On Thu, 28 Feb 2008 11:02:31 -0500, Fred Bloggs
wrote:



The circuit below simulates fairly well. You don't want to use a
conventional snubber across the contacts because on opening the relay
coil voltage reverses and adds to the 220VAC source. Placing a snubber
in shunt with the coil with peak current limiting resistor as shown
increases operating power by about 10% but tends to maintain the contact
voltage and results in a very slew rate limited 0.5V/us contact voltage
peaking in the 450V range. There should be no arc at all with this
circuit, with or without zero crossing logic. K1 are the 12V relay
contacts and K2 is the 220VAC coil. I did not consider contact bounce on
closure, will leave that to you.
View in a fixed-width font such
as Courier.


.
.
.
.
. 220VAC
. o o
. | |
. | |
. | - K1
. | -
. | |
. | |
. | R 100R, 1.5W
. | |
. | |
. | +-----.
. | | |
. | - |
. | |\| |
. | K2|\| ===
. | |\| | 47nF
. | |\| |
. | - |
. | | |
. | | |
. '------+-----'
.
.
Placing impedance in series with the working solenoid could produce a
reduction in speed/dropout performance in the armature of the relay
switching the main working load. (not shown in the above drawing)

I'm not sure how you modelled the relay coil, but if it used a linear
inductor, it will not likely reflect actual performance. A relay
drive coil is coupled to a mechanically changing magnetic circuit.

As the OP already has a cost-free solution involving programmed timing
adjustments, perhaps it's best to let the issue drop?

RL