[time-nuts] HP 5370A socket problem

[Magnus Danielson]

Didier skrev:
> One is S/N 2128A01306 and the other is S/N 2213A01361, so they were quite a
> bit apart.
> 
> I have heard of several units with similar problem, on this mailing list in
> particular, which is the reason that prompted me to suspect the sockets in
> the first place. I do not think the problem was isolated.
> 
> It is quite possible that the history of these particular units (possibly
> stored in less than ideal conditions) exacerbated the problem. Tin wiskers
> have the well deserved reputation of seemingly random occurrence, and socket
> issues in general, other than tin wiskers, are very environment dependant.
> 
> Regarding gold-lead issues, I have only seen that problem (quite
> extensively) in military hardware that operates undre extreme conditions of
> temperature and humidity. Way back, we had to retrofit over 100 expensive
> pieces of equipment because of that. I have never seen it any other place. I
> am not worried, but I thought I would mention it. Most commercial hardware
> now avoids gold anyhow because of cost. Military specs require to remove the
> gold from the solder area by pre-tinning in a solder pot, which has to be
> replaced regularly.

We have had one issue at one time with gold. It only occured due to a 
particular batch of boards having way too much gold in them. All other 
boards have been without problems and none of all our boxes in the field 
have not given any indication of a problem over the years.

While not running extreme temperatures, for our more normally 
temperatures gear, using gold have not been a problem.

Cheers,
Magnus