Back to Basics video: Why diodes are used around relay coils

[w2aew]

This is a pretty common question that I get asked, so I figured it was a good time to put together a tutorial on why flyback / snubber diodes are used around relay coils, solenoids, electric motors, etc.

[janengelbrecht]

An inductor coil will send its energy into the driver transistor when it is turned off....or try to One could also use an arrangement with a PTC resistor though...

[Andy Watson]

Nice vid. I reckon the "ringing" that you refer to around 11:30 is mechanical due to the armature of the relay being released.

[w2aew]

Nice vid. I reckon the "ringing" that you refer to around 11:30 is mechanical due to the armature of the relay being released.

Ooh - good point!  That's probably it!

[w2aew]

An inductor coil will send its energy into the driver transistor when it is turned off....or try to One could also use an arrangement with a PTC resistor though...

I'm not sure a PTC resistor would react fast enough due to the thermal lag...

[jlmoon]

An inductor coil will send its energy into the driver transistor when it is turned off....or try to One could also use an arrangement with a PTC resistor though...

I'm not sure a PTC resistor would react fast enough due to the thermal lag...

Are PTC devices the  same as Posistors?  I those are used to soften or buffer the input startup currents for any temporarily low resistance or impedance load? .. such as really large low ESR discharged capacitor on the output of a bridge diode array.  They are great for the inflow but not good after the fact.. if I understand correctly.

[SeanB]

The ring also is indicating when the armature is being released from the frame, and thus is the delay of the relay opening. Shws how slow a relay is at opening, and the time to the hump is the closing time.

Instead of a PTC device use a MOV or transient suppressor ( or a zener diode) to clamp the voltage to something under what the transistor will withstand. Faster turn off and still protecting the transistor. You just place it across the transistor and leave the other diode  out.

[w2aew]

The ring also is indicating when the armature is being released from the frame, and thus is the delay of the relay opening. Shws how slow a relay is at opening, and the time to the hump is the closing time.

Thanks Sean - I hadn't thought of that until Andy brought it up in an earlier post.  Makes perfect sense.  I've added an annotation in the video to indicate that this is where the hump is coming from.  I should have probed the relay contacts, and that would have cinched it.

[timb]

The ring also is indicating when the armature is being released from the frame, and thus is the delay of the relay opening. Shws how slow a relay is at opening, and the time to the hump is the closing time.

Instead of a PTC device use a MOV or transient suppressor ( or a zener diode) to clamp the voltage to something under what the transistor will withstand. Faster turn off and still protecting the transistor. You just place it across the transistor and leave the other diode  out.

Interesting. So if I have a 60V NPN, I could put a, say, 50V Zener between the coil and collector and the relay would turn off faster and the transistor would still be protected?

The Zener would still only need to be rated for the coil current like a traditional snubber, right?


Sent from my Smartphone

[Andy Watson]

Interesting. So if I have a 60V NPN, I could put a, say, 50V Zener between the coil and collector and the relay would turn off faster and the transistor would still be protected?

The Zener would still only need to be rated for the coil current like a traditional snubber, right?

Yes. But note that the voltage swings above the positive rail (in this example) so the supply would have to be limited to 10V. Also, remember that zeners conduct like a normal diode in the forward direction - so you would need another normal diode in series with the zener.

[gxti]

Alternately, use just the zener but connect the anode to ground. Then you not only don't have to worry about the supply voltage (because the diode is directly across the voltage-sensitive part i.e. the transistor), but it also means the coil current isn't being dumped into your positive supply where it could end up affecting other circuit components. If it dumps into VCC then depending on how much energy was stored in the coil, how quickly it is released, and how much supply capacitance you have, it could end up moving VCC -- in theory, anyway. The only downside I can think of is that it's not physically as close to the coil so there could be a bigger "loop area" and end up creating more EMI. If the transistor is right next to the relay on the PCB then it ought to be fine, but if it's a big off-board contactor with hundreds of milliamps of coil current or something it could be problematic.

[HighVoltage]

Great video, thanks.

[rs20]

Alternately, use just the zener but connect the anode to ground. Then you not only don't have to worry about the supply voltage (because the diode is directly across the voltage-sensitive part i.e. the transistor), but it also means the coil current isn't being dumped into your positive supply where it could end up affecting other circuit components.

Er, I'm not sure that makes sense... with the arrangment in the video, isn't the relay current coming from the supply, through the relay, through the diode, and then back into the same supply? That cancels out, doesn't it? There's certainly no risk of the supply voltage being raised dangerously... unless the sudden cessation of current draw by the relay circuit causes a similar effect due to a supply with a ridiculously high inductance itself.

[Circlotron]

^^ A zener from emitter to collector, "pointing" upwards. I use a BZX79C75 with 80V BC639's and 12V relays all the time. Nice fast release time and the relays never stick. This can be a problem in some circuits if the relay de-energises too softly because of a diode across the coil.