Relay vs christmas lights

Relay specs.
G3MB-202PEG-4-DC20MA-Omron-datasheet-111010.pdf (69.0 KB)

The output specs quote a minimum as well as a maximum spec.

Sound like Triac leakage current.

Tom... :grinning: :+1: :coffee: :australia:

The leakage rating for this SSR (Omron G3MB-202P) is 1.5ma max at 200VAC.

Assuming the actual is 1ma actual it is likely the LEDs driven with this current will be able to be seen.

So Yes it is normal.

Thank you, guys!
First: I connected in parallel with the led strip an incandescent light bulb, as you suggested, and the led didn't light on.
Second: @LarryD : I added a 10uF capacitor between 5v and GND in the circuit, do you thonk I have to add 3 100nF ceramic capacitor, 1 for each shift register?
Third: @LarryD : connecting the ground to the GND terminal of the relay module on the left is the same than connecting the GND to the power supply, don't you agree?

However: what could I do to avoid this behaviour? I choose these relay module 'cause of the speedy of changing on/off state. Christmas lights has to be fast, in several cases I had 0,1 secs interval time from on and off (or vice-versa) state. I have also modules based on Songle SRD-05VDC-SL-C relays, I don't know these can fit up my necessities.

Could someone help me again, please?
Thanks a lot,


I'm confused, in both statements you put an incandescent bulb as the mains load.
In the first instance it worked without the LED strip.
In the second instance, basically the same configuration, but with the LED strip in parallel and the LED strip didn't work?

Is that right?

Tom... :grinning: :+1: :coffee: :australia:

No, excuse me Tom if one made confusion, it was me.
As you can see in the first picture, I have 24 output to connect (hopefully) as many christmas lights.
In one output I connect a christmas light and light on weakly when relays are opened, and normally when relays are closed. This is the strange behaviour that you have explained with the riac leakage.
In a second output I connect (to test purpose only) a incandescent bulb light. In this case, when the relay id opened the lamp is off, when closed the lamp lights on. This is the normal behaviour I hoped also for the led strip.

Someone in the thread tol me to test a bulb light in parallel with the led strip (therefore on a single output): in this case both lamp and led stripe are off when the relay is open and light on when the relay is closed.

I hope I've explained myself.
Thanks again for helping me.

Don't MOSFET based relays leak a whole lot less?

I am going with Leakage, from the board documentation:

  • The output control terminal must be connected to the circuit, and the working current must be> 50mA, so that the solid state relay can be disconnected normally

Unless your load current exceeds 50 mA you will get the described symptoms. The experiment with placing a lamp across the load demonstrates this. Rather than always having a lamp out there for each SSR a resistor can be used of enough wattage for each SCR output. Figure you have a 2.0 Amp SCR so you may want to place a 2.4 K resistor which will always be a 100 mA load.



So your problem is identified, leakage through the Triac in the SSR.
In the open position, the lamp resistance is very low and so the majority of the leakage current flows through the lamp and not the higher resistance of the LED strip.

The 1.5mA leakage is not enough to light the lamp.

Tom... :grinning: :+1: :coffee: :australia:

The TRIAC will not "leak".

The SSR no doubt contains a snubber.

@Paul_B : I really don't know what a snubber is...

However, I think I've learned some things from this thread (thanks again to you all):

  1. I have to insert a decoupling ceramic capacitor (100nF) for every shift register close to the vcc of the IC; and I have to leave where it is the electrolytic 10uF capacitor

  2. to avoid ssr leakage that lights on the led strip I have two chance:
    a) insert a resistor between the AC outputs (correct? which kind?) in parallel with the led strip
    b) choose a led strip that requires more corrent draw

Thanks for your patience.

So you figured out the green component is in fact, a 2 Amp thermal fuse? :sunglasses:

But the resistor draws power, so an incandescent bulb it traditionally a practical way of dissipating that power.

It may still light up.

Third option - just ignore it.

Yes, the F designation gave it away.


My thoughts:

  1. Yes the 100 nf is required at the IC (as close to the Vcc, Vss pins as possible. The 10ยตf on each board is helpful as well.
    Note the exact value is not critical.

2a) Resistor value must be determined by how much current needs to be diverted from the LED's. I would start with 500k then 100k.... etc.
Keep in mind the power in the resistor is:

  • P = E^2/R
  • P = 200 * 200 /R
  • P = 40,000/R

FOr reference, 100k ohm resistor will dissipate 0.4 watts you must purchase at least a 2 W or better yet a 5 W resistor.

You might try something greater than 500k and maybe get the LED brightness down to an acceptable lever without it going completely off.

Thank you @Paul_B ,
I'm afraid that the third is not a viable option. The led strips have to be on or off, not weakly light...

Thank you @JohnRob ,
I'll make some tests with resistor (and decoupling capacitor, of course).

This PDF is, I think, very informative.
Of particular interest, please refer to Snubber followed by Off-state leakage current on 'page 7 of 16' ("2-16"):
SSRs always have off-state leakage currents. The thyristors, control circuitry and
snubber network all supply small off-state currents, which usually total from about 1
to 10 mA rms.
These leakage currents should be taken into account when either indicators are used,
or the circuit may actually be touched, say for servicing. A resistor across the indicator and a line safety breaker are the standard means by which these limitations can be overcome.
ssr.pdf (954.9 KB)

As I read your response another idea hit me, then I read @runaway_pancake's post which supported this idea.

You should try a capacitor across the 200V (is it 200 or 220?)

I would start with a 0.001 to 0.005 ยตF 600Volt AC rated capacitor. These are used mostly in motors. The beauty of a capacitor is they don't consume energy nor create heat. The capacitor draws current during part of the phase then returns it the next (kinda). I think it is definitely worth a try.

@JohnRob : we say that it's 220V (energy supplier indicate 220/240 V, but mostly it'd between 220-230 V.
@runaway_pancake : "across the indicator and a line safety breaker" sounds foggy to me :slight_smile: