Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Using MOV to supress voltage spikes on coils

Status
Not open for further replies.

WasteWaterTim

New Member
I understand the concept behind using MOV's (Metal Oxide Varistors) to surpress voltage spikes on a collapsing coil.

My problem is this, I need to know, specifically with AC relays at 120 VAC if a spike can be created when the relay is first fired depending on the part of the cycle the coil is first engergized in. e.g switched on right at the point of zero crossing of the AC line vs turning on at the peak.

Additionally, as the coil of an AC relay in essence charges and collapses 60 times a second, what keeps the coil from spiking with every cycle?

Help! I need to understand this in detail.
 
The voltage spike can occur if the coil is suddenly de-energized (as by another relay contact or switch) when the coil is at other than the zero current part of its cycle. The nearer to the peak current this occurs, the larger the spike.

This voltage spike does not occur when you energize the coil (although you may get a current surge).

The spike doesn't occur during normal operation because you are not opening the circuit to the coil and trying to suddenly stop the current. Instead the current normally increases to the max and then goes to zero under the sine wave excitation of the power. But as you may know the current through an inductor is delayed by 90 degrees from the voltage. This is due to the inductance of the coil trying to keep the current flowing.

It's this same effect which causes a voltage spike when the coil circuit is suddenly opened. The inductance attempts to keep the current moving, but since it now has no place to go, there is a large voltage spike generated (limited primarily by the stray current capacitance).

Think of inductance as inertia that requires a push (voltage) to get the current flowing but then tries to keep the current moving in the same direction even after you remove the push (which then can generate a voltage in an to attempt to keep it moving in the same direction).
 
Last edited:
[FONT=&quot]I explain the “back EMF” generated by a collapsing magnetic field in a completely different way.
When a coil is de-energised, the magnetic flux (magnetic lines of force) (magnetic field) collapses and if this field is able to collapse very quickly, it cuts the turns of the coil and produces a voltage in the opposite direction that can be 10, 100 or even 1,000 times greater than the original applied voltage.
You could use a MOV to limit this reverse voltage but many transistors can only withstand a very small reverse voltage , so by using a diode to suppress the voltage , only a very small voltage is allowed to be produced (less than 1v).

[/FONT]
 
Last edited:
  • Like
Reactions: jjw
[FONT=&quot]I explain the “back EMF” generated by a collapsing magnetic field in a completely different way. [/FONT]
[FONT=&quot][FONT=&quot]When a coil is de-energised, the magnetic flux (magnetic lines of force) (magnetic field) collapses and if this field is able to collapse very quickly, it cuts the turns of the coil and produces a voltage in the opposite direction that can be 10, 100 or even 1,000 times greater than the original applied voltage. [/FONT]
[FONT=&quot]You could use a MOV to limit this reverse voltage but many transistors can only withstand a very small reverse voltage , so by using a diode to suppress the voltage , only a very small voltage is allowed to be produced (less than 1v). [/FONT]
[/FONT]
Using a collapsing magnetic field to explain the effects of inductance is fine but you are putting the cart before the horse. You say "if this field is able to collapse very quickly" as if it were an independent action, but it is not. How rapidly the field collapses is determined by the external circuit. If there is no path for the current when a switch is opened then the coil current must rapidly stop which causes the external back emf to build up very quickly (as determined by external capacitance) and that is what causes the field to rapidly collapse (V = L di/dt).

Standard diodes work fine for DC coils, but not AC, as the OP was discussing.
 
The voltage spike can occur if the coil is suddenly de-energized (as by another relay contact or switch) when the coil is at other than the zero current part of its cycle. The nearer to the peak current this occurs, the larger the spike.

This voltage spike does not occur when you energize the coil (although you may get a current surge).
The above is true if there is no contact bounce.

Switch & relay contacts usually bounce when opening & closing.

This means that the transition from open to closed and vice vesa is not "clean".

For example, when the contacts close, they can open briefly a few times before settling to a permanent closure.

The same applies in reverse when the contacts open. They may close briefly a few times.

For a relay operated by AC you could either:-

1. Use Zener diodes in reverse series or

2. Rectify the AC and put a normal diode across the coil.

For the Zener option, you would need to put enough of them in reverse series so the Zener voltage is greater than the peak AC voltage.
 
True, I neglected the effects of any contact bounce for the purposes of my discussion. But, as ljcox noted, any mechanical switch or relay will bounce when opening or closing which will generate inductive spikes across the contacts.
 
As this room is discussing a similar topic, perhaps someone can assist me with my own problem. I have a large 2Kva 100v ac Transformer which is feeding into a Bridge Rectifier the input to the Bridge Rectifier is being pulse switched with a 100A SSR. I have put an assortment of CR snubber circuits and MOV's across both the AC lines into the rectifier and the DC lines out of it. I have also put an MOV across the AC contacts of the SSR.

