Preventing Arcing Across Relay Contacts

[cepheid]

Hi,

I'm switching an inductive load. Let's say a DC motor (20 A continuous current, 40 A peak surge) is going full on in one direction. I want to shut it off. Its power circuit is switched by the contacts of a very large electromechanical relay. Two questions:

1. First of all, I want to dissipate all of that magnetic energy safely without my control circuit facing an enormous back EMF. So I was thinking of putting a flyback diode in parallel with the load. Something like this:

http://www.nteinc.com/specs/6000to6099/pdf/nte6013.pdf

Is this a good idea?

2. I want to prevent arcing across the relay contacts. I came across this ridiculously old (1957) paper on the subject. Heh, it even predates the IEEE.

http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=01135909

I'm a bit confused by what they're saying. On the front page, they say that a certain voltage is required for arcing, because the contact material has to be vapourized and ionized. But a bit further down they contradict themselves and suggest that it is in fact dV/dt that matters more. Their first plot clearly shows that a set of contacts protected by a capacitor reaches a higher voltage than an unprotected set, however it rises much more gradually. So which is it?

Regardless of the answer, I know I need a capacitor across the relay contacts in order to give the current somewhere to go as the contacts are opening. However, I'm not sure how big it needs to be. Also, the paper suggests that a resistor should be put in series in order to reduce the discharge current of the capacitor that, when combined with the load current (that is re-established when the contacts are closed again) might tend to weld the contacts together. However, it also points out that resistors are not very good, and that diodes might be better. The paper doesn't talk about inductive loads at all. Anyway, my second question is actually two questions:

How big a capacitor do I need, and how do I make this determination?

Does anything need to be in series with it?

Here is the relay I am thinking of using:

http://www.kgtechnologies.net/products/100.html

I would appreciate any insights those of you who are experienced in such matters may have.

 

[dlgoff]

If it's a DC motor, then a diode that can handle the current will work. Resistor-capacitors circuits will also work. Here's a little article about how to size R and C.
http://www.industrologic.com/mechrela.htm"

 

[cepheid]

Hi dlgoff

Thanks for the link. It's interesting that two of the sources give fixed sizes for R and C (which seems weird), whereas the third suggests that they should scale with current and or voltage. If I follow the rules for this source I get (V = 28 V, I = 20 A):

R = 28 - 56 ohms
C = 10 - 20 uF

Although the R value is consistent with the other sources, the capacitor is quite a discrepancy (tens of microfarads vs. tenths of microfarads). I'm inclined towards the tens of microfarads. What do you think?

Also, when you say that a diode that can handle the current will work, do you mean in series with a cap across the relay contacts (like in that paper)?

 

[denni89627]

A time honored technique to prevent arcing across relay contacts is to series the arcing contact to another pair of contacts (as long as they are not too close to the arcing one). This will basically double your air gap and help prevent arcing.

 

[dlgoff]

I'm inclined towards the tens of microfarads. What do you think?

Either would help suspress the arcing but I would think the 10-20uF is would be better. At say 50 working volts, they won't be physically too large to work with.

Also, when you say that a diode that can handle the current will work, do you mean in series with a cap across the relay contacts (like in that paper)?

The diode would be across the coil to short the back EMF. So for a DC coil the Cathode of the diode would be connected to the positive terminal of the coil and the Anode connected to the negative terminal.

Edit: The diode is not really for suppressing arcing but will suppress noise that could effect the circuit doing the switching of the coil.

 

[cepheid]

The diode would be across the coil to short the back EMF. So for a DC coil the Cathode of the diode would be connected to the positive terminal of the coil and the Anode connected to the negative terminal.

Right, as I said in my first post, I've already got a flyback diode in there, to deal with the back EMF (because I'm dealing with an inductive load). But my understanding is that this is a separate problem from arcing across the relay contacts (which might happen even if the load
weren't inductive) and therefore I need a separate solution for it (resistor + cap as discussed)

 

[dlgoff]

I'd just try the cap and resistor and see how well it works.

 

[Danger]

I'm going to chirp up from the 'total ignorance' perspective. Couldn't you just flood the relay with oil or grease?
Really, I have no clue, but it seems to me that it would suppress arcing while still allowing continuity when the points are closed.

 

[dlgoff]

With big circuit breakers used in electrical substations, some use nitrogen (or other gasses) to blow out the arc. They're called gas-blast or puffer-breakers.

 

[famousken]

In r/c motors, we place schottkey diodes across the terminals of the motor to protect the mosfets in the speed control from overvoltage, the same thing would work for a relay