Solid State Relay latches due to low current problem

After some hours of troubleshooting, I finally discovered that my Arduino is in fact working, --turning on and off a pin, which controls a solid state relay. But the problem is my load isn't enough, and so the really never shuts offs, or shuts off randomly, like a minute later. I tested my load (an aquarium bubbler) and it draws 45ma of AC current. If I use a lightbulb or something, my SSR turns on and off perfectly, so I know this is the problem. I'm just not sure the best way to fix it. If it were a DC circuit, I could add a resistor or something. but it's 110VAC.

included is a schematic of my SSR. I actually have 4 of them on my PCB, and the other 3 are fine, because the load is more substantial.

ps. Here's the data sheet of the SSR, and it does have something on there about a latching current, but it's not really supposed to BE a latching relay. So I'm assuming if the current is below this level, (which mine is right at the level), that this SSR just happens to not turn off...

SSR datasheet

Try a snubber circuit across the bubbler/pump.
Something like 100ohm/1uF.
Leo..

I suggest you read the relay data sheet and see what its minimum current is. If I remember correctly they bias the electronics and that bias current appears across the load contacts.

The holding current is the current required to keep the Triac on once the gate voltage has been removed. So if you had problems with that then you would see your device fail to switch on or at least only have power on it for the duration of the gate pulse.

The latching current has nothing to do with a latching relay. See this

The problem you have is with the phase angle between the current and the voltage not allowing the commutation ( turning off ) of the Triac. To turn a Triac off the voltage across it must be below the latching voltage and at the same time the current through it must be below the latching current. An inductive load puts the voltage and current out of phase so these conditions can not exist at the same time.

Wawa suggestions would basically correct that phase angle an make it work. Another way would be to place a resistivity load in parallel and let that correct the phase angle. Something like a low wattage incandescent light or a heater, like a small aquarium heater.

If it were a DC circuit, I could add a resistor or something. but it's 110VAC.

You can still add a resistor to an AC circuit, you just have to ensure that the wattage is high enough to dissipate the heat it will generate. That is in effect what the other loads I suggest are doing.

I tested my load (an aquarium bubbler) and it draws 45ma of AC current. If I use a lightbulb or something, my SSR turns on and off perfectly, so I know this is the problem. I'm just not sure the best way to fix it. If it were a DC circuit, I could add a resistor or something. but it's 110VAC.

Your minimum SSR switching current is 100 mA and you mention the load is 45 mA. You are switching 110 ~ 120 VAC. Your problem is leakage current is my guess. I would just place a resistor across the load. This will increase your load current. I would try about a 1200 Ohm resistor. The problem becomes Figure 120 V @ 100 mA = 12 Watts so you need a pretty large resistor to dissipate 12 watts when the pump is running. Your SSR is rated up to I believe 3 amps so you can lower the resistance. The idea being you want a load that exceeds the 45 mA you have. I agree with Grumpy Mike in that increase the load be it a lamp or resistor or anything else.

Ron

I'm not an electronics expert, but could you not use a Triac rather than an SSR in this situation?

countrypaul:
but could you not use a Triac rather than an SSR in this situation?

And this makes a difference because?

Grumpy_Mike:
And this makes a difference because?

Grumpy_Mike:
And this makes a difference because?

If I knew the definitive answer I wouldn't have asked the question. I read somewhere that a Triac could be used when trying to control low current mains voltage. Like everythig on the Internet - whether it is true or not needs someone with a full understanding to explain.

It is true. But a Triac is just like two SCRs back to back. In fact that is what the OPs data sheet showed. In effect there is very little difference between the two.

What do you mean by SCR? I assume that its not a typo for SSR as that would not apper to make any sense.

Adding a panel mount indicator wouldn’t be a bad idea for my project. Would this panel mount lamp be sufficient? For 120VAC indicators, it’s about the highest current rating I found, but it’s only 25mA. But current isn’t the problem, it’s the phase/angle thing which I just barely grasp an understanding of. So I better ask if this lamp would be sufficient. Thanks.

https://www.digikey.com/product-detail/en/visual-communications-company-vcc/1092C3-125VAC/1092C3-125VAC-ND/6175228

Seems they got that wrong. 25mA is more than 3watt.
That will melt the plastic holder in seconds.

Just try that snubber circuit.
Leo..

countrypaul:
What do you mean by SCR? I assume that its not a typo for SSR as that would not appear to make any sense.

