25 Dec
2015
25 Dec
'15
10:56 a.m.
Years ago, we had EPROMs walk out of machine pin sockets from
vibration when the leads still had the out bent pins for auto machine
inserters. The problem when away when we straightened the pins.
Since then, I don't use one of those IC inserters on machine pin sockets.
I always first straighten the pins and manually insert them.
The IC inserters work well when doing PC boards and solder.
I never really understood why they'd walk out. With the tension
outward, you'd think it would pull them in tighter. After several times
putting them back in, we found no more problems with straightened pins.
Dwight
________________________________________
From: cctalk <cctalk-bounces at classiccmp.org> on behalf of tony duell <ard at
p850ug1.demon.co.uk>
Sent: Friday, December 25, 2015 9:32 AM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: RE: Piggybacking 74LS logic chips to confirm a suspected fault
In my experience, the type of socket you used is quite
unreliable and
will tend to develop bad contacts. There is a reason they are cheap... I
would use a turned pin type socket instead, e g an Augat socket. They
will not let you down. More expensive, but considering the work involved
in replacing a socket and the risk of messing up the PCB, the cost is
negligible.
My feeling is that for the sort of things most of us do here, a turned pin
socket is the most reliable of all. Yes, soldering the IC directly may have a
(slightly) lower risk of bad connections, but given that we do (or at least I
do) component level repairs, may want to remove ICs for testing, etc, there
is a risk to the PCB if the IC is soldered directly. As you say, cheap sockets
are cheap for a reason. They do develop bad contacts!
I don't think I've ever had a problem due to a bad contact at a turned pin
socket. Not saying they can't happen, but it's very rare.
I use nothing else, both for prototyping and for repairs.
-tony