Electronics: why does the technical data for many switches and relays
specify a *minimum* operating voltage and current, as well as a
maximum?  For example, the data sheet at
http://www.farnell.com/datasheets/36353.pdf shows a keyboard switch
with a *minimum* switch voltage of 20mV and a *minimum* switch current
of 1mA. The datasheet for a relay,
http://www.farnell.com/datasheets/17343.pdf, indicates a similar
"minimum switching capacity".

---

Clarification of Question by doctordavid-ga on 11 May 2005 13:02 PDT

Wizawuza: I think that's part of it, but not the whole explanation.
Consider this example. I have an input to a logic gate pulled up to
+5V by 100K. I short the input to ground with a switch, specified to
have a contact resistance of < 100mOhm - so there ought to be no
problem "seeing" the switch. (It would need a contact resistance of
kilo-ohms to cause a problem). But there is only 50uA flowing in the
switch, and it has a minimum rating of 1mA. Am I going to get
unreliable switching?   And, supplementary: is it the *switching* of
low current that is advised against, or merely the *carrying* of low
current?

I believe the reason is due to the contact resistance and noise. 
Since there is contact resistance, such a small value of voltage and
current may be "lost" in the noise of the contact resistance.  If that
value is lost, you'll never see the input on the output.

The minimum current rating for the switches are to make sure that there
is no erratic variations in the contact resistance when the switch is put
into "on" state, especially under low current switching.Basically you have
better noise margin due to stable contact resistance, when the current is more
than the minimum allowed value. The low value of the current also depends on
the switch construction(type of contact coating(gold plating prevents formation
of oxidising film), mechanical pressure on the contacts(reduced and stable
contact resistance)). Carrying of low currents is beneficial to the 
switching contacts, but switching under such conditions is not always 
reliable.The contacts in relays are further influenced by the susceptibility
to mechanical shock, variation in contact pressure(between active/holding 
states).


I hope the above explanation helps in getting a better understanding
on the minimum current requirements for a switch.

Regards
ldavinci-ga

Thoughts,

From google search:
://www.google.com/search?q=relay+minimum+voltage+rating&hl=en&lr=

Voltage, off state. In solid state relay, the following determine
whether the relay will stay off under each load voltage condition:
(1) Critical rate of rise of commutation voltage, dv/dt. The maximum
value of the rate of rise of principal voltage which will cause
switching from the off state to the on state.
(2) Maximum off state voltage (VDMax)(RMS). The maximum effective
steady state voltage that the output is capable of withstanding when
in off state.
(3) Maximum rate of rise of off state voltage, dv/dt. The rate of rise
of the off-state voltage which the output can withstand without false
operation.
(4) Minimum off state voltage (VDMin)(RMS). The minimum effective
voltage which the relay will switch.
(5) Non-repetitive peak voltage (VDSM).The maximum off-state voltage
that the output terminals are capable of withstanding without
breakover or damage.
http://www.leachintl2.com/english/english2/vol6/properties/00009.html
----

That would be solid state relay... an analog relay, across a coil, I
believe, it is stay voltage and current.  Meaning, there is a spring
which pulls the contacts back to the 'off' state, position, and to
counter the force of the spring the coil must drop below the minimum
voltage and current.  There is also some force applied by the current
traveling through the contact that the spring has to overcome in order
to open the circuit (that would be very small but still should be
accounted for in the purest sense, which is what I always thought the
ratings where given to show).

I look forward to seeing what the full answer is to the question.

When exposed to the environment, a surface film (from smoke, air
pollution, etc) is deposited on the metal contacts of the mechanical
relay or switch. Additionally, for contacts other than gold, a surface
film may be formed as a result of oxidation or corrosion.  Even if the
metal contacts were sealed in a vacuum, the flow of electrical current
causes arcing whenever the metal contacts meet.  This arcing can cause
pitting of the contact surface, i.e. tiny balls of metal are melted
and re-deposited.

If the surface film (oxide, etc) on the metal contact is an insulator,
a minimum applied voltage is required to "punch through" this very
thin layer of insulation.  This punch-through phenomenon may be
visualized as a lightning storm, occuring on a microscopic scale.  To
give a sense of scale:  the breakdown voltage of air (an insulator) is
25,000 volts per inch (between two metal surfaces which are flat, like
the switch contacts), which scales to 25 millivolts per microinch.

larryg999-ga is on the right track.

Silver contacts are well known to experience problems when switching
low voltages and currents (signals). The current required to remove
the oxide layer and establish a metal to metal contact and hence an
electrical connection is also described as a wetting current. It is
normally in the order of a few milliamps.

If you wish to reliably switch signals in the mV uA region then a gold
contact is likely to give a much better performance. Wetting currents
are in the uA region.

Less than this then a wiping contact or a mercury wetted reed would be required.

I found this link looking for the wetting current on a specific switch
where the customer has designed in a silver switch and we believe it
should probably be gold as we are getting poor performance and the
current is ~1mA.