relay Snubber Questions

[R.G.]

could you use an RC series circuit in parallel with the coil instead of just a capacitor?

Yes; however, this has some quirks.

A simple C in parallel with the coil makes for a damped ringing resonant circuit. The coil has a moderate inductance. Some relays even specify the inductance range in their datasheet; it changes depending on relay open or closed. The coil has a non-trivial resistance. Putting a capacitor across the resistor/coil forms a damped RLC circuit. When you put a transient into it, it rings at the natural frequency of the RLC. Putting more resistance in series with the C changes the damping by making the C less effective. The voltage on the coil is less constrained by the capacitance. This ... may ... be a good idea or not, depending on the amount of total damping it introduces. Damping eats energy out of the stored current in the coil as it oscillates back and forth into the capacitor. Prolonged oscillation is not a good thing.

Every RLC circuit reacts to a transient step by producing an initial smaller step plus ringing. You juggle how much of a step versus how much ringing by setting the damping.

this would slow things down quite a bit no?

No. Imagine that the resistance in series with the cap can be between 0.001 ohm and 1M ohm. At 1M, the capacitor is essentially not even there, and exerts little effect on the inductor. At 0.001 ohm, the resistor is effectively not there. In the middle, the resistor damps things some. But the resistance of the inductor already damps things. The exact value needed for least audibility and fastest relay response depends heavily on the values of R, L and C.

However, the ring frequency of the LC is always determined by the value of L and C. R's have no frequency selectivity.

if you used the right values maybe you could even make a tremolo

?? Not sure how that would work  

[R.G.]

I messed with the circuit some more. I made the circuit in the schematic in the "A remote indicating Bypass" article. I measure with a meter that indeed the pop is supressed quite a bit. I put a 220uf cap across the coil.
I don't have a 47uf available to test. My relays are hi sensitivity. Are there any values i should experiment with resister wise?

Without knowing the exact values of the coil inductance and resistance, can't say. What you might try is putting a resistor in parallel with the coil. This eats DC current when you have the relay energized, but also damps the coil a lot when you turn it off. This suppresses the coil's swing, too.

Also, in your circuit RG, does establishing a ground reference and using ground as my common going to be an issue at all?

Strictly speaking, there is no need for the relay power to have a common ground with the audio at all. However, you could inject power line hum if you powered the relays from a completely isolated power supply run from the AC mains. What you might do is isolate audio ground and power ground for the relay coils, and connect them through a resistor, then try different resistances. Something between 10 ohms and 1K should let you experiment and get best hum reduction and most isolation.

[trixdropd]

I messed with the circuit some more. I made the circuit in the schematic in the "A remote indicating Bypass" article. I measure with a meter that indeed the pop is supressed quite a bit. I put a 220uf cap across the coil.
I don't have a 47uf available to test. My relays are hi sensitivity. Are there any values i should experiment with resister wise?

Without knowing the exact values of the coil inductance and resistance, can't say. What you might try is putting a resistor in parallel with the coil. This eats DC current when you have the relay energized, but also damps the coil a lot when you turn it off. This suppresses the coil's swing, too.

Also, in your circuit RG, does establishing a ground reference and using ground as my common going to be an issue at all?

Strictly speaking, there is no need for the relay power to have a common ground with the audio at all. However, you could inject power line hum if you powered the relays from a completely isolated power supply run from the AC mains. What you might do is isolate audio ground and power ground for the relay coils, and connect them through a resistor, then try different resistances. Something between 10 ohms and 1K should let you experiment and get best hum reduction and most isolation.

Cool RG, Thank you for the help. I will experiment some more and report back.

[slotbot]

?? Not sure how that would work  Huh

well i was just making a joke. but really i think you could if you picked nice values. like if you go here

http://www.falstad.com/circuit/

and import

$ 1 5.0E-6 10.20027730826997 46 5.0 43
v 544 112 544 48 0 0 40.0 5.0 0.0 0.0 0.5
w 464 368 544 368 0
w 544 32 544 48 0
r 208 208 208 256 0 10.0
c 208 256 208 368 0 1.4999999999999999E-5 -3.841805128025876E-4
w 208 208 208 176 0
w 208 176 272 176 0
w 208 368 272 368 0
178 144 224 64 224 0 1 1.0 2.4498178926322914E-6 0.05 1000000.0 0.0010 0.1
w 144 256 176 256 0
w 176 256 176 176 0
w 176 176 208 176 0
w 144 272 144 368 0
w 144 368 208 368 0
t 336 128 272 128 0 1 -5.0 4.086786860476207E-4 100.0
w 272 144 272 176 0
r 336 128 336 32 0 100.0
w 544 32 336 32 0
w 272 112 272 32 0
w 272 32 336 32 0
s 336 176 336 272 0 0 false
w 336 128 336 176 0
w 544 272 336 272 0
w 544 368 544 272 0
w 544 112 544 272 0
w 400 368 464 368 0
r 272 368 400 368 0 100.0
x 61 415 602 421 0 24 turn it back on and watch the relay flip on and off.
x 45 458 440 464 0 24  maybe tremolo was the wrong word.
x 367 181 616 185 0 15 turn this switch off and relay goes on...
x 456 451 673 454 0 10 but i was imagining the relay going on and off
x 51 53 103 59 0 24 hello

i was just making a joke though... because you have oscillations (like you mentioned)... maybe to turn the signal on and off... like a tremolo

[amptramp]

Just looking at the layout in the first post, it appears the relay has a diode across the coil.  The diode is oriented so that it does not conduct unless the coil voltage at the collector of the transistor goes above the supply voltage, so there is no conduction when the relay is on or off in the steady state.  When the relay switches off, the coil voltage will go above the supply (sort of like an ignition coil) with an uncontrolled rise until the diode conducts.  This delays the opening of the relay because the current in the coil is still flowing in the same direction.  I have seen series diodes used for redundancy or in series with resistors to permit the relay to open faster by permitting more voltage drop in the circuit around the snubber.

R-C snubbing draws power during the turn-on whereas a diode does not, so a diode will eliminate the need to size the transistor for the coil current and the capacitor current.  A diode will also tend to eliminate ringing, which may be unacceptable in an audio application.

Good luck with the build!