zero crossing relay problem

Hi,
I would like to switch on/off relay for 16A(rms), voltage of 230v. the switching must take action as close as possible to the zero-crossing point. I cannot use SCR relays, because of the price / heat.

So I build a simple relay driver – using transistor. The problem is that all simple relays (the cheap relays..) have tolerance of 10-20 ms operation time.
I built a zero – crossing detector and it works great – but since the tolerance of the relay is >10ms, ant the line frequency is 50hz (20ms), all the design is useless.

I have two questions:
1. How can I reduce the time to operate the relay, or at least stable it to constant value?
2. Is it possible to put protection element (like move) in parallel to the output pins of the relay, to protect the appliance from spikes (if the zero cross detection doesn’t work properly)

Thanks,
Nir

baba_sanfur said:

Hi,
I have two questions:
1. How can I reduce the time to operate the relay, or at least stable it to constant value?
2. Is it possible to put protection element (like move) in parallel to the output pins of the relay, to protect the appliance from spikes (if the zero cross detection doesn’t work properly)

Thanks,
Nir

Click to expand...

hi,
You could introduce some phase shift in the zero detect pulse.
If you assume the relay operate time is 10mSec and the mains period is 20mSec, then a 10mSec delay to the zero crossing signal would bring it close to the crossing point.

You can add MOV or resistor/capacitor 'snubbers' across the relay contacts to minimise sparking.

Hello there,


There are two identical posts so i wasnt sure which post to reply to,
so i guess this one is ok.


Ok, first off, it's hard to time a relay to turn OFF at a given time because
as you found out the de-energize time is a little unpredictable. BUT,
it might be possible to get the turn ON time to within an acceptable
limit, as long as you can change your circuit to turn the relay coil ON
to turn off the external circuit connected to the contacts.
The way to do this is to energize the relay with a voltage that is 5 to
10 times the rated voltage (it must be a dc relay too) for a short time
and that causes the relay to pull in much faster than it will open
when turned off.
To do this, you can energize the relay with a high voltage and regulate
the max current to an acceptable limit or perhaps use a spare contact
on the relay to turn the high voltage off after it closes completely.
I guess you could also limit current casually and then just switch the
high voltage off after a certain time period, like 5ms or something like that.

The idea is the same as what is used when driving stepper motor coils.
The coils are rated for some CONTINUOUS voltage like 5v, but they
are driven as high as 40v for short time periods in order to speed the
stepping action up appreciably. You could do the same thing with a
relay, driving the coil in the same way as a stepper.
Once the high voltage is switched off, the normal holding voltage would
take over. For a 10v coil relay this would be 10v, and the high
voltage might be 50v (current limited).

The idea for the relay coil and the stepper comes from this:

v=L*di/dt

solved for dt:

dt=L/v*di

and as this shows as v (voltage) goes up dt (time) goes down
for any given inductor L. For a relay coil it may be possible to
get the turn on time down to 2ms with a small tolerance. This
could be timed much more accurately with the zero cross circuit
plus time delay circuit, to turn the power off close to the
zero crossing.

If this isnt possible (turn the relay coil ON to turn off the circuit)
then it may be possible to use a dual coil relay where one coil
turns it off and another coil turns it on. These might be
more expensive though at 20 amps.
Another idea would be to 'roll your own' return coil that is
physically connected to the original relay, where that second
coil is used to speed up the opening of the original relay.
This requires a little knowledge of good mechanical design
however, but given that a really fast relay could be designed.