quickly.

All this is driven by the need for accuracy. A 1 millimeter shift means a 3 meter shift firing at the bad guys 3 kilometers away. 3 meters off means a miss, and that can mean the bad guy gets me instead.

DaveG · March 5, 2015 1:27PM

I just happened to be skimming through Randy Wiley's 1986 article on Die Varieties of Liberty Seated Half Dollars (in America's Silver Coinage - from the Coinage of the Americas Conference of the ANS), in which he estimated that average expected die life was 155,000 half dollars. He went on to say that dies for larger coins had a shorter life and that during the Liberty Seated era, the die life for quarter dollar dies was about double that for the half dollar and the die life for dimes was four times as great.

Off the top of my head, I don't know if more recent research has resulted in different estimates.

BillJones · March 5, 2015 9:54PM

<< Thanks for the numbers. I really had no clue. Very interesting.

I wonder how 19th century dies performed? >>

A die's useful life during the 1790s and early 1800s varied widely. Years ago I collected half cents by die variety with a special interest in the Draped Bust pieces, 1800 to 1808. The stemless wreath die was used to make the most die varieties for the 1804, 1805 and 1806 half cents. The 1804 and 1806 stemless wreath half cents, listed in the Red Book, are the most Draped Bust half cents of all, and the most common of all half cents. That die probably lasted for a few hundred thousand impressions. As the reverse die, it was the anvil (bottom) die which had a longer expected life. It was obviously made of some very good steel at a time when the term "steel" was a very lose concept at the first U.S. mint.

Conversely, the first die that was combined with the 1808 over 7 half cent variety may have lasted for only a few hundred impressions at most. It cracked early on creating a rare variety. During the time I was collecting I think five examples were known. Today I think it is somewhere around 10, but I don't know for sure.

This is a medal, not a coin, but it is the only piece in my collection I have use to illustrate the point. Joseph Wright made this obverse die for a gold medal that was to awarded to General "Light Horse" Harry Lee (the father of Confederate general, Robert E. Lee) for his victory at Paulus Hook (now Hoboken, New Jersey) during the Revolutionary War. The medal was supposed to have been made by the French, like the other American Revolutionary War medals, but our officials in Paris forgot to order it. As a result the U.S. mint tried to make.

During the die making process, both the obverse and reverse dies broke. It is not known if Lee ever received a medal, but in the 1870s the U.S. mint did trot out the old broken die, paired it with a new reverse die, and made a few medals from it. This piece is also known with a blank reverse. You can see all the cracks on the obverse. This die did not last long. The mintage was probably less than 40 or 50 pieces before it was replaced.

This is was made from the replacement for the reverse die, not the original.

keyman64 · March 5, 2015 10:58PM

For my 10,000th Post I give you....

If you want to know what things were like at the San Francisco Mint in 1896 then
there is nowhere else to turn than this original handwritten US Mint document.
I think it provides a full picture of what was going on. Enjoy!

Based on communications that I have had with a couple of researchers, most if not all of the
documents like this in the 20th Century have been destroyed. Very sad.

CaptHenway · March 6, 2015 12:04AM

A most excellent 10,000th post, Sirrah!

Of course, as Mark Twain wrote, there are lies, danmed lies and statistics! The average life shown for the eagle dies is suspiciously low. I suspect (but cannot prove, of course) that they simply used two pair of dies to make their roughly 125,000 coins at some point in the year, never made any more, and retired the dies at the end of the year with a lot of life left in them.

And FWIW, the Redbook shows a mintage of 1896-S Eagles at 123,750. This document shows 125,202. Should we assume that the 125,202 figure shows the actual number of pieces struck, which would be important for a study of die usage, with some of that number having been remelted as unfit for issue?

TD

keyman64 · March 6, 2015 12:32AM

<< A most excellent 10,000th post, Sirrah!

Of course, as Mark Twain wrote, there are lies, danmed lies and statistics! The average life shown for the eagle dies is suspiciously low. I suspect (but cannot prove, of course) that they simply used two pair of dies to make their roughly 125,000 coins at some point in the year, never made any more, and retired the dies at the end of the year with a lot of life left in them.

And FWIW, the Redbook shows a mintage of 1896-S Eagles at 123,750. This document shows 125,202. Should we assume that the 125,202 figure shows the actual number of pieces struck, which would be important for a study of die usage, with some of that number having been remelted as unfit for issue?

TD >>

Yes, the 1896 document answers many questions for the time but it also causes us to ask more, as you have demonstrated.
It is also interesting to note that I do not believe this document was available when the Redbook Mintage figures were
derived/published originally. As for the eagle production that year, RWB might be able to shed some light on the "needs"
of the mint that year which in turn might support your idea that there could have been life left in the dies when they were retired
at which point the mint opted to produce no more.

As for assuming things, I would not. When I was a young lad, my parents taught me that not to be wise.
However, I would not be opposed to a possible strong hypothesis if you have evidence that the mint had practiced similar actions previously.
On a "State of the Life of Dies" document, I am also not sure if they would report the total number of items struck or the total of items struck well
and therefore fit for circulation.

