MOSFET Pool for Matching Qs (Aleph X or Other)?

[lgreen]

I'm thinking about all the wasted MOSFETs out there as we all buy a lot of them to try and come up with a few matched sets.

Anyone interested in starting a MOSFET pool?

It would work something like this-
everyone sends their MOSFETs to someone, who puts them in a big box and sends the box around to members, each recepient of the box--

1. looks through it for matched sets, removing however many MOSFETs they want,
2. adds to the box at least as many transistors as they took out, of the type they took out (ensuring new supply in the pool, we could make this 1.0x or 1.5x or something),
3. sends the box to the next person.

Other variations could be marking indiv. mosfets in a manner we all agree with so that they can be easily matched with new ones. Or maybe a deposit is required so no one sells the big box of transistors? time limit of 1 week or less to whoever gets the box? Having a max # of how many people can take, scheduling and totals updated on wiki?

Also, if someone has a good test apparatus it would be really nice to include this in the box as well so those who get the box don't have to build their own, so this would avoid slowing down the rotation of the box.

When the pool ends, and we are left with a box of 1000s of Q's, we have many options - send it to Nelson as a thank you, have a sale and contribute the proceeds to diyaudio.com, or whatever else we can agree on.

Anyone think this is a good idea or know of any reason why it won't work?

[kilowattski]

If everybody is honest I think this is a fantastic idea, especially for those who do not have the equipment to do the matching themselves.

[jackinnj]

if everyone is mindful of static -- if everyone tests at the same ambient temperature ---

[TheJudge]

I'm interested.
Maybe it's nice to only add MOSFET's from the same supplier?

[GRollins]

The one fly in the ointment is that the matching of devices is based on the assumption that all the devices were cut from the same wafer. If someone gets the box, sees a couple of IRFP044s and uses them with some that he already had on hand, then they won't be from the same wafer. Matching the Vgs alone won't overcome this difficulty.
I use leftover MOSFETs for other projects. That's part of the reason I put the Mini-A project together. It was a sink for orphaned MOSFETs.

Grey

[lgreen]

Ok, so how does one go about getting a bunch of transistors from the same wafer?

If so, maybe a group purchase of a few wafers would be better?

[GRollins]

Buy a tube of MOSFETs. Twenty, fifty, a hundred. However the manufacturer packages them. Many devices are embossed with a code, e.g. 3G, that is related to the sequence when they cut the wafer. If you've got 3G in your hand, it's a fair bet that 3F and 3H are going to be pretty similar in characteristics.

Grey

[lgreen]

So how do you do that? If I go to digikey or mouser for online ordering how many irfp044 and 9610 do you order? Or just get 100 and hope for the best?? It doesn't list the size you mentioned as far as I can see.

[jacco vermeulen]

So, what will be the total size of lots 3F, 3G and H ?

I can imagine there will be plenty of interest for a group Mosfet buy.

[Brian Donaldson]

Group buy?

I just had one for IFRP244. Bought 1/2 of digikeys stock(300 pcs). I figured 10% waiste, and sent everyone 4 extra ba$tards and had nothing left over. Most peoples sets were within .01V at .5A. I also ended up with 2 different lots. My first inclination was to keep the lots seperate, but there was much more variation within the lots than between the lots. Both had a simular bell curve, and the transconductance difference was less than I was able to measure. ( I tested at .48A and .74A, and all had the same slope within .01V/.26A)

A moving pool is a nice Idea, but I think getting in on a group buy or doing a group buy your self is much more practical.

As far as shipping around the testing gear, A good liear precision power supply is heavy and costly, and not something you would want to ship around.