Announcement

Collapse
No announcement yet.

OT relay coil snubber

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • OT relay coil snubber

    Ok the classic way to prevent voltage spikes when turning off a relay coil is a parallel diode. I'm using a relay to switch mic level signals, and I want no induced signal at all from the relay coil. Rather than the diode, I'd put a capacitor across the coil terminals, then feed the voltage to it through a resistor. I realize this slows the action of the relay, but that doesn't matter- what does matter is that no transient signal is developed that could transfer to the mic input. The diode clamps the voltage spike, but it does so by shorting out a transient spike, which creates a current spike. I want to eliminate the spike entirely by allowing the relay a soft close and a soft open.

    These are 24 volt relays with 600 ohm coils, so 40 ma operating current. The switch controlling the relay will be in the same box as the relay and the input and output jacks. I'm thinking of using a 100 ohm resistor in series with the switch, and something like 10 uf cap across the relay coil.

    Have I just lost more neurons, or does this sound like it will work?
    I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-

  • #2
    Won't happen just like that.

    You have energy stored in the coil, and when you open the switch, the stored energy has to go SOMEWHERE. So there is a transient.

    What you can do is minimize it.

    One way is to use a diode and add a resistor in series. That will get rid of the energy, but may be not much different from using only the coil resistance for a 600 ohm coil.

    If you are concerned about transient snap voltages, then put a capacitor either across the coil, or from coil to ground, with a resistor in series with the transistor driving the coil, to avoid a short circuit through the capacitor. Value should be enough to avoid damaging the transistor, but not enough to interfere with closing the relay.

    If you put the resistor in series with the capacitor, it will change the capacitor's action and permit some "snap" voltage change to occur.
    Last edited by J Tiers; 01-31-2021, 09:17 PM.
    2801 3147 6749 8779 4900 4900 4900

    Keep eye on ball.
    Hashim Khan


    It's just a box of rain, I don't know who put it there.

    Comment


    • #3
      Is there a reason you are using a magnetic relay rather than a semiconductor to do your switching?
      "A machinist's (WHAP!) best friend (WHAP! WHAP!) is his hammer. (WHAP!)" - Fred Tanner, foreman, Lunenburg Foundry and Engineering machine shop, circa 1979

      Comment


      • #4
        With Mic level signals it is going to be extremely hard to eliminate transient spikes while using a magnetic relay. The better way to do this would be to use either a semiconductor switch of which there are many available or if it works with your design idea, use a semiconductor attenuator. This would give you more options including variable level control. This is how professional mixer boards work today.
        Robin

        Happily working on my second million Gave up on the first

        Comment


        • #5
          Current day mixer boards are all digital, and have digital routing and so forth, up until (and in some cases even after) the signal leaves the mixer.

          It appears that this is not applicable to the OP's situation. With mic level signals, semiconductor switches are not too great. They really need line level, or at least something post preamp, to work decently.

          Ther IS a way to do this. Use an OPAMP to operate the coil, with transistor booster if needed. Make the opamp an integrator, with a capacitor as feedback. Size the input resistor so as to get a sufficiently slow transition.

          That will give a slow (you set the slew rate) increase of voltage, or decrease of voltage, and should work fine. You may need to put a diode in series with the coil if the opamp supply is bipolar.
          2801 3147 6749 8779 4900 4900 4900

          Keep eye on ball.
          Hashim Khan


          It's just a box of rain, I don't know who put it there.

          Comment


          • #6
            I wonder if putting a MOV across the coil will help.

            Comment


            • #7
              Yes, this is the much preferred way. You do not need to know a lot about circuit design, just find an appropriate IC and look at the application notes. In the past I have seen 4066s used for this, but I am sure there must be much better IC by now. Digi-Key and Mouser have search facilities that will make finding one easier.



              Originally posted by rdfeil View Post
              With Mic level signals it is going to be extremely hard to eliminate transient spikes while using a magnetic relay. The better way to do this would be to use either a semiconductor switch of which there are many available or if it works with your design idea, use a semiconductor attenuator. This would give you more options including variable level control. This is how professional mixer boards work today.
              Paul A.
              SE Texas

              And if you look REAL close at an analog signal,
              You will find that it has discrete steps.

              Comment


              • #8
                So just put your preamp before the switch. Duhhhhh!

                There's your use for your op-amp.



                Originally posted by J Tiers View Post
                Current day mixer boards are all digital, and have digital routing and so forth, up until (and in some cases even after) the signal leaves the mixer.

                It appears that this is not applicable to the OP's situation. With mic level signals, semiconductor switches are not too great. They really need line level, or at least something post preamp, to work decently.

                Ther IS a way to do this. Use an OPAMP to operate the coil, with transistor booster if needed. Make the opamp an integrator, with a capacitor as feedback. Size the input resistor so as to get a sufficiently slow transition.

                That will give a slow (you set the slew rate) increase of voltage, or decrease of voltage, and should work fine. You may need to put a diode in series with the coil if the opamp supply is bipolar.
                Last edited by Paul Alciatore; 01-31-2021, 11:13 PM.
                Paul A.
                SE Texas

                And if you look REAL close at an analog signal,
                You will find that it has discrete steps.