Wikipedia!

Fourth hit on my Duck2Go search, :roll_eyes:

countrypaul:
What do you mean by SCR? I assume that its not a typo for SSR as that would not apper to make any sense.

A SSR is a Solid State Relay. It contains in it two SCRs - Silicon Controller Rectifiers wired back to back and together they act as a Triac a portmanteau word consisting of transistor and AC.

I think you are out of your depth here, learn a bit before wading in about things you know nothing about.

But current isn't the problem, it's the phase/angle thing which I just barely grasp an understanding of. So I better ask if this lamp would be sufficient.

You can’t tell if any resistive load would be enough because you don’t know exactly what phase angel you have got that you are trying to counteract. The correct word for this phase difference is called the power factor.

You need to know by how much you need to bring this power factor down by. You then take the vector addition of the power factor of the two loads to find the final power factor of combining two loads.

Fluorescent lights have a capacitor in them to bring up the power factor onto the real axis because household meters can only registers power projected onto the real axis. If you remove the capacitor then you get two things:-

  1. you run the light for free.
  2. you break the law because it is an offence to do this.

Grumpy_Mike:
Fluorescent lights have a capacitor in them to bring up the power factor onto the real axis because household meters can only registers power projected onto the real axis.

Wow! :astonished: That is new!

Whatever happened to the good old fashioned Ferraris/ Bláthy meter which automatically compensated for the power factor? I thought that was the standard - at least in my lifetime here.

(Apparently things must be different in England. :roll_eyes: )

Yes the data sheet says 2 back to back SCR's. We did that long before triacs were readily available at a reasonable cost at the ratings needed. The data sheet tells the Holding Current - IH - 44 50 mA, that states you must have at least a 50mA load to keep it on. You can put a LED and probably pick up the current you need. For many years and still today SCR's are used in many high horsepower DC drives. Try to find a triac rated at about 300A at 1000V, for use on a 3 Phase 550V mains to drive a DC motor.

The data sheet tells the Holding Current - IH - 44 50 mA, that states you must have at least a 50mA load to keep it on.

Yes but that is not the problem the OP is having. His problem is that it won’t go off.

Paul__B:
Wow! :astonished: That is new!

Whatever happened to the good old fashioned Ferraris/ Bláthy meter which automatically compensated for the power factor? I thought that was the standard - at least in my lifetime here.

(Apparently things must be different in England. :roll_eyes: )

Obsolete because they can’t report home. Modern UK meters are all smart and take & report readings
automatically. They’ve been rolling out for over a decade IIRC.

The old mechanical meters were subject to a lot of fraud as sticking a big magnet on them slows them
down. Mechanical meters aren’t very flexible for multiple-tariffs (although this has been done, ie
Economy 7 meters had two meter clocks driven from the meter movement on a time-switch),
Mechanical meters are useless for future requirements such as supporting load-shifting where the
cost varies across the day, and have all gone now I think (certainly they were due to be replaced by 2020)

Energy use monitors typically report the consumption live from the smart meter too.

Billing for apparent power rather than real power is a great incentive to have a sensible power factor
in the first place (bad power factors mean the power distribution infrastructure has to be rated more
conservatively, as higher currents are flowing - and more power is lost in wires). Highly reactive
loads cause instability problems in the distribution network too, so its important domestic electricity
use is kept relatively well-behaved (in the old days the bulk of the load was resistive heaters, not
any more…).

And smart meters measure real power and apparent and can support complex tariffs involving both
values (in theory at least - no idea if this is done).

Don’t get me wrong, the old style electro-mechanical meter was a miracle of technology, its just that
it was essentially Victorian(*) technology and doesn’t meet today’s requirements.

(*) slight exageration, but not far from the truth.

Grumpy_Mike:
Fluorescent lights have a capacitor in them to bring up the power factor onto the real axis because household meters can only registers power projected onto the real axis. If you remove the capacitor then you get two things:-

  1. you run the light for free.
  2. you break the law because it is an offence to do this.

Interesting, power factor capacitors here in Australia were only needed in industrial installations.
(Large numbers of factory lighting units)
Possibly to also make sure 3phase current balance.
Ahh those were the days, always a flickering "fluoro" somewhere in the building. :slight_smile: :slight_smile: :slight_smile:
Household fluoro batten units here only had to have the inductor to start them.
Tom... :slight_smile: