Gold Finger Yield from Gaylord of PCI cards

[MMFJ]

Lots of info on the board about yield of gold/lb of fingers (Flb), but haven't found anything that addresses the approximate Flb to total PCI weight (PCIlb), or ratio of Flb/PCIlb.

Of course, it varies a bit by the type/brand/manufacturer/etc as to total card weight, and a thousand variables of amount of gold per finger, etc. but I'm wondering if anyone has some data on overall 'average', which would come from several hundred pounds of 'mixed' cards.

We have an opportunity to purchase PCI cards by the gaylord (about 600 lbs a shot) and get maybe 2-3 gaylords (or more) per month.

This particular supplier tosses into their 'PCI' box just about anything that isn't the motherboard, processor, power supply or carcass (which means there will be a mix of any/all 'true' PCI cards as well as Expansion Boards and such). We'll of course have to sort through those and sell them as different types - some of which currently pay (per boardsort.com rates) far below what we will have to pay per PCIlb.

We can't just 'flip' the load to boardsort and make any money (just below a break even, depending on total shipping) and although we are looking for other rates, currently, they are 'the best', although shipping costs are killing us!

To bring more cash when we sell, we will remove all the fingers from the cards as well as the 'easy' (socket) chips and sell them separately.

We have limited data on just what to expect in the way of fingers from such a load and could really use your help/data on approximate yields.

Over the past couple months, we have sold 37.9 lbs of trimmed PCI cards and collected 2.5 lbs of 'clean' (snapped off) fingers (we have a guy building us a jig that will make the work simpler and more consistent).

The challenge is that we also collected a lot of other fingers from various/sundry parts like power supplies and such - perhaps .5 Flb, but we have no solid data there and so the best we can do is GUESS, which is not very preferable on these tight margins!

Presuming that we have a 'sample' worth evaluating, from our previous runs;

AMOUNT OF FINGERS PER POUND OF PCI CARDS

• ~5% Flb/PCIlb <<< ===== CRUCIAL data - please comment on your yields!

That being the case, we can estimate

• ~30 Flb from a 600 PCIlb Gaylord.

Taking that to data readily available in the forum

• (2g Au/1 Flb and/or 1 ozt Au/15 Flb)

can we presume that

• 1 PCIGaylord = ~2 ozt Au?

This info is critical to this purchase as we must beat (by at least $.05/PCIlb) an existing, long-time purchaser that recently raised their rates (plus they pick up for free so there's no shipping loss to the supplier...). We will actually be required to pay very close to the 'With Fingers' price we currently get from boardsort - plus our current shipping rate (which will go down somewhat with pallets, though we don't have solid data on that just yet), which makes 'flipping' the load impossible (although we have really tried to go that route! :wink:

Considering the time/effort of removing all the fingers, etc. the venture is 'worth it' (i.e., better than break even), though only because it gets our foot in the door for buying other 'bigger' parts from this (and other, friendly-to-them) supplier. Depending on the actual yield (if it really is 2 ozt Au), that offsets our shipping and labor and we can offer what the supplier wants and still come out with a small button from each load, which will be nice...

We already pay slightly more than anyone on their processors and last month picked up 46 lbs of them (with 28 Pentium Pros in the lot.... and will pick up another batch next week, which is why they are now giving me an opportunity to bid on the other gaylords of stuff they get each month - and told me exactly what price we'll have to offer to get it, so I know just what we have to shoot for at least!

We could really use some input from anyone that has kept data on the Flb/PCIlb to make sure we won't be PAYING to take this load!

Your comments are encouraged and welcome!

 

[glorycloud]

If you use your sample data, you got 2.5 pounds of fingers from
40.4 pounds of boards (37.9 + 2.5) or a yield of .06188 per pound of boards.
A 600 pound gaylord should yield for you 600x0.06188 = 37.13 lbs of fingers.
That 37.13 pounds of fingers sold on ebay will bring spot price I bet:

2 grams Au per pound x 37.13 pounds = 74.26 gr AU / 31.1 = 2.395 Troy Oz
2.395 x $1,600.00 = $3,831.54 for your gold fingers ($103.19 per lb)

If you then sold the trimmed boards to boardsort at $3.10 you would
see 600 - 37.13 = 562.87 x $3.10 = $1,744.90 for a total of $5,576.44
less your ebay fees, paypal fees and labor costs.

if you just sold the 600 pounds to boardsort at $4.75 per pound you would
get $2,850.00. Seems like the math says snap off the fingers and double your money or am I missing something. 8)

The numbers come out lower if you go with your suggested 30 pounds of
fingers per 600 pounds of PCI boards per gaylord but it is still a better yield
depending on how you value your time or your workers time.

