Issues using high current contacts at signal levels

[Mike Harrison]

Wonder if anyone has encountered this sort of problem....

I'm vaguely aware that there are issues with using contacts (e.g. relays) designed for high power at
low levels, due to buildup of oxides etc. that get punched through by current flow on closure.

I was wondering if this only applied to switching, or if it may be an issue for things like thermal
trips when spend all their life 'closed'.

The device in question is this :
http://www.sekicon.com/eng/product/st-22.html
Which mentions 'Min current 50mA'.

This is used in a commercial switchmode PSU (Mean Well USP-350), and is used in a signal path where
it only passes about 1mA.
I recently replaced one of these PSUs which had stopped working after 3 months in an installation,
but now works OK when tested on the bench. It was running well below capacity so I'm pretty certain
it never went over temp.

I have a strong suspicion that it might have been this switch that caused the problem. The contacts
have a large contact area and low contact force, so would probably be susceptible to oxide issues.
The switch seems somewhat sensitive to mechanical shock (I've tested another two unused PSUs and the
switches behave similarly.).

I am rather worried, as the installation contains 9 of these PSUs located 12 metres in the air at
the other end of the country, so further failures would be a major PITA.

I've sent an enquiry to the switch Mfr - I'd like to get their opinion before approaching the PSU
mfr.

The device is clearly aimed at high-current apps, and I'm sure there must be a reason they list a
min current - I'm mainly interested in opinions of whether this min current may be relevant to
non-switching apps.

I opened up the switch - contacts looked clean and bright contact material is silver colour, fairly
soft and about 0.25mm thick - doubt it's silver due to quantity and lack of any sign of tarnish.

 

[Phil Allison]

"Mike Harrison"

I'm vaguely aware that there are issues with using contacts (e.g. relays)
designed for high power at
low levels, due to buildup of oxides etc. that get punched through by
current flow on closure.

I was wondering if this only applied to switching, or if it may be an
issue for things like thermal
trips when spend all their life 'closed'.

The device in question is this :
http://www.sekicon.com/eng/product/st-22.html
Which mentions 'Min current 50mA'.

** That appears to be a fully sealed unit - so one would not expect to see
any corrosion on the contacts.

For silver (or silver alloy) to tarnish, it must be exposed to sulphides in
the air.

http://en.wikipedia.org/wiki/Silver_sulfide

Of course, it is always possible that something wrong with the plating, the
post plating cleaning or the material the case is made of the might
eventually contaminate the surface of the contacts.

The contacts being normally closed mitigates against this eventuality too.

I suspect the 50 mA minimum switching rating is just the maker playing it
safe.

FYI the majority of " dry switching " in audio is done with silver plated
contacts.




...... Phil

 

[Mike Harrison]

"Mike Harrison"

** That appears to be a fully sealed unit - so one would not expect to see
any corrosion on the contacts.

For silver (or silver alloy) to tarnish, it must be exposed to sulphides in
the air.

http://en.wikipedia.org/wiki/Silver_sulfide

Of course, it is always possible that something wrong with the plating, the
post plating cleaning or the material the case is made of the might
eventually contaminate the surface of the contacts.

The contacts being normally closed mitigates against this eventuality too.

I suspect the 50 mA minimum switching rating is just the maker playing it
safe.

FYI the majority of " dry switching " in audio is done with silver plated
contacts.

Thanks - the sealing is something I didn't think about.

I've subsequently found that this wasn't the problem with the PSU - what initially made me suspect
it was that the contact mass combined with the low contact force made it somewhat susceptible to the
sort of 'bang with a screwdriver handle' type event used to locate intermittent faults.