16 Aug
2019
16 Aug
'19
3:57 p.m.
On 8/16/19 5:46 PM, Brent Hilpert via cctalk wrote:
On 2019-Aug-16, at 11:56 AM, systems_glitch via cctalk
wrote:
Back in the dark ages when MITS Altair (and dirt) was new....
Initial board were tin and not the fancy stuff either, sockets were
commonly gold. then came the occasional gold. What was nasty was the
gold over copper not gold/nickel/copper... Can you say Electromigration
and green plague? It was the cause of the shake well disorder as in
before powering up, pull the board and wipe the edge connector,
re-insert boards and it would be hopefully stable, maybe. I had to
retire that machine after about 2 years it was so flakey due to that.
By then the suspect boards were retired and never used again.
Looking back and having it to look at part of the issue was crappy gold
plating (looked good) and also some of the sockets did not have a hard
wipe or high spring tension both of which were likely causative.
I've not see that anywhere else. Dec connector blocks are hard wipe
and very good at what they do, make a connection. Even tin plate seems
to be no trouble at all.
Allison
On Fri, Aug 16, 2019 at 2:53 PM Paul Koning
<paulkoning at comcast.net> wrote:
>
>> On Aug 16, 2019, at 2:43 PM, systems_glitch via cctalk <
>>
>> I'm sure DEC wouldn't have bothered with hard gold plating if their
>> connectors were metallurgically incompatible :P The few busted DEC
>> connectors I've replaced did indeed have selective gold plating on the
>> contact surfaces. Most quality edge connector slots are similarly
>> constructed.
>
> It's been a while and I never looked in depth, but it most definitely is
> not true that gold is only compatible with gold.
>
> From what I remember, the detailed analysis involves an "electrochemical
> series", which has metals like sodium at one end, copper closer to the
> middle, and gold at or near the other end. Metals are compatible if their
> potential value differs by less than a limit. The limit depends on the
> environment; in an office you can have a larger limit than on a ship where
> you have salt spray, or a tire factory with lots of SO2 in the air.
>
> There are also some twists; I think stainless steel is compatible with
> many things thanks to the alloy ("stainless") properties. In fact, I
think
> the subject came up in connection with failure analysis of coin cell
> battery holders. The battery cases are stainless steel; the question is
> what contacts are acceptable. Gold is; there may be others but some things
> that are used in the market are not good choices.
You can look it up in an electronegativity chart
for a quick "will these
ruin each other" check.
I think a lot of this comes from the SIMM era in PCs, where folks were told
to only use gold-flash SIMMs in gold sockets, and only tin plated SIMMs in
tin plated sockets.
I've seen pieces of HP high-end lab equipment from thru the 60s that used tin plating
on the PCB edge fingers, mating into gold-plated edge connectors on the backplane.
Never quiet understood it, they (HP) were doing gold-plated edge fingers on other
equipment at the same time.