dc coil surge supression 11

[electricpete]

I was asked to provide a “quick” (i.e. not my real job…helping someone else out) general recommendation for dc coil suppression for relay coils powered from 125vdc. . There are about 30 different coils in the cabinet, all powered by 125VDC and all drawing about 0.25A or less. Various problems have been experienced that are believed attributable to voltage spikes from coil switching. Response time is not critical in this application… 1 sec delay would not hurt anything. It is important for relays to change state reliably and even more important not to short out the dc power supply if the surge suppression device fails.

I did a quick search and it has been discussed many times on eng-tips.
Also I found:

http://relays.tycoelectronics.com/appnotes/app_pdfs/13c3264.pdf

http://relays.tycoelectronics.com/kilovac/appnotes/fig48.asp

http://www.sprecherschuh.com/library/techdocs/get/TECH_Surge_Suppression_109.pdf

I’ve read through the above and formed my own conclusions, submitted for your comments.

I think the 2 most common discussed options are:
1 – flyback diode
2 – varistor.

The first link especially seems to push the option of varistors. They highlight a concern that flyback diode can make the coil be so sluggish that it might not even operate. Also apparently when it operates slowly, it’s output contacts can be degraded. I have to admit I have not heard much about these concerns before (other than time response).

I really don’t like varistors in this application. I think every time the coil switches open there can be fairly high current at high breakdown voltage drop and these things degrade. Sure there is a rating, but they use a little life every time they cycle. If they ever short circuit, life is not good.

So I like the flyback diode. Diodes can short, but then again I have an easy solution: put two diodes in series…. Makes me feel a lot better. Either diode can fail short and not a problem. Also I’m thinking I would put a resistance about equal to the coil resistance (R = 125V/0.25A = 500ohms) in series. That should tend to minimize concerns about effect of slow field collapse upon the relay discussed in the first link. I picked 1 times coil resistance since when the full coil current switches into flyback loop the voltage is limited to the original 25VDC voltage (if I had double the resistance I could have double the voltage).

In summary I am thinking about two reverse-biased diodes (during normal operation) and a resistor 1x coil resistance... all connected directly in parallel with the relay.

What do you think? Am I grossly overlooking anything?

By the way, any tips for diode selection?

=====================================
(2B)+(2B)' ?

 

[btrueblood]

FWIW,

I have spoken to people using high-performance (fast, high current, DC) relays, and they were adamant that any reduction in the transfer speed of the contacts would result in and increase in arcing across the contacts. I suspect they were also concerned about control responses within the system if relays were delayed, but that's a more difficult problem to quantify. I also know of cases (solenoid valve coils on small spacecraft thrusters) where any variability in coil drop-out delay would cause unacceptable variations in the controlled variable (impulse of the spacecraft thrusters in that case). The suggestions by Skogs of course are a good solution for epete's posted problem, but recall that resistors have a typically uncontrolled/variable TCR, and the drop out time can thus vary if the duty cycle of the resistor changes (i.e. it heats up and its resistance changes).

In those cases, the use of a backwards-facing zener of sufficiently high voltage is preferred, as the transfer times are much more repeatable and less susceptible to variation. Have to agree that the use of MOV's is not a good idea due to breakdown over time/number of pulses.

 

[electricpete]

there is no real need to worry about the diode break-down. If you were using 1N4148 - yes. But never with 1N4007. Never.

That’s valuable input, but can you explain why you say that? I’m not very familiar with various diode types and their differences.

I have spoken to people using high-performance (fast, high current, DC) relays, and they were adamant that any reduction in the transfer speed of the contacts would result in and increase in arcing across the [relay output] contacts

That seems to be the main message of the first article linked above (tyco), repeated here:

http://relays.tycoelectronics.com/appnotes/app_pdfs/13c3264.pdf

I take a small comfort from the fact that we already have a bunch of applications where someone has put in simple diode with NO resistor… at least we are doing better than that in trying to bring the current down quickly with the 500ohm resistor.

recall that resistors have a typically uncontrolled/variable TCR, and the drop out time can thus vary if the duty cycle of the resistor changes

The linked resistors have 100ppm/C spec and we don’t be cycling that often.

In those cases, the use of a backwards-facing zener of sufficiently high voltage is preferred, as the transfer times are much more repeatable and less susceptible to variation. Have to agree that the use of MOV's is not a good idea due to breakdown over time/number of pulses.