 

[niteliteone]

MMFJ
But don't forget to subtract the average of nonfinger material. As in the power supplies and other stuff.

A guestimate of nonbearing material must be factored into the yeild.

Even a +/- 10% gross for nonbearing material can mean a difference of several hundred dollars.

Sorry about this one.
Tom C.

 

[ericrm]

i find it kind of strange ,are you sure about your 2.5lbs for 40lbs of cards??? to me it sound excessive
let say at 4 card a pounds it means you got 2.5lbs from 160 pci/isa cards, i rather warn you about your scale apears to be broken...

 

[publius]

Hijack alert: I have to ask. What is a "Gaylord?"

 

[Barren Realms 007]

Hijack alert: I have to ask. What is a "Gaylord?"

A big cardboard box that sits on a wooden pallet.

 

[ericrm]

since you all elped so much i has choose to help you freeely
after the seeing your number i made a little experiment and here are my result
for 486g of card 1pci ex,1big isa,2small pci,1medium videocard (3 set of finger), i end up with 486 g and 9.4g of finger
for a percentage of
9.4/(486+9.4)=0.01897 of gold finger by weigt
so if we redoo the math

A 600 pound gaylord should yield for you 600x0.01897 = 11.382 lbs of fingers.

11.382lbs x 2g = 22.764 g for a total (at 1600$oz)= 1171$
i will add if you sold the remaining card at 3.10 589x3,10=1826 you would get paid total before expense 2997
now with expense for me 100$of gaz collecting those boards everywhere,500$of gaz to ship,my time(free?),my chemical(60$)
if i just sold the 600 pounds to boardsort at $4.75 per pound i would get $2,850.00.-gaz-time=2350
deeling with them myself 2997-100-500-60-time=2337
the math talk for themself ... dont get fooled THEYR IS NO EASY MONEY

 

[MMFJ]

here are my result
for 486g of card 1pci ex,1big isa,2small pci,1medium videocard (3 set of finger), i end up with 486 g and 9.4g of finger
for a percentage of
9.4/(486+9.4)=0.01897 of gold finger by weigt
so if we redoo the math

A 600 pound gaylord should yield for you 600x0.01897 = 11.382 lbs of fingers.

11.382lbs x 2g = 22.764 g for a total (at 1600$oz)= 1171$
i will add if you sold the remaining card at 3.10 589x3,10=1826 you would get paid total before expense 2997
now with expense for me 100$of gaz collecting those boards everywhere,500$of gaz to ship,my time(free?),my chemical(60$)
if i just sold the 600 pounds to boardsort at $4.75 per pound i would get $2,850.00.-gaz-time=2350
deeling with them myself 2997-100-500-60-time=2337
the math talk for themself ... dont get fooled THEYR IS NO EASY MONEY

THIS is the type thing I am looking for - as I am not really sure of my numbers (as stated in the original post, I had a mix-match of various other fingers all mixed together, not considering that I would need such data..... [now I get the idea of why many on the forum have the mantra of 'document, document, document!'

Although your sample is small, it is at least something - we haven't had any new cards come in awhile and just don't have anything to measure it before we need an answer to buying the gaylord (they sell on the last day of the month).

As for your numbers, much is similar, except that I have negotiated $0 gas collecting (they will deliver) and I think the $500 is a bit high for shipping the gaylord via freight from here (though I have used $.82/lb as my number even though that is UPS boxed rates - always prefer to go a bit high on it and have a "good surprise" at the end!)

The whole thing may be out of the question now as I spoke with the guy and he now wants $4.35/lb for me to get the deal (we started at $3.50, which was looking pretty lucrative...). Of course, doing a 'flip' to boardsort won't work at that (because I'd lose the shipping cost), which is why I looked at doing the fingers for the difference anyway.

I do understand that boardsort is 'close' on the value of the fingers and that selling them with fingers is not too far off. However, with a large pile and cheap labor, I think there is something to be made there in the difference (and when I can sell on fleabay for the same price plus shipping, I make about $6/lb more than boarsort pays anyway... - not a fortune, but it all adds up).