Despite all of this now and again the Rectifier goes short circuit and has to be replaced. Sometimes the SSR also goes short circuit and needs replacing, but less often. Both devices are rated for 1600v operation. which is well in excess of the operating voltage of my circuit. I have concluded that my MOV and filter circuitry is not helping as anticipated. I believe the failures are caused by back emf spikes from the transformer, produced when the SSR disconnects. I would like to know if perhaps I am connecting my MOV's wrong ?

It had occurred to me that if my MOV is directly across the transformer output as it presently is, when it clamps it may aggravate the problem by creating a closed loop instead of diverting the excess energy out of the circuit. So what I would like to know is this.... Should I be connecting an MOV from each side of the transformer AC output to my Earth grounding point instead ?. Also if I do so, how would I do it without creating a a further closed loop as both MOV's would be joined together at the grounding point ?.

Would greatly appreciate some help on this issue asap !.

Thanks
Rob
 
Can you post a schematic of your circuit as you do not say if the SSR is between the transformer and the bridge or between the bridge and the load. Can you also geve details of the nature of the load. An SSR sould not have the same problems as a normal relay as it does not cut the power to the load as soon as the drive signal is removed. It remains conducting until the next zero crossing of the current waveform.

Les.
 
Can you post a schematic of your circuit as you do not say if the SSR is between the transformer and the bridge or between the bridge and the load. Can you also geve details of the nature of the load. An SSR sould not have the same problems as a normal relay as it does not cut the power to the load as soon as the drive signal is removed. It remains conducting until the next zero crossing of the current waveform.

Les.

Sorry, no Schematic im afraid, as the whole thing came from my own head and has been designed as an ongoing project for the past two years. If I was asked to rebuild it without the original to copy, I couldn't.

The SSR is between the Transformer and the Bridge. The surge current figures can be pretty high and DC-DC SSR's in the 100A range can be very expensive. However DC-AC 100A SSR's are pretty cheap and easily obtainable, which is why I went that route.

The High current pulsed supply produced from the output of the Bridge feeds into a "very" large capacitor. This then feeds another device, which is is a not an easy one to explain as the nature of the load can change. It is an intelligent device which adapts itself to the supply, in order to get the maximum power. It works in a similar way to a Solar Inverter.
 
Last edited:
If you can't trace out the circuit and will not give the nature of the load I have no idea what the cause may be.

Les.
 
If you can't trace out the circuit and will not give the nature of the load I have no idea what the cause may be.

Les.

I am also 99% certain the cause of the spikes are when the SSR disconects between pulses.
 

Attachments

  • 20161203_155821.jpg
    20161203_155821.jpg
    58.4 KB · Views: 205
I understand the concept behind using MOV's (Metal Oxide Varistors) to surpress voltage spikes on a collapsing coil.

My problem is this, I need to know, specifically with AC relays at 120 VAC if a spike can be created when the relay is first fired depending on the part of the cycle the coil is first engergized in. e.g switched on right at the point of zero crossing of the AC line vs turning on at the peak.

Additionally, as the coil of an AC relay in essence charges and collapses 60 times a second, what keeps the coil from spiking with every cycle?

Help! I need to understand this in detail.
Hi WWT,

May I suggest simply turning on the solid state relay (SSR) at zero volts. You can use a standard chip to do this or do it yourself.

The SSR, if it is an AC type, can only turn off at zero volts, so you will not get a voltage spike at turn off.

By the way the voltage generated by an inductor is, -L * dI/dT, where, L is the inductance of the inductor in Henries, dI is the change of current in Amps, and dT is the time in seconds. It, thus follows that, if you disconnect an inductor from its current source instantly the voltage generated by the inductor will be infinite, in theory that is. Because the inductor is not perfect, and because of parasitic capacitances and resistances, the voltage will not be infinite, but it could be 2 to 5 times, typically for a relay, of the input voltage (that is causing the current flow).

spec
 
Hi WWT,

May I suggest simply turning on the solid state relay (SSR) at zero volts. You can use a standard chip to do this or do it yourself.

The SSR, if it is an AC type, can only turn off at zero volts, so you will not get a voltage spike at turn off.