                Comment


                • #9
                  With mic level signals, the ICs do not work as well. They have some switching noise, and at that level, the small switchig transient can be pretty huge coming out of the "other end". There can be signal to noise issues of other types as well.

                  Newer ICs may be better, I have not looked at them for a few years.

                  The OP has the mic level switching, and wants it to be quiet. He can do that with the integrator switch driver that I mentioned earlier. Pretty much ZERO induced noise and no spike.

                  There will be a need perhaps, for ensuring that there is no voltage difference that can cause a noise when switching. It does not take much noise to come through as a loud thump. Both sides of the relay or other switch element must be at the same potential as far as DC, or there will be an audible thump, snap, or worse. Usually it is easiest to keep them at DC ground.

                  The ultimate in quiet switching is with an opto system.... optovariable resistors (Vactrol or similar). I have done that, and it can be dead quiet.

                  There are many ways to do this job without requiring a complete change of the system.
                  2801 3147 6749 8779 4900 4900 4900

                  Keep eye on ball.
                  Hashim Khan


                  It's just a box of rain, I don't know who put it there.

                  Comment


                  • #10
                    Sorry, lil late was off doing something else.

                    Relay snubber? Very nice. Have not heard that language for a second.

                    I thought resistors were the common method way back.

                    Yeah, yer smart. It will be good. JR

                    My old yahoo group. Bridgeport Mill Group

                    https://groups.yahoo.com/neo/groups/...port_mill/info

                    Comment


                    • #11
                      Have you tried the circuit with just the diode?

                      I've built many custom AV systems with relay switching of both mic- and line-level signals, using diodes as snubbers. I've never heard any switching noise on the mic lines.

                      If you use DPDT relays, you can also wire them so that when the mic is turned "OFF", you short out the line going to the destination. That will make sure that if anything does get through, it is only for a very brief instant.

                      Comment


                      • #12
                        The better switching circuits that I have seen for turning audio and video on and off have employed dual switches: one in series to block or pass the signal and the second one from the output to ground to short the output when the signal is turned off. They are very effective. A logic chip (inverter) is used to generate the opposite polarity control signals needed for these two electronic switches. By quickly putting a ground on that output line when it is being switched off, you will eliminate a lot of noise.

                        Another thing to keep in mind is the impedance of the circuits. I did a long discussion about the effects of circuit impedance either here or on Practical Machinist a week or two ago but what it amounted to is that you should use low impedance circuitry if you want to minimize noise, like relay spikes. In this case, using a low impedance microphone is a very good idea: 600 Ohms, 150 Ohms, 50 Ohms or even less. This can go a long way towards having a low noise circuit. Likewise, the circuit that follows the switch should also exhibit a low impedance input. Many common, consumer style audio devices will have a relatively high impedance, like 10,000 Ohms. That may allow the use of cheaper, high impedance microphones, but it is not going to help with noise in the circuit.



                        Originally posted by andywander View Post
                        Have you tried the circuit with just the diode?

                        I've built many custom AV systems with relay switching of both mic- and line-level signals, using diodes as snubbers. I've never heard any switching noise on the mic lines.

                        If you use DPDT relays, you can also wire them so that when the mic is turned "OFF", you short out the line going to the destination. That will make sure that if anything does get through, it is only for a very brief instant.
                        Paul A.
                        SE Texas

                        And if you look REAL close at an analog signal,
                        You will find that it has discrete steps.

                        Comment


                        • #13
                          Originally posted by JRouche View Post
                          ... I thought resistors were the common method way back...
                          Certainly was standard practice in the electrical action of pipe organs - may still be - dating from the days when diodes were big, bulky and expensive. It also meant that the organ builder, who was usually more at home with wood, leather and animal glue, didn't have to know which way round to connect it. But that's a very agricultural application compared with the OP's requirements.

                          George B.

                          Comment


                          • #14
                            The resistor does work. Any method that extracts the energy (current) from the inductance of the coil will do that.

                            The resistor, if of optimum value, will approximately double the transient voltage at the coil terminals vs the diode alone, when the relay drive is turned off. And it draws current when the coil is energized. The less current it draws when the coil is energized (higher resistance) the higher the transient voltage, and potential noise input to the circuit.

                            Adding a diode to the resistor will eliminate the extra current when energized, but does not affect the transient voltage significantly. You would have better results from simply using the diode alone, which has half the transient voltage.

                            The only "snubbing" method which actually reduces the noise input is to use a capacitor snubber, similar to those used in a flyback type SMPS. That generally involves a diode, capacitor, and a resistor to discharge the capacitor after the transient is over. it will lower the noise on turn-off of the drive, but does not affect the turn-on, which needs a separate slow-down circuit.

                            The "integrator" method of relay drive can do a true "slow transition" at any speed desired, and will have essentially zero noise induction into the circuit from the relay drive. It can be done with either a transistor drive, or a more complicated IC drive.
                            Last edited by J Tiers; 02-02-2021, 09:53 AM.
                            2801 3147 6749 8779 4900 4900 4900

                            Keep eye on ball.
                            Hashim Khan


                            It's just a box of rain, I don't know who put it there.

                            Comment


                            • #15
                              Or you could get something like an Omron mosfet relay. It's like an optocoupler but then with a conducting fet as output. No coil, no noise....

                              Comment

                              Working...
                              X