I somewhat concur that "There is no easy money", however, there are some deals where it is not very difficult! (I've been making a living off JUNK for years and just yesterday selling 26 Pentium Pro chips for $1000 was pretty 'simple' money!)

If anyone else has some cards laying around, it would be great to gather some more samples! When we get another bunch, I'll sure be happy to share the yield data. I'd also be interested in hearing from the 'gurus' on here as to just how useful data would be for the finger yield from a gaylord full of 'non-specific mix' (of course, with disclaimers galore of 'your results will LIKELY vary'!)

 

[niteliteone]

The whole thing may be out of the question now as I spoke with the guy and he now wants $4.35/lb for me to get the deal (we started at $3.50, which was looking pretty lucrative...).

Just because he wants $4.35 /lb doesn't mean you have to pay it.
Bid what you feel comfortable with. Don't offer to pay a penny more than you are comfortable with. He might be fishing for more money.
Maybe the other guy is offering $3.51/lb . That's more than your first price.

The profit margin is to slim for you get into a bidding war. Just offer what you think is fair and leave it at that. Don't loose any sleeep over it since most sellers shop around for the best prices anyway.
I know I do 8)

Tom C.

 

[ReapHer_6]

MMFJ.

I am a real novice but I wanted to throw this out there. I have been collecting / stockpiling broken computers for 10 years from my small computer shop. I have a lot of this stuff. I had a great success (for me) in the reduction process for the fingers. The problem is I got so excited that my idea worked that I processed a large box of card fingers only to find out that was not a great idea. Anyway this is what I did.

I took one large box of cards. I failed to weigh the starting amount. (I made several mistakes but I am on a learning curve). I trimmed the cards of their fingers. I did a reasonable job cutting them close but I did not know at the time the trimming has a lot to do with the percentage of Au / weight. (Less fiberglass on the trimmings means better yield) Anyway the box was about 2.5 x 2.5 x 2.5 feet sq. As unremarkable as the box of cards seems, I did error in another noob way that helps with your quest.. I did not sort any of the cards out. The box is a very realistic sample of all types of cards from the year 2000 (when I started my repair business, so the machines I was seeing in the shop that needed to be scrapped were most likely 3-5 years old making them late 90's machines). So the box (sample) contained PCI and ISA- modems, Network cards, Sound cards, video cards etc. (including AGP but not many - if any PCIE graphics cards). So the box represents a great cross section of all types of cards right up until last month Dec 1 2011 when I sold the residential repair side of my company.

This gets better, or worse.. While I was gathering knowledge and glassware needed to begin my precious metal reclamation experiments, I was looking for a way to reduce the fingers so I could AR them. Yes I know, I have since learned from the forum that AR is not the correct approach to processing fingers. But in my quest to find a good reducer for the fingers I noticed our industrial(ish) paper shredder. It dawned on me that the shredder will reduce compact disks. .. CD's I thought? This machine must be able to shred up fingers. Guess what... it worked! Really, really well. I weighed the final reduced cross section of cards spanning 12 years. I am home now, but I know for a fact that I separated the sample into Ziploc snack baggies. I weighed them (accounting for the baggie, I did zero the scale with an empty one on it to TARE it)... the final tally I remember was 8oz per bag times 2 bags and 1 bag with 6 ounces in it. I took out 2 ounces to work with. So the boxed sample of unsorted cards where I removed the fingers, netted 1lb 6oz of really shredded fingers.

So outside of me going in and counting and or classifying the cards ( I still have them in the box awaiting further processing) I don't have any idea what I started with. I am willing to go back and do this if you are anyone is interested, but if there is no use for this data or the idea is stupid, I don’t want to waste any time on it as I am super busy at work.

Ok on to my weird idea.


The reason for this ramblings is my actions did not help me, but in the sample is the exact data you, I (and others might need). BUT I do not have the skill-set (yet) to do a proper job in the recovery, precipitation, and smelting to get REAL data from this great sample. In other words I will screw it up and it will be useless to all of us as far as data is concerned. So I am presenting this idea to anyone who might be interested (and I am braving the possibility that I might get voted off the island for posting a possibly idiotic idea.)

If the forum is interested I would like to offer up the sample to an individual that is willing to process the sample. I am not looking to profit from them or this idea, but I am not looking to give the sample away either. It seems like an accidental occurrence that could result in some really useful data for us all (as long as this has not already been done). In the case this was just a really stupid idea, please go easy. I have thick skin, but even at 41 rejection from the crowd brings back terrible memories of being the kid that got picked last in dodge ball :x I hate rejection.. oh, hell, who am I kidding I must love it, I have been married for 21 years... :|

So if this is interesting, I will gladly go to the lab and itemize the cards in the box that produced the fingers and harvest the weight, qty, and component list and present it back to the forum.

