Capacitor for across-the-mains usage

[pmbrunelle]

I am in Canada here, so mains power is 120 VAC, 60 Hz.

So I have a small AC synchronous motor (lets say pill-bottle-sized) in my device.

I have added an RC snubber in parallel with the motor to deal with the inductive kick upon opening the solid-state-relay.

The RC consists of:
R = 120 Ohm, 1 W, taken from scrap bin
C = 1.5 uF, 250 VDC, taken from scrap bin, looks kind of like some generic red gumdrop, 15 mm lead spacing: http://i00.i.aliimg.com/photo/v0/629393316/100v_Metallized_Polyester_Film_capacitor_155K.jpg_200x200.jpg

At line frequency, the capacitor's impedance is much greater than that of the resistor, so the capacitor sees most of the mains voltage across it.

Will this be okay? Is the ripple current typically an issue? Should I get an X2-rated capacitor for this?

Do X and Y ratings actually mean something in practice, or is it just regulatory red tape?

I estimate the duty of this gadget to be 200 hours per year, for 8 years.

[rs20]

Class X1 or X2 probably make sense. Class X is used when a short circuit in the cap might cause a fire, which sounds like your case. Typical use is line-to-neutral, like what you're doing.

Class Y is unnecessary in this case. Class Y is used when a short circuit in the cap might cause electric shock to the user, which isn't true in your case. Typical use is line-to-earth.

[station240]

X1 and X2 capacitors are designed for connection across the mains, including safety aspects.
When an X class capacitor fails, it fails open circuit. Also designed not to burn up or catch fire.

Naturally cheap Chinese junk uses 400V generic green cap capacitors, or if you're lucky 1-3kV blue disc caps.

[Aodhan145]

Why don't you use a cheap MOV?

[pmbrunelle]

So if this cap fails shorted, then the resistor will see the full mains voltage and overheat (by a lot).

The circuit is fused 2 A, so the fuse won't blow if the cap fails shorted, due to current limiting from the 120 Ohm resistor. So short-circuit failure is not too good.

I don't like if the cap fails open either... then the snubber will lose effectiveness. EMI will go everywhere into the digital circuitry, which will silently corrupt the data acquisition. This will cause frustration until the cause is determined...

As for a MOV, the device has been tested for a few months now with this RC combination, so I have confidence in this snubber. I don't want to change the circuit too much now, besides changing a component rating (but same value) for better long-term reliability.

I know I should be using an X2 capacitor for this, I just don't want to wait an extra week to get this thing out the door...

[rs20]

To be clear, X1 is better (has a higher breakdown voltage) than X2.

It's not like the cap is actually going to fail, neither short nor open. It's just that if it does fail, it'll fail open. The main reason it might fail is that it is exposed to a surge; this is where having a MOV can help. Adding a MOV in parallel with your cap will have no effect on its normal operation; it'll just protect the cap.

[Gyro]

So if this cap fails shorted, then the resistor will see the full mains voltage and overheat (by a lot).

The circuit is fused 2 A, so the fuse won't blow if the cap fails shorted, due to current limiting from the 120 Ohm resistor. So short-circuit failure is not too good.

----

I know I should be using an X2 capacitor for this, I just don't want to wait an extra week to get this thing out the door...

I don't like the sound of the "out the door" reference. If it's going out of yours and into anyone else's then you really do need to use an X cap as, with the fuse rating, you have identified a potentially dangerous failure mode.

You might also get away with using an approved flameproof fusible resistor, but the X cap sounds easier.

EDIT: Actually the fusible resistor might be preferable in this case because X caps need current in order to self-heal and the resistor might restrict that.

[rs20]

Actually the fusible resistor might be preferable in this case because X caps need current in order to self-heal and the resistor might restrict that.

We don't care whether the cap self-heals, we care whether it catches on fire. An X class cap will thus not catch fire no matter how restricted he current is, so there's no reason not to use one.

[Ian.M]

Actually the fusible resistor might be preferable in this case because X caps need current in order to self-heal and the resistor might restrict that.

We don't care whether the cap self-heals, we care whether it catches on fire. An X class cap will thus not catch fire no matter how restricted he current is, so there's no reason not to use one.

X class cap and fusible resistor so the resistor doesn't cause a fire either.

[Gyro]

My point was that if the X cap doesn't get enough current to self heal (ie, vaporize the metalization around the puncture) then it will remain short circuit - frying the resistor.

Edit: That's how X caps work, it's not that they never break down, they just self heal without catching fire. (unless they're a Rifa MP that is  )

[Seekonk]

1.5uF is way overkill, .1 to.47 would be sufficient.  Must have one of those X2 in your scrap box.

[pmbrunelle]

I ordered an X2 capacitor and a UL-1412 (apparently that's the magic word) fusible resistor... so I expect to replace the existing parts with these ones.

As

1.5uF is way overkill, .1 to.47 would be sufficient.  Must have one of those X2 in your scrap box.

Actually, since I work in automotive, everything we have is for 12 V electrical systems. We don't have scrap boxes with X or Y capacitors.

As for 1.5 uF, here's the logic in how I came up with it:
I used an LCR meter to measure the electric motor's resistance and inductance.
At the moment of relay opening, a series RLC circuit (including the snubber) is formed.
I picked the capacitor so that the current wouldn't oscillate back and forth - it was close to critical damping.

So from this analytical sort of logic, I picked 1.5 uF as a sort of nominal value. I then tested 1 uF and 2.2 uF to ensure some robustness, then I settled on 1.5 uF.