Microswitch Wetting Current

[Tim Auton]

Hi,

I'm looking to solve a switch bounce problem I've got with a couple of
mice (Microsoft Intellimice, it's a known hardware problem - a single
click registers as a double-click). Replacing the microswitches solves
the problem for a month or so, but it always comes back. I was just
going to put together a deboucing circuit with a 555 (because I've got
one to hand), but while skimming through the relevant bits of The Art
Of Electronics for inspiration (Hi Win, great book) I noticed a brief
mention of switch wetting current. This is a new one on me and I
couldn't find much information on it elsewhere, nor the datasheet for
the (Omron) microswitch in question. Curious, I measured the current
through the microswitch when closed - a mere 0.06mA.

So, my question is might this tiny current (rather, lack of current)
be the cause of (what looks like) excessive bounce? If it could be the
cause might slipping a resistor in series to get a bit more current
flowing solve it? Not having the datasheets for the mouse chip I'd
probably risk frying it by asking it to source a couple of mA, but the
mouse is buggered anyway, and I've got two


Tim

 

[Gareth]

Hi,

I'm looking to solve a switch bounce problem I've got with a couple of
mice (Microsoft Intellimice, it's a known hardware problem - a single
click registers as a double-click). Replacing the microswitches solves
the problem for a month or so, but it always comes back. I was just
going to put together a deboucing circuit with a 555 (because I've got
one to hand)

Have you tried adjusting the double click speed software? In Windows that's:

My computer -> Control Panel -> Mouse.

Your OS may be different, but I'm sure you can adjust it.

Gareth.
--

 

[Tim Auton]

Have you tried adjusting the double click speed software? In Windows that's:

My computer -> Control Panel -> Mouse.

Your OS may be different, but I'm sure you can adjust it.

Thanks Gareth, but it's kind of the opposite problem - my OSes allow
you to set the maximum time it will count two clicks as a double. My
problem is that a single click of the mouse button registers as two
clicks in ultra-quick succession. If I lived permanently in Linux-land
I'd solve it in software, but sadly Windows is a necessity some of the
time and I really don't want to go writing drivers for that. I'm
amazed Microsoft haven't through, I'm sure it would save them a lot of
money in replacement mice.


Tim

 

[John Fields]

http://www.ibiblio.org/obp/electricCircuits/Digital/DIGI_4.html#IDX133
With all this discussion concerning the reduction of switch contact arcing,
one might be led to think that less current is always better for a
mechanical switch. This, however, is not necessarily so. It has been found
that a small amount of periodic arcing can actually be good for the switch
contacts, because it keeps the contact faces free from small amounts of dirt
and corrosion. If a mechanical switch contact is operated with too little
current, the contacts will tend to accumulate excessive resistance and may
fail prematurely! This minimum amount of electric current necessary to keep
a mechanical switch contact in good health is called the wetting current.

Normally, a switch's wetting current rating is far below its maximum current
rating, and well below its normal operating current load in a properly
designed system. However, there are applications where a mechanical switch
contact may be required to routinely handle currents below normal wetting
current limits (for instance, if a mechanical selector switch needs to open
or close a digital logic or analog electronic circuit where the current
value is extremely small). In these applications, is it highly recommended
that gold-plated switch contacts be specified. Gold is a "noble" metal and
does not corrode as other metals will. Such contacts have extremely low
wetting current requirements as a result. Normal silver or copper alloy
contacts will not provide reliable operation if used in such low-current
service!

 

[Tim Auton]

I'm looking to solve a switch bounce problem I've got with a couple of
mice (Microsoft Intellimice, it's a known hardware problem - a single
click registers as a double-click). Replacing the microswitches solves
the problem for a month or so, but it always comes back. [wetting current]
Curious, I measured the current
through the microswitch when closed - a mere 0.06mA.

So, my question is might this tiny current (rather, lack of current)
be the cause of (what looks like) excessive bounce?

http://www.ibiblio.org/obp/electricCircuits/Digital/DIGI_4.html#IDX133

Thanks Baphomet, I understand the theory a bit better but I still
don't know if this might be my problem or not.

What I'd still like to know is:

Is 0.06mA sufficient wetting current for a "typical" microswitch? If
not, what is? Is this kind of thing on datasheets? I can't get hold of
the one for this switch, Omron's site says "coming soon" or something
like that.

and

If 0.06mA is insufficient wetting current would the switch's
degredation manifest itself in a way that looks rather like worsening
switch bounce?


Tim

 

[Baphomet]

Hi,

I'm looking to solve a switch bounce problem I've got with a couple of
mice (Microsoft Intellimice, it's a known hardware problem - a single
click registers as a double-click). Replacing the microswitches solves
the problem for a month or so, but it always comes back. I was just
going to put together a deboucing circuit with a 555 (because I've got
one to hand), but while skimming through the relevant bits of The Art
Of Electronics for inspiration (Hi Win, great book) I noticed a brief
mention of switch wetting current. This is a new one on me and I
couldn't find much information on it elsewhere, nor the datasheet for
the (Omron) microswitch in question. Curious, I measured the current
through the microswitch when closed - a mere 0.06mA.