SaorAlba · March 6, 2015 12:45AM

<< It has varied widely over time. I have read the average die life for early shield nickels was only about 15,000 strikes. >>

And they usually were still poorly struck. I want a shield nickel for my collection, but I have yet to see one I like enough to ante up the dosh to buy it. I really do believe the mint had a rough time with cupro-nickel during that time, even the FE and IHC cents have some striking issues.

CaptHenway · March 6, 2015 3:08AM

Number struck does not have to correspond to number released. Struck coins can be remelted for various reasons. Look at the 1873-S No Arrows half, or the 1873-S dollar.

I don't know if any Mint knew on January 2nd what they were going to strike in the upcoming year. I believe that a lot of the denominations were struck to demand. If they had an order for eagles early in the year, they would strike eagles and then wait to see if they had any more orders for eagles during the year. If they did, they would strike more.

Stuart · March 6, 2015 3:25AM

KeyMan64: Congrats on your 10,000th forum post, and for sharing that most interesting and very informative 1896 San Francisco Mint die life historical document.

keyman64 · March 6, 2015 3:38AM

<< KeyMan64: Congrats on your 10,000th forum post, and for sharing that most interesting and very informative 1896 San Francisco Mint die life historical document.

>>

Thanks, happy to share. I have shared it on the forums previously once or twice,
including when I got it from Roger W. Burdette a couple of years ago. Good stuff.

Type2 · March 6, 2015 4:00AM

<< KeyMan64: Congrats on your 10,000th forum post, and for sharing that most interesting and very informative 1896 San Francisco Mint die life historical document.

>>

Same here Congrats and nice post..

CaptHenway · March 6, 2015 6:55AM

I referred the question to Mr. Burdette, and received this reply. Reprinted with his permission.

--

I'm not quite sure of the question(s) being asked; however, maybe this will help….

The mint tracked die life by the number of pieces struck. This was done officially to evaluate the quality of die production and to determine if steel from one supplier was better than that from another. I don't know if either of these were actually influenced by the numbers since the two most consistent factors observed in die usage are wide variability in die life, and a tendency for average die life to improve over the span of several years. (This latter is from small changes made to the master dies and master hubs. Think of changes to the Buffalo nickel from 1913 to 1916, for example.)

The count was always of pieces struck – not coins accepted by the coiner as "good coin." Thus, the 1896-S $10 required striking 125,202 planchets to produce 123,750 good coins. The 1896-S $5 required 158,475 strikes to make 155,400 coins. (Also, gold was counted and sacked in units of $5,000 which is why most annual coinages -in dollars - are given in round numbers evenly divisible by $5,000.)

The struck but not counted coins were a mixture of defective pieces and remainders that were not usable in assembling full bags. Good-coin remainders were carried forward during the calendar year (called a "weigh back") with any left at the end of December being rolled out and melted for reuse. (Keep in mind that the US Mints used a dual test of tolerance for gold coin weight: individual tolerance and a bag tolerance. The legal weight of $5,000 in gold coin was 268.75 Troy ounces, with a deviation of only one-hundredth of an ounce per bag from this legal weight. The weight range for a bag of new double eagles was 268.74 to 268.76 Troy ounces. Bags were assembled from coins that were standard weight and ones that were within individual tolerance but slightly heavy or slightly light.)

I hope this has at least partially answered the question

astrorat · March 6, 2015 7:18AM

Great 10K post, keyman!

I am not sure how relevant, but two questions came to mind with 1896 data.

1. Are the data based on fiscal year or calendar year? Much of the US Mint stats are from fiscal years and not calendar years. Of course, fiscal year coverage includes portions of two different calendar years. So, data based on coin dates can get confusing.

2. Is there a definition of "die life" as used in context of the 1896 data? Does "die life" refer to a die that was used to failure or one that was used until it was pulled for evaluation?

Cool topic.

keyman64 · March 6, 2015 7:21AM

Well done Captain for reaching out.

I thought about reaching out to him over the weekend but you beat me to it.

It is for reasons like this that it makes me very mad that he is not an allowed member of these forums.

CaptHenway · March 6, 2015 7:22AM

The totals given seem to correspond with the calendar year figures given in the Redbook, though as we have discussed they seem to be the "gross mintages" rather than the "Net mintages" given in the Redbook.

keyman64 · March 6, 2015 7:26AM

<< Great 10K post, keyman!

I am not sure how relevant, but two questions came to mind with 1896 data.

1. Are the data based on fiscal year or calendar year? Much of the US Mint stats are from fiscal years and not calendar years. Of course, fiscal year coverage includes portions of two different calendar years. So, data based on coin dates can get confusing.

2. Is there a definition of "die life" as used in context of the 1896 data? Does "die life" refer to a die that was used to failure or one that was used until it was pulled for evaluation?

Cool topic. >>

Thanks.

1. Calendar Year
2. Dies were pulled for failure and visible indications of being worn out but not necessarily failing I believe.
The die's life was determined to end when it was pulled permanently from production, I would think.

I wish Roger could participate on these forums.

He would be setting people straight all the time.

How many coins are minted from a pair of dies?

Comments

Leave a Comment