Tin whiskers in Grayhill switches

Well, that sucks. At their prices they should be the best switches on earth.

--sp

rote:

roduce an instrument w/ 7 switches, purchase about 1000 switches a year. Un its come back because of failures, 2-5 years old. Internally, a switch cont act gets shorted to the body, so the signal on that contact gets shorted ou t to the panel the switch is mounted on.

ee the whiskers? Our attempts to open the switch just ends up w/ a mess.

) is an old

that they

e original

just

switch to

h tin

ven

Right, I should have said nothing. (for whatever reason you rub me the wrong way.)

George H.

e:

produce an instrument w/ 7 switches, purchase about 1000 switches a year. U nits come back because of failures, 2-5 years old. Internally, a switch con tact gets shorted to the body, so the signal on that contact gets shorted o ut to the panel the switch is mounted on.

see the whiskers? Our attempts to open the switch just ends up w/ a mess.

e) is an old

that they

he original

or

just

switch to

ue

th tin

ven

Right, I've got switches with a forest of whiskers. scary.

GH

Very little current. I don't have the numbers but the ohmmeter section of my Fluke 12 DMM would reliably cause it to blow. NASA says it requires over 50ma to blow some whiskers: "Stable short circuits in low voltage, high impedance circuits. In such circuits there may be insufficient current available to fuse the whisker open and a stable short circuit results. Depending on a variety of factors including the diameter and length of the whisker, it can take more than 50 milliamps (mA) to fuse open a tin whisker."

Who are you going to believe, me or NASA? (Don't answer that).

We had problems with tin whiskers inside silver and tin plated RF cavity filters. I don't recall the exact metallurgy:

The typical air gap was about 3mm and took about a year for a whisker to form. Only about 20% of the filters exhibited the problem. I got stuck with the usual "fix it, but don't change anything" problem. My compromise was to drill a hole in the cavity, have the customer blow compressed air into the cavity, and then cover the hole with a plastic hole plug. That literally blew away the problem, for a year.

The clue here is that tin whisker only form in still air or no air (vacuum). If the air is moving, then they just crumble and fly away. The limits the problem to anything that is sealed, enclosed, evacuated, packaged, or lacks a cooling fan, such as inside the Grayhill switch.

Incidentally, I've only seen a tin whisker with a microscope once. It was only for a few seconds and it disappeared when I exhaled. They are VERY fragile.

I had some Xilinx chips that got tin whiskers that could be seen with the naked eye, and my eyes are not all that great, even!

They plated the lead frames with pure tin, assembled and wire bonded the chips, molded the package and then sheared and bent the leads into the gull wing shape. This put stress in the bends. They expected you to reflow the chips hot enough to anneal the pure tin! I think you had to get to 270 C or some totally insane temperature to properly do this on the entire board. I use tin/lead solder on these particular boards (customer request) and reflow at about 230 C. So, I got some truly AWESOME whiskers!

I've seen them in some other cases where they were really fine and you could probably destroy them by blowing hard on the board.

Jon

Yeah, I don't like rubbing guys the *right* way. :-P