Voltage rail spike when switching heavy duty relay

[Olivier]

Hi all,

I'm building a pulse counter and have some phantom pulses being counted when the micro releases the relay.

I'm using an arduino mega 2560 to drive a transistor. (BC337)
The transistor switches a relay (Omron G5Q-14-EU)
The relay switches 230V ac line to a contactor (Schneider Electric iCT A9C20732)

There is a voltage ringing on VCC 5V when engaging but primarily when disengaging the contactor.
This leads to phantom pulses being detected.
When the contactor is not connected, no phantom pulses are detected.

Attached is the schematic, pcb and capture of the ringing.
1- Yellow being VCC 5V
2- Cyan being Pulsar 1 input
3- Pink being Nozle input (relay output is driven by this after debounce)
4- Blue being PulseOut
https://www.dropbox.com/sh/da25vt2naqago5d/AACF7Oi5xGS_Oyrvgo4O_Yeqa?dl=1

Can you guys give me some tips why this is happening?

[Signal32]

When the contactor is not connected, no phantom pulses are detected.

So you're still switching the relay on / off but you don't get any spikes ?
Have you tried manually switching the power to the contactor and seeing if that gives you a spike on your arduino ?
How are you providing the 5V ? Have you tried a battery to see if it still happens when powered via battery ?

[RoGeorge]

How do you power the Arduino?

It looks like the spikes are produced by the contactor's load, then they are propagated through the mains lines and through the power supply, upto the Arduino board.

Try adding some filters on the Arduino's mains power supply path, or at least a coil in series with the Arduino's +9V wire.

[T3sl4co1l]

Try a snubber across the contacts, say 0.01uF 250V + 47 ohm.

While it may be that your power rail really is that sloppy, I suspect you will find that you measure the same transients even when probing ground to ground.  That is, if you clip the probe tip to ground clip, and probe around with that, you will find the same signal on GND in most locations in the circuit, and the same signal on +5V in most locations in the circuit (assuming your circuit and oscilloscope are isolated of course, so it is safe to connect probe ground to +5V).

This occurs because common mode noise is coupled into the circuit from the contacts opening.  You need to provide an alternate path for that current to flow, or to dampen the voltage in the first place, that's causing that current to flow.

Tim

[Olivier]

When the contactor is not connected, no phantom pulses are detected.

So you're still switching the relay on / off but you don't get any spikes ?
Have you tried manually switching the power to the contactor and seeing if that gives you a spike on your arduino ?
How are you providing the 5V ? Have you tried a battery to see if it still happens when powered via battery ?

Yes. When switching only the relay without the contactor attached, no spikes are measured.
Manually switching the contactor using a switch gives the same spikes.
I am powering the arduino using a Vigotronix VTX-214-010-105 directly to the 5V pin.
Using a wall wart and an LM7805 results in the same observations.

How do you power the Arduino?
It looks like the spikes are produced by the contactor's load, then they are propagated through the mains lines and through the power supply, upto the Arduino board.
Try adding some filters on the Arduino's mains power supply path, or at least a coil in series with the Arduino's +9V wire.

Load or no load on the contactor does not seem to matter.
I'll look into the filters

Try a snubber across the contacts, say 0.01uF 250V + 47 ohm.

While it may be that your power rail really is that sloppy, I suspect you will find that you measure the same transients even when probing ground to ground.  That is, if you clip the probe tip to ground clip, and probe around with that, you will find the same signal on GND in most locations in the circuit, and the same signal on +5V in most locations in the circuit (assuming your circuit and oscilloscope are isolated of course, so it is safe to connect probe ground to +5V).

This occurs because common mode noise is coupled into the circuit from the contacts opening.  You need to provide an alternate path for that current to flow, or to dampen the voltage in the first place, that's causing that current to flow.

Tim

Well this explains it nicely.
I'll look into a snubber.

[TheDane]

I would suggest adding a diode across the relay, if not already done, as this discharges/shorts the reverse induced EMF energy when the relay is disengaged.

(I can't view your schematics, so I don't know if you already have implemented this)

https://en.wikipedia.org/wiki/Flyback_diode
https://electronics.stackexchange.com/questions/110574/how-to-choose-a-flyback-diode-for-a-relay

[Olivier]

I would suggest adding a diode across the relay, if not already done, as this discharges/shorts the reverse induced EMF energy when the relay is disengaged.
(I can't view your schematics, so I don't know if you already have implemented this)

I did in fact already added a flyback diode
I have included the schematic and the pcb as images

So you're still switching the relay on / off but you don't get any spikes ?
Have you tried manually switching the power to the contactor and seeing if that gives you a spike on your arduino ?
How are you providing the 5V ? Have you tried a battery to see if it still happens when powered via battery ?

Yes, but the Omron relay is not inducing the spikes.
I observe the same induced spiking when manually switching the Schneider contactor using a ZBE-101 switch.

Try a snubber across the contacts, say 0.01uF 250V + 47 ohm.

The snubber I made using a ceramic capacitor of 0.01µF rated for 50V and a .25W 47R resistor did not help.
I also had a RIFA PMR 209 100nF 100R lying around. This caused the contactor to remain engaged after disengaging the relay. Why does this happen?
I have ordered a few similar items with different values to test.
Is there any way to calculate the ideal snubber values?

How do you power the Arduino?
It looks like the spikes are produced by the contactor's load, then they are propagated through the mains lines and through the power supply, upto the Arduino board.
Try adding some filters on the Arduino's mains power supply path, or at least a coil in series with the Arduino's +9V wire.

I've added an EMIS MF410 in the mains path. Seems to have marginally to no impact on the results.

[T3sl4co1l]

The snubber I made using a ceramic capacitor of 0.01µF rated for 50V and a .25W 47R resistor did not help.
I also had a RIFA PMR 209 100nF 100R lying around. This caused the contactor to remain engaged after disengaging the relay. Why does this happen?
I have ordered a few similar items with different values to test.
Is there any way to calculate the ideal snubber values?

It depends on load.

When you say "engaged", do you mean you measured AC voltage, with a DMM, with no other load present?

That's such an example.  A DMM is a very light load indeed, so the impedance of the snubber allows voltage to pass (0.1uF is 31.8kohms reactance at 50Hz, much less than the 10M impedance of the DMM).

Which spike did it not help with, turn-on or turn-off?  The solutions are different for both cases!

Tim

[TheDane]

Diode - check.
Capacitors - missing?

I would suggest adding some 100nF decoupling caps across all your supplies. I would add some at the relay, and at the controller - at the very least, as these are high-frequency - 'high power' draining devices.
I would also recommend a few µF as a low-frequency decoupling item, you can easily do 1000's of µF on the supply - it 'shorts' fast and slow oscillating currents, so they do not cause trouble elsewhere in the circuit - like you seems to be experiencing.
- the result is a bit like a snubber works. Adds power when needed, removes excess power.
This is recommended for all designs!  Big caps are good for slow (ms), small are good for fast (ns) speeds.

There's no need to re-do the pcb for now, just solder the caps in parallel across the pins and do a function check again. I think it might solve your issues.
Thanks for attaching the schematics and the pcb-layout.

Good luck