question re. stability of IC's in sockets.

[BrendanP]

I have a 40X1 running a sensor. In protos I've been using the 40X1 mounted in a socket which is then soldered to the board. This allows changing the part if need be.

Has anyone had the experience of a IC vibrating out of socket in a in field application?

Has anyone had problems with the legs of the IC not making a good electrical connection with the socket after a long period of time in service?

Of course I can avoid all this by simply soldering the picaxe right into the board, problem is of course if there a problem with the part the whole board is a write off. Ive tried unsoldering a 40 pin part and getting it out, it aint easy no matter how much wick and sodler sucker you give it.

 

[Dippy]

In work and hobby I have used hundreds without a problem ever. Applications included outdoor devices which were subject to shock and large temp variations.

The ic sockets were all 'known brand' types (e.g. Harwin,Aries,Tyco) and were of both the dual wipe and turned pin types. We never buy cheapies 'off of Ebay'.
We found the ultra-low profile turned-pin sockets (eg from RS) to be particularly good when testing. Maybe you get what you pay for.

The only problem ever experienced was corrosion between ic pin/leg and a dual-wipe contact caused by poor sealing in an outdoor product.

Having said that, I probably wouldn't use sockets in something subject to vigorous continuous vibration - unless there was a mechanical means of making sure the chip din't fall out. I can't see any problems with the actual contact/mating side of things though.

As I've never used cheap tat I can't comment on it.

PS. Save up the pennies and buy a desoldering station, or, design your boards so they don't blow I've found that course of action quite beneficial.
-

 

[moxhamj]

I can't imagine it vibrating out, but if unsure, dip the whole thing in varnish after it is working.

Re corrosion - I have some 20 year old sockets that don't solder very well due to corrosion, but that is exposed metal. If a chip is in a socket then metal is touching metal and won't be corroding.

In general terms, try to keep moisture out.

I tend to solder in most ICs but put picaxes in sockets. I've invested in machined sockets more recently and I think they are worth the extra price.

 

[hippy]

I've seen chips fall out of sockets, the bent-wire sockets not turned pin. Vibration does it but so can high-G drops and jolts.

Ideally use turned or other high-quality sockets. One solution is to mount a fairly firm rubber block to the chip so the casing keeps it pushed in place, another under the board if it's going to vibrate. Another solution is not to fix the board to the enclosure but mount it in a sandwich of rubber or use rubber 'eyelets' to dampen vibrations between mounting bolts and board.

Soldering the chip is the simplest solution. Removing even a 40-pin isn't too hard. Cut the body from the legs then unsolder each leg individually.

 

[BeanieBots]

Only problems I've ever seen have been with ham fisted "couldn't care less" temporary production staff who 'somehow' manage to curle the IC legs around back towards the IC. It results in light contact with the socket which often passes the board test but fails shortly afterwards.
Always use turned pin sockets. The folded spring type can have a habbit of holding onto the legs during removal and the type of operator described above can manage to remove the IC but leave the legs in the socket

 

[Peter M]

BB, I must have been the recipient of one of these some years back... damn there hard to find (the curled legs that is.. the ham fists seem to be dime a dozen) but you do get what you pay for.

I have had a chip vibrate out of a dual wipe socket, but is was mounted on an off road race buggy (these this are like paint mixers with added vibration just for good measure). It was a pic used for controling timing of a variable camshaft.

One solution is to mount a fairly firm rubber block to the chip so the casing keeps it pushed in place, another under the board if it's going to vibrate

another way to get your "rubber block" on top of the chip the right size (so the lid keeps it in place) is to use hot melt glue on top of the chip, put a piece of paper on top of that, then put the lid on (paper stops glue sticking to lid so you can open it again).

 

[BrendanP]

Thanks for the imput once again guys. It's like having a personal electronic engineering consultancy service at my beck and call. I'm very gratefull.

Hippy your tip re. cutting the ic out is a good one.

I'll probably go the solder the thing in and be done with it route espescially hearing that they can vibrate out. I can't eliminate the prospect that someone will clamp this thing to a D9 and then flog down a few thousand acres of bush over the next few months and then expect it to work flawlessly.

Dippy you're a hard man.... My boards are second to none! (just ignore the couple of thousand dollars worth of stuff ups and mistakes Ive got laying in the box under the bench).

 

[Dippy]

"In general terms, try to keep moisture out."- how do I answer that without meaning to sound sarcastic ?!
Really? I hadn't thought of that

"If a chip is in a socket then metal is touching metal and won't be corroding."
- if only.


PS. Go for good Q turned-pin type. They are pretty rock-solid. I know you'll buy the Best Quality. But would I trust them in the Space Shuttle, no.

PPS. When I said "mechanical means" I meant some method of keeping them in with some sort of pressure applied via a compliant cushion. Maybe like hippy and his rubber thing.

 

[slurp]

Many of the board I've seen for use in vibration environments have the components strapped in. I've often used cable ties on bread boarded robots to help keep things in place before I get round to PCBs.

regards,
colin