I’m not sure I follow. Is the “backward facing zener” configuration you’re talking about the same as Figure 3 in link above? In that one it seems like the zener takes the place of my resistor and must break down every time the coil is interrupted, which I think would be an aging mechanism.


=====================================
(2B)+(2B)' ?

 

[Skogsgurra]

Pete and BTB. I will be back later with comments. Have to do one of my lab things to find out about contact separating speed. I do not think that the diode makes a big difference. It delays the drop-out. But since the armature doesn't start moving until flux has reached the drop-out value, I see no reason why the separation speed would be affected. I will set up a few relays and do some 'high speed' recordings. Stand by.

Re the 1N4148. It is a small signal diode that I have seen used across relay coils. I do not like it. They break down easily and have caused paper machines to stop. The 1N4007 is a rectifier diode with 1000 V reverse voltage and 1A continuos forward current and 30 A half-wave 60 Hz current. It will not break down if the system isn't hit by lightning. And then, you have a few other problems.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[BrianE22]

We had a problem with contacts releasing too slowly when using just a zener diode across the coil. The contacts would burn and stick together. We changed to the Tyco method using a zener and a regular diode and haven't had any problems with the contacts since.

 

[Compositepro]

A small capacitor is often used to reduce contact sparking. It allows the coil current to keep flowing for a millisecond after the contacts open so they can separate far enough so the voltage cannot arc across.

 

[electricpete]

Thanks Gunnar. So we have a beefy diode in a relatively mild application so should not suspect many problems. It will be interesting to see what your testing shows about effects of slower dropouts. Although the linked article included traces on the input side of the coil they didn’t include anything related to the output contacts to prove their point.

Brian – thanks that’s a good datapoint.

Again at least I have a rectifier diode plus resistor in there, which I think is better than rectifier diode only… and the resistor roughly plays the same role as their zener (with the advantage that my resistor behaves much better during the hypothetical possibly-unlikely scenario of rectifier diode short).

Let me think about that comparison (my resister vs their zener) a little closer… in figure 3 at moment of switching the coil voltage changes from +100% to -200%. So I assume the zener breakdown voltage much be around twice the supply voltage. Is it too simplistic to think that quates to an initial resistance of Rzener_eq = 200%voltage/100%current = twice coil resistance? In that case my resistor is at least in the same ballpark, right? Maybe 150-200% coil resistance (1000ohm) would be better since I doubt 150-200% overvoltage would hurt anything.

=====================================
(2B)+(2B)' ?

 

[btrueblood]

"The linked resistors have 100ppm/C spec and we don't be cycling that often. "

Should be fine, and as stated, the variability in transfer time with even cheap carbon film resistors (where the spec. can look like +/-1000 ppm/C) doesn't sound like it would give you any real worries. I agree with Skogs, by the way, that (some) of the issues about transfer times are overblown. Certainly when switching AC relays, the AC current itself acts to snuff arcs. But with even 12 volt DC relays, again at current levels across the contacts approaching the ratings for the contacts, I have had experiences similar to Brian's, with contacts welding with only a diode for snubbing. But adding the resistor as Skogs suggests eliminates the worry about transfer speed, and is just as good if not more robust than a zener diode, and the transfer time variability is really only a concern to persnickety system engineer types (like those who send spacecraft to Jupiter and beyond...)

Yes, the backwards zener is as shown in your posted links, sorry. Zeners won't degrade any faster than the diodes will, i.e. the breakdown voltage won't change dramatically over time, at least not the way a MOV will. Assuming you get a zener that can take the heat pulse without releasing its magic smoke, which is not always possible.

 

[electricpete]

Zeners won't degrade any faster than the diodes will, i.e. the breakdown voltage won't change dramatically over time, at least not the way a MOV will.

I am still struggling with this. I think of varistor as roughly a pair of back-to-back zeners. The varistor breaks down whenever it conducts. The zener in figure 3 will break down whenever coil is deenergized causing it to conduct. Why wouldn’t this cause similar aging in the zener as we see in varistors? I can see that the shorting failure mode of the figure 3 zener (loss of damping during the deenergization) is not as severe as the failure mode of a typical varistor (shorting of the power supply voltage), but I think they are both seeing similar stresses, right?


=====================================
(2B)+(2B)' ?

 

[electricpete]

....Maybe it's the continuous voltage that gets the mov (?)

Anyway that's actually not a big factor in our currrent thought process. The bigger problem with the zener in our mind would be that it doesn't handle shorting of the rectifier well which is the main reason I prefer the resistor. Gunnar may be very right