By the way the voltage generated by an inductor is, -L * dI/dT, where, L is the inductance of the inductor in Henries, dI is the change of current in Amps, and dT is the time in seconds. It, thus follows that, if you disconnect an inductor from its current source instantly the voltage generated by the inductor will be infinite, in theory that is. Because the inductor is not perfect, and because of parasitic capacitances and resistances, the voltage will not be infinite, but it could be 2 to 5 times, typically for a relay, of the input voltage (that is causing the current flow).

spec

Hello

The SSR is a Zero Point crossover type and the control circuit already does this. However there are still spikes getting through !. The Inductor is not being suddenly disconnected from its source. The voltage into the inductor primary coil remains constant. However it is the load which is being disconnected suddenly and thus the spikes are being generated within the secondary coil.

I really did not want to get into an in depth conversation about the circuit design. And I don't want anyone to try and figure out ways to stop the spikes in the first place. I simply wished to know if I am using the MOV correctly by placing it across the output of the secondary coil. Or should I be using two MOV's linked together at an Earthing point ?.
 
Last edited:
Hello

The SSR is a Zero Point crossover type and the control circuit already does this. However there are still spikes getting through !. The Inductor is not being suddenly disconnected from its source. The voltage into the inductor primary coil remains constant. However it is the load which is being disconnected suddenly and thus the spikes are being generated within the secondary coil.

I really did not want to get into an in depth conversation about the circuit design. And I don't want anyone to try and figure out ways to stop the spikes in the first place. I simply wished to know if I am using the MOV correctly by placing it across the output of the secondary coil. Or should I be using two MOV's linked together at an Earthing point ?.

OK GGEL, got the message, I was answering the OP in error.

spec
 
Last edited:
OK GGEL, got the message,

I think that it has been explained to you that the normal sine wave voltage supply does not create voltages higher than itself.

So on to the voltage spike suppression that you inquired about:

It is normally best to suppress any transient voltage generators at source, if possible. In your case the only source, as you state, is the secondary coil, so it would be best to connect the voltage suppressors across the secondary coil, remembering to use as short leads as possible and also baring in mind that large currents will flow (for a short period) when the suppressor is doing its job so sturdy conductors and joints are necessary.

There are better suppressors than MOVs, which I was going on to describe, but you are obviously are not interested.

spec

I also have a CR Snubber circuit across the secondary side and have seen with my scope that it does help with the smaller spikes. However the larger spikes which are causing the problem are rare and may occur a couple of times a day. Unfortunately I am not going to sit an stare at a scope screen all day to catch it.
 
I also have a CR Snubber circuit across the secondary side and have seen with my scope that it does help with the smaller spikes. However the larger spikes which are causing the problem are rare and may occur a couple of times a day. Unfortunately I am not going to sit an stare at a scope screen all day to catch it.
See my revised post 15
 
See my revised post 15
OK GGEL, got the message, but I was not answering your post- you have hijacked this thread which is against the rules of ETO- you should start your own thread!

spec

I have hardly hijacked the post, you are seriousley over reacting !.... I had one simple question about the installation on an MOV, which should only have taken "one" simple reply. Plus this thread is quiet old with its last post in 2011 and was dormant when I posted to it. As I said, my question was so similar to the original poster, it made sense to see if I could get an answer here.
 
I have hardly hijacked the post, you are seriousley over reacting !.... I had one simple question about the installation on an MOV, which should only have taken "one" simple reply. Plus this thread is quiet old with its last post in 2011 and was dormant when I posted to it. As I said, my question was so similar to the original poster, it made sense to see if I could get an answer here.
Whoops, sorry GGEL- my error. I have modified post #15.

spec
 
If your rapidly switching the SSR, you may want to include some type of current limiting. My crystal ball is useless in this regard, but phase angle fired SCR's can operate into transformer loads when current limited. Commercial controllers can be had to do so and we used to do this all current limit, and at the time. I inherited this type of system at work. They were popping $25.00 SCR fuses at a rapid rate. Shorts after the transformer were common (exposed in a vaccum chamber). I made changes so that only a $1.00 USD fure would likely blow first and implemented current limit with the SCR unit (hooked it up in most cases), 25 A SCR units, and a 3AG fuse sized for the load (around 1Kw). The system basically operated 120 VAC into a variac to match our low voltage loads. with stupid analog metering.

Later, I tuck my nose out on the line and said (This new method I was proposing) would be cheaper and work better and it did. 1RU power supplies and a temperature controller with an isolated analog output and ramp/soak capability. The tough part was powering the mess, so we hooked a service panel to a 60 A 4 wire power cord and powered the stuff from there. The first use was a mix of old and new stuff. 1/2 wide 4RU stuff at 208V and 1 RU 95-240 V stuff. It saved a lot of labor in the long run.

The system was smaller in the long run. The only thing missing was a "true power readout" on the power supply, but we never had that except in one system on one heater. That power meter was like $900.00 USD. A power supply was $1500.00. The controller was obsolete.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top