So, my question is might this tiny current (rather, lack of current)
be the cause of (what looks like) excessive bounce? If it could be the
cause might slipping a resistor in series to get a bit more current
flowing solve it? Not having the datasheets for the mouse chip I'd
probably risk frying it by asking it to source a couple of mA, but the
mouse is buggered anyway, and I've got two


Tim

http://www.ibiblio.org/obp/electricCircuits/Digital/DIGI_4.html#IDX133
With all this discussion concerning the reduction of switch contact arcing,
one might be led to think that less current is always better for a
mechanical switch. This, however, is not necessarily so. It has been found
that a small amount of periodic arcing can actually be good for the switch
contacts, because it keeps the contact faces free from small amounts of dirt
and corrosion. If a mechanical switch contact is operated with too little
current, the contacts will tend to accumulate excessive resistance and may
fail prematurely! This minimum amount of electric current necessary to keep
a mechanical switch contact in good health is called the wetting current.

Normally, a switch's wetting current rating is far below its maximum current
rating, and well below its normal operating current load in a properly
designed system. However, there are applications where a mechanical switch
contact may be required to routinely handle currents below normal wetting
current limits (for instance, if a mechanical selector switch needs to open
or close a digital logic or analog electronic circuit where the current
value is extremely small). In these applications, is it highly recommended
that gold-plated switch contacts be specified. Gold is a "noble" metal and
does not corrode as other metals will. Such contacts have extremely low
wetting current requirements as a result. Normal silver or copper alloy
contacts will not provide reliable operation if used in such low-current
service!

Hope this helps

 

[CFoley1064]

Thanks Gareth, but it's kind of the opposite problem - my OSes allow
you to set the maximum time it will count two clicks as a double. My
problem is that a single click of the mouse button registers as two
clicks in ultra-quick succession. If I lived permanently in Linux-land
I'd solve it in software, but sadly Windows is a necessity some of the
time and I really don't want to go writing drivers for that. I'm
amazed Microsoft haven't through, I'm sure it would save them a lot of
money in replacement mice.


Tim

Hi, Tim. The concept you're looking for is "dry contact". Most switches will
require a certain amount of current to refresh the surface of the contacts.
This can be improved somewhat by using gold contacts, but you will still get
this problem. Switches rated for dry contact have self-wiping contacts, with
the combination of contact metal chosen and switch construction resulting in a
refresh of the surface with friction each time the switch is closed. In fact,
even self-wiping contacts work better with a li'l bit of current, but it's not
as necessary.

For your Omron replacement microswitches, you probably just got old switches
(the contacts oxidize in storage, too). If you can't find dry contact
microswitches for this application (doubtful in "mouse"-sized switches without
paying $$), just try going with another batch from another mfg (and look at the
date code). Either that or, if you're desperate, you might try removing the
Omron switches from the circuit and switching maximum rated load through them a
couple of dozen times. That will refresh the contacts, and you can reinstall
them -- they'll be a lot happier for having switched some real current for once
in their lives.

In days of yore (mid- to late-'80s), mice were manufactured to last, and some
manufacturers did use dry contact microswitches. I've still got a couple of
'80's vintage mice that still are working on DOS machines in the lab at work. I
just have to clean out the mouse pad fuzz every once in a while. But over
time, people have become accustomed to cheap, throwaway mice, I suppose. Those
old mice were pretty expensive. If you've got any '80s vintage mice in the
closet, you might want to cannibalize those switches ;-)

Good luck.
Chris

 

[Tim Jackson]

What I'd still like to know is:

Is 0.06mA sufficient wetting current for a "typical" microswitch? If
not, what is? Is this kind of thing on datasheets? I can't get hold of
the one for this switch, Omron's site says "coming soon" or something
like that.

Small silver contacts need about 10mA to combat tarnishing but for gold,
microamps should be fine. Gold contacts in my experience (in industrial
relays) usually fail through wearing away the plating or through
contamination build up, rather than lack of current. Gold doesn't tarnish
like silver, it adsorbs a coating of pollutants and a wiping action
effectively removed this in normal service, eliminating the minimum current
requirement.

and

If 0.06mA is insufficient wetting current would the switch's
degredation manifest itself in a way that looks rather like worsening
switch bounce?

No, it fails to connect at all. You would get far more missed click events
than double click events You can get a 'bounce' effect from a self-wiping
switch that is pitted or contaminated, as the contact rides over an
insulating patch. You must have clean metal surrounding a dirty patch, not
vice versa. At a guess your switches are gold plated with a very thin
(cheap) gold layer which has worn away leaving a patch of base metal which
has then oxidised. It is the manufacturer's "electrical life" figure that
you need to consider, 10,000 ops is not uncommon and wouldn't be a long life
for a mouse.

You might be able to give your mouse a new lease of life by soldering a
small capacitor across the switch contacts to slow down circuit response.
I'd try about 1uFd.


Tim Jackson

 

[Baphomet]

I don't mind your saying so at all. Unfortunately, I copy and pasted it
right from the website link.

 

[John Fields]

Hi, Tim. The concept you're looking for is "dry contact".

---
Contrary to popular belief, "dry contact" or "dry switching" refers to
the make of the contacts before current is passed through them and the
cessation of the flow of current through them before they are opened.

This is in contrast to "hot switching", which refers to the contacts
making while there is a voltage across them, and breaking while current
is flowing through them.

Gold is usually chosen for low current dry switching in non-hermetic
relays because of its resistance to oxidation/corrosion and because no
current will ever cause the electrical disruption of the contact
surfaces. It can also be used for hot switching but, as Baphomet
pointed out, a wetting current will have to be used to ensure that, once
the gold plating (or cladding) is displaced any contamination which
forms on the base metal will be "blasted off" every time the relay
switches. However, in my experience I have always found that contacting
the relay manufacturer is the preferred method of determining the
contact material/wetting current required for any given application.