Solid State Relay question

[ricko_uk]

Hi,
I haven't used a SSR before and am looking at this one: http://www.crydom.com/en/products/catalog/pf-series-ac-pcb-mount.pdf

Three questions:

1) on page 1 it says: "Zero-crossing (resistive loads) or random-fire (inductive loads) output". Should it not be the opposite? i.e. Zero-crossing (inductive loads) or random-fire (resistive loads) output?

2) is it suitable for capacitive loads?

3) how does it work? Does it switch on the output at zero crossing whenever the LED is on?

Many thanks

[Doctorandus_P]

You would want to switch the SSR when the current is (near) zero.
For a resistive load, the current and voltage are in phase, so detecting the zero crossings of the voltage is just fine.

For an inductive load you also want to switch at the point where the current is (near) zero, but there is a phase shift between voltage and current.

For a 600VA toroidal transformer project I once did just for fun, I had extremely large current surges If I switched it at the wrong moment. After I figured out which phase delay to trigger the solid state relay, I still had very high surges about half the time I turned the transformer on.

It turned out that the remanent magnetism in the turned off transformer was important enough to cause an extra phase delay. When I figured this out, I always turned the transformer off during a negative going flank, and turned it on at a positive going flank, and then the current surges stopped.

During development I've measured the full mains peak voltage (>300V) over a 1 ohm shunt resistor, which I found quite scary. The transformer also hummed loudly for a few seconds after such an event.

[ricko_uk]

Thank you Doctorandus_P

[TimNJ]

General rules of thumb for any type of switch (mechanical switches, relays, transistors):
1.) Inductive loads: High stress at turn off
2.) Capacitive loads: High stress at turn on

For capacitive loads, the stress is caused by inrush current. SSRs usually have a 1 cycle max current rating. You can determine the peak inrush current through calculation, measurement, or simulation. For a zero-crossing turn-on triac/SSR,I’d say there shouldn’t be much issue, unless you are really switching a very large capacitance.

[ricko_uk]

Thank you Tim,

You can determine the peak inrush current through calculation, measurement, or simulation.

How can I determine it through calculation, at least a rough/safe value? Would it be just through ohms law using the capacitor ESR as the only resistance? Would that be the peak inrush current value?

For a zero-crossing turn-on triac/SSR,I’d say there shouldn’t be much issue, unless you are really switching a very large capacitance.

What would be a "very large capacitance"? I have 1,500uF up to 3,000uF (i.e. 1.5 to 3 Farads)

Many thanks

[OM222O]

What would be a "very large capacitance"? I have 1,500uF up to 3,000uF (i.e. 1.5 to 3 Farads)

No! That would be very "small" capacitance! I have fitted multiple 2500uF caps on a board without inrush issues. 3000uF is only 3mF not 3F.

Super caps (>1F) don't come at mains voltages anyways!

[ricko_uk]

ahah, sorry, my mistake in the rush of replying.

Thank you for the infos and feedback!

[TimNJ]

The absolute worst case is when the switch is turned on at 90 degrees (peak of the line voltage). The peak current in this case is simply the peak AC voltage divided by the ESR of the cap. (Just as you said.) However, an SSR turns on at the voltage zero crossing so the peak current is a little harder to calculate. I usually just plug in the equivalent circuit into LTSPICE. The peak will not occur immediately at switch on but is usually delayed until 45-60 degrees (I think, don’t quote me on that. I don’t know an easy way to do that by hand.)

You can judge what capacitance is too big by adjusting the value until the inrush current starts to approach the 1 cycle peak rating. Since different SSRs have different ratings, it’s best to judge that on a case by case basis.

And really, it’s the ESR driving the worst case peak current, so check the respective capacitor’s datasheet.

[ricko_uk]

Thank you Tim

[TimNJ]

You're welcome. In general, if you only use the ESR of the capacitor..you get the absolute worst case if the source impedance/resistance of the power system was 0 ohms. In reality, of course that's not the case, and if you're using a large cap with low ESR (<1 ohm), you can get a pretty big error in your simulation if you don't account for source resistance. Small cap with high ESR, the ESR of the cap dominates, so error is smaller.

I often model a typical building source resistance with about 0.5ohm. So, if you used a cap with an ESR of 1 ohm, and neglected to include ~0.5 ohm source resistance, your simulation will probably show a peak current about 1.5x higher than what you'd probably actually measure.

Still, designing for 0 ohm source resistance will give you a more conservative design, and that's not necessarily a bad thing if you can afford it.

[Zero999]

I have 1,500uF up to 3,000uF (i.e. 1.5 to 3 Farads)

You've got the decimal point wrong. 1500µF and 3000µF is 1.5mF and 3mF, not Farads.