I am hoping the board will help me sort out this next statement. I would like to send the sample (1lb 4oz) to a respected and trusted member for processing and analysis to gather accurate data for the good of the community. (Maybe we could split the final product for our time.)

If we (the forum) are not interested in this data then I will just leave the samples in storage and process them later when I have proven the necessary skills needed to correctly process them and not loose values.

Let the flaming begin, I'll get my asbestos undies on.

John S.

 

[element47]

Can you post a pix of your chopped salad?

Lazersteve might be interested in doing a refining job for you. PM him. He did some excellent platinum work for me.

 

[ReapHer_6]

I guess I take it this was a dumb idea?

John S. :|

 

[MMFJ]

I guess I take it this was a dumb idea?

John S. :|

I'm not sure it was a dumb idea (in fact, I thought it was great!), but perhaps the 'family' is waiting for pictures, as requested by element47. Also, did you PM lasersteve (or any of the 'guru' types)?

One thing I have seen for sure around this forum is that, if your message is not clear or needs more info and it was requested by one, it is likely that many others are waiting for that tid-bit before moving forward as well.

So, get the picture(s) and make some PM messages (if you need assistance on that, PM means 'private message' and the way to do it is to just click on someone's name - pretty easy once you have done it. You can also go to the Member list and look up someone - many ways to get there...)

One kindly bit of advise - around here or really anywhere - don't apologize for your 'ramblings' or ideas and forget the "ok, let the flaming begin" stuff - it doesn't come across well and if your idea isn't well thought out to the point it would deserve 'flaming' (I would not say yours fits that category at all), you shouldn't be posting it anyway! By presenting your idea in a clear manner, which you did, the only thing you should do from that point is to just be open to what comes and process any requests or replies in a timely and polite manner. Certainly, that will get a LOT of respect and 'free training' around here (the folks on this board are not real responsive to 'drama queen' stuff, which is why I like it so much!)

BTW, I hadn't responded to your message as it was pointed more toward the 'seasoned' types, but I did appreciate both the info and the offer to get this sort of data for all that might use it.

My offer on the gaylord had to be low to cover the numbers as discussed, so I did not (yet) get one, however, I'm researching other sources and hope to have something to contribute to this thread soon.

 

[Harold_V]

and forget the "ok, let the flaming begin" stuff - it doesn't come across well and if your idea isn't well thought out to the point it would deserve 'flaming' (I would not say yours fits that category at all), you shouldn't be posting it anyway!

Those who have been on this forum long know that flaming is not acceptable. Perfectly acceptable to post contrasting points of view, in a mannerly way, but if a reader makes a decision to become disorderly, pretty good chance it's going to blow up in his/her face. I don't tolerate folks that can't keep things polite, or can't speak without profanity. (Yeah, I'm a hypocrite---I use vulgar language in real life---but I have a job to do on this forum, and I do it).

Harold

 

[reelman65]

Hi all, thanks for the awesome information source. I am a very small e waste hobbyist. I am interested in the gold, but have not refined anything. I mainly sell enough to pay for the material and then hoard the material with the highest concentration of PMs, like CPUs and RAM etc.

When I process slot cards, I take into account the overall weight of the board that I will be reducing the value by trimming the fingers. For example, some slot cards are barely bigger than the gold finger itself. I definitely trim these. however a huge video card I will not trim.

Just a thought,

thanks again

 

[ReapHer_6]

Damn, I forgot this thred. I will post pics today. I figured the reply to post pics and possibly have LaserSteve process it for me was a quick way of saying it was not a good idea. I was thinking it was a shame as the sample is a great cross section of cards spanning 10 years. So the data harvested (if it was processed by a pro not a newb like me) would help in the quest for data that the origonal poster and the community could use.

Let me get my camera out.

John S

EDIT...Ok, no camera here at the moment so I took them with my phone. I emaild them to myself but they are not here yet.. lol
Anyway I am going to get them posted really soon. I have 2 things to ubpate to the composition of the salad. First I only know know that the closer you cut the fingers the hight the yeild / weight. I see that I did not cut them perfectly as when I did cut them I had not read the post on cutting them close. But I did cut them consistantly. I would say I left about 1/16 of an inch on average of board still attached to the fingers. Secondly I noticed there were some memory chip fingers in the mix. I have the bucket of memory that ended up in the mix so I can get a count of the type and number of chips in the 1lb 4oz sample. I guess it makes the sample data less valuable in answering the average card finger value over a large broad sampling of cards...Unless the gaylord could contain memorychips also. If so, then it makes the data more closely match the real world.

