Why Do Relays Stick?

[Irish_Alley]

i "had" door poppers in  my 79 with cheap relays. someone popped the door and the relay stuck open and burnt up the wiring. i didnt put them in so it wasn't protected by a fuse. so the popper also went u. talked to a couple friends about relays and they said the cheap ones will do that

[rich weyand]

Relay contacts can burn in the closed position when driving inductive loads like motors and especially coils.  When the contact breaks, the inductive load tries to keep the current going and you get a voltage spike across the contacts that burns the contacts.  Enough cycles and the relay contacts weld together in the closed position, leaving the circuit continuously on.

That having been said, there is a way to defeat the inductive spike of a relay breaking the circuit to an inductive load.  When wiring the relays, put a diode with the same amperage rating as the current draw across the load in the opposite direction of the applied voltage.  When the current through the inductive load is broken, it winds down through the diode and kills the voltage spike.  It is called a snubber diode.  I've been doing this for forty years on any inductive load and can't recall ever having a relay failure, even with cheap relays.

See the lower right circuit in this diagram.  The switch on the right is your relay contacts, the inductor at the top is the door lock solenoid or window motor, and the applied voltage E (V) on the left is your battery connections.  This will do for the diode: http://www.radioshack.com/product/index.jsp?productId=12673865&filterName=Type&filterValue=Diodes

[bd]

Caution:  Don't install a diode across any DC motor, including a power window motor, with which the polarity on the circuit reverses to change motor rotation/actuation direction or you'll 'pop' the diode like a fuse. 

[rich weyand]

Caution:  Don't install a diode across any DC motor, including a power window motor, with which the polarity on the circuit reverses to change motor rotation/actuation direction or you'll 'pop' the diode like a fuse. 

Ack.  That's very true, for both the power locks and the power windows.  I did a lot of one-way motors and twin-coil solenoids in my day.  For reversible circuits, the snubber circuit can be the R/C circuit in the upper left hand corner of the above diagram.  For a 5A motor at 14 volts, the values are C=2.5uF and R=0.15ohm.

The circuit on the lower left will also work.  This varistor will do: http://www.radioshack.com/product/index.jsp?productId=2062574.

[VileZambonie]

Use circuit protection and don't sit on your switches and you'll be just fine.

[rich weyand]

Use circuit protection and don't sit on your switches and you'll be just fine.

No, the arcing across the switch points when breaking the current through an inductive load is a very real problem regardless of circuit protection or user action.  That's why big relays are rated in three ways: carrying current, switching current (resistive load) and switching current (inductive load).

[Irish_Alley]

rich i see what your talking about. and would have to give it a try. circuit protection is good but  it still wont prevent the relay from freezing "on".

[Irish_Alley]

from what i understand those relays will work and i know they will. but without the diode they will eventually arc and fuse themselves hot. cause more problems if you dont have circuit protection. if you use the diode in the wrong direction you will pop that. but the part im confused about is if you use the diode in the incoming power for the relay how would it be reversed if there is a seperate relay for each function

[rich weyand]

For arc suppression at the switches, use the varistor, then you don't have to worry about polarities and reversing and the like.

[bd]

from what i understand those relays will work and i know they will. but without the diode they will eventually arc and fuse themselves hot. cause more problems if you dont have circuit protection. if you use the diode in the wrong direction you will pop that. but the part im confused about is if you use the diode in the incoming power for the relay how would it be reversed if there is a seperate relay for each function

The diagrams above are with regard to a circuit's switched inductive load (i.e., door popper solenoid, door lock solenoid, window motor, etc.).  The bottom two diagrams also apply to relay coils and will suppress reverse current surges resulting from collapse of the magnetic field surrounding the relay coil when the relay is deenergized.  Each inductor (or coil) in the circuit should be 'snubbed' independently.  That is, each switched load should be 'snubbed' as discussed, and every relay coil should be snubbed or clamped separately.  'Load' snubbing protects the relay contacts from arcing, while 'relay' clamping protects the control circuit switch (e.g., power window switch, door lock switch) from arcing.  Generally, relay coils (being part of the remote switches that control the inductive loads) don't see "reversing" control current in a DC circuit, so a clamping diode works well; a ~220 Ω resistor across the relay coil in a low voltage circuit would work equally well with no concern for polarity.  Do you follow?

[Irish_Alley]

lol no. i think i understand the relay part, but whats snubbed? and relay coils change the flow so thats were it would matter where you put the diode

[bd]

lol no. i think i understand the relay part, but whats snubbed? and relay coils change the flow so thats were it would matter where you put the diode

I used the term "snub" to tie to Rich Weyand's posts.  Essentially, 'to snub' is industry jargon synonymous with "to suppress," "to dampen" or "to minimize."  I also used the term 'to clamp,' which means "to limit to a specific threshold amount" (which in the case of silicon diodes is ~0.6 volt).  Both terms imply similar purpose within the context of our discussion - to bleed the energy stored in an inductor to zero.

Inductors, whether used in a solenoid for linear motion, in a motor to rotate an armature, or in a relay for arcuate motion of a hinged arm, obey the exact same physical laws.  When current stops flowing through an inductor, its magnetic field collapses and a reverse current is generated in the wire.  That reverse current has opposite polarity to the current that created the magnetic field in the first place.  The same methods of clamping the voltage to prevent arcing of switch contacts applies to all inductors, motors, relays, solenoids, alike.  But, the case of a motor or solenoid that reverses direction by reversing the polarity of the applied current, requires a method of clamping that can tolerate the change in current polarity without damage to the components.  In this case, a diode must be avoided and substituted by a varistor, resistor, or RC filter.

[rich weyand]

Close, bd.  An inductor has momentum.  When voltage is applied, current does not flow immediately like it would in a resistor.  It has to build up, to accelerate, if you will.  Similarly when voltage is cut off, the current wants to keep flowing.  When the switch opens in an inductive circuit, the current wants to keep flowing in the same direction.  and piles up a voltage on the switch points.  That voltage appears across the points in the opposite direction.

The definition of inductance is related to the rate of change of the current due to an applied voltage as V= L (dI/dt), where L is the inductance.  The problem is, when you open the switch, the rate of change of the current, dI/dt, is basically infinite because it stops abruptly, and negative (the current is decreasing).  Thousands of volts appear across the switch points for a brief instant.  10,000 volts will arc across a centimeter of dry air between pointed electrodes, so this huge induced voltage across the opening switch points creates a large spark that degrades the switch points.

That's why the varistor works.  Most varistors are non-conductive up to a couple hundred volts, then conduct.  When the voltage surge from shutting off the current appears across the varistor, the varistor conducts all of it away except a couple of hundred volts, which is not enough to spark and damage the switch points.

So just hook a varistor across any inductive load -- here, the window motor, the door lock plunger, and the coils of the relays -- and the switches and relay contacts -- even cheap ones -- will last until the points mechanically rub themselves away.  No arcing damage.

Why didn't GM put the varistors in there?  Cost of the part and cost to install.  Simple as that.  In old-style points distributors, though, GM and other mfrs did use a condenser as a snubber to increase the life of the points, as the primary winding of the ignition coil is an inductive load.  Without the condenser, ignition points don't last very long.

Rich (B.S. and M.S. degrees in Physics)

[Irish_Alley]

so when you wire them on one side get soldered to b+ then the other side gets Ground?
does each "application" call for a different varistor (depending on amps)

[nlauffer]

OK.  I only took basic electricity in College.  And as a Lineman, I was only worried about not blowing my arms off.  The discussion about the varistor, diode, resistor, etc, has me confused.  Is one of these necessary to wire up the system safely or a really good idea for longevity of the system?