Chip socket

From Computer History Wiki
Revision as of 17:37, 16 December 2018 by Jnc (talk | contribs) (copy edit)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

A chip socket (usually plain socket) is a device which can hold an integrated circuit chip and allow it to make electrical contact on all the pins of the chip plugged into it. They are commonly used to hold DIPs, but sockets exist for other packaging formats as well.

Use of sockets allows failed chips to be easily removed and replaced; for this reason they are often found on DRAM main memory boards. However, some manufacturers, such as DEC, eschewed them, probably for a mix of reasons:

  • Cost (sockets do not provide any additional functionality, but increase the parts cost).
  • The potential for bad contacts between the pins on the chip, and the socket (on older boards, it is common to have to remove and replace chips to clear poor contact caused by corrosion).
  • Companies which had a large field service organizations would be better set up to take in 'broken' boards and repair them; allowing the customers to repair boards by chip replacement would make life easier for small manufacturers.

On DRAM boards, socketing the memory chips also allowed upgrading the boards when larger DRAM chips came out. A number of boards (e.g. the National Semiconductor NS23C and NS23M memory cards).

Sockets are often found on large chips, even on boards from manufacturers which didn't use them much; whether this is because it limited the chances of damaging them with heat when soldering them in, or because it's harder to remove large chips to replace them if/when they fail, is unclear.

Boards made using wire-wrap usually have all their chips in sockets; the wire-wrap pins come straight out of the socket, with no intermediary.