Ok, I am getting swamped at work. I will try to post them later.

Good Day.

John S.

 

[Marcel]

My experience with high quality cut PCI card connectors using A/P is:
1%
That means 1% of the weight of the cut sections will be the weight of the collected gold fingers.

Please note that PCI edge connectors are gold plated in various thicknesses, this ratio refers to high quality PCI cards such as server hardware etc.
Soundcards, low cost vga cards will return lesses results ( maybe 0.5%). Also some asian designers do not use all connectors, so the number of PCI fingers on a card can vary.
Same ratio goes with the connectors for memory boards. Their goldplating is usually thinner, because the card will only be put once into the socket. So expect 0,3%-0,5%

Cards that are plugged in and out multiple times have thicker goldplating ( centronics connectors f.i.) because otherwise the gold would wear off very fast and contacts corrode.
If the goldplating is very thin you must expext losses due to the gold going into solution or beeing dissolved into miscoscopic particles that pass any filter.
I use these rough numbers when buying scrap material. I expect 1% from high grade and 0,3% from low grade material.
To determien wheter it is low or highly plated one can run a test batch or scratch the surface of the goldfingers. High grade material will not reveal undelying copper when beeing scratched once with a screwdriver or sandpaper f.i. wheryby low grade will very easy reveal silver or copper colored layer underneath.

 

[Palladium]

I like this guy already. :mrgreen:

 

[MMFJ]

Well, I finally got a decent load of PCI fingered cards and remembered to weigh them before getting started, so I have a bit of data to add on this topic.

For anyone new that is trying to get an IDEA of what yield you can get from X number (or pounds, etc.) of PCI cards, know that it is IMPOSSIBLE to give any sort of accurate count - for the reasons that should be obvious from the pictures. This data is for 'example' estimates only and WILL vary by any given set of cards. The data can be considered one datapoint of an 'average' load of PCI cards (it would be great if we could get several dozen 'example' points and then average them to get an 'average of averages' sort of thing...)

OK, here's what I started with;

23.4 pounds of MIXED (including vintage and recent, large and small and a variety of interface types as well as some Chinese and some American made ones)


After about 30 minutes (time between pictures, though I did get a phone call and someone walked in.....), I got the following result;


So, for 23.4 pounds of MIXED PCI cards, I got .5 of 'clean' gold fingers (mostly 'clean' - less than 2% waste, I'd guess, so I'm calling my .6 as .5 'clean') in 30 minutes work.

Therefore, as a "BALLPARK" number (something to keep in your head, NOT as real expectations!), it would seem from this sample that approximately 4% of the total weight of the lot could be counted as gold fingers. As this varies wildly from lot to lot, it is best to estimate on the low side. Plus, you should look at the samples I pulled out of this lot (at total random, just grabbing anything that looked different than the rest) and compare to your lot before applying this 'ballpark' math!

So, was it worth the effort to take the fingers off, or is it better to sell them whole? (another question, perhaps, but certainly pertains to this sort of lot!)

At today's rates, and presuming the fingers are processed by a competent refiner (not me! , this results in approximately $10/hr plus a bit of profit for the 'house' - i.e., it is a decent job for a minimum wage worker to 'earn his keep'. The difference in the price you get for selling cards with fingers vs. those without is something like that 'house' profit, so it really boils down to - do you do the work or sell it where someone else can afford to?

These numbers are nothing really new to me, other than the fact I actually kept track of them this time and duplicated the results that have been anticipated (through discussions with several people).

Hopefully you will find this data helpful in your endeavors.

 

[ericrm]

So, for 23.4 pounds of MIXED PCI cards, I got .5 of 'clean' gold fingers (mostly 'clean' - less than 2% waste, I'd guess, so I'm calling my .6 as .5 'clean') in 30 minutes work.

Therefore, as a "BALLPARK" number (something to keep in your head, NOT as real expectations!), it would seem from this sample that approximately 4% of the total weight of the lot could be counted as gold fingers.

0.5 finger / by 23.4 = 2% not 4%........
just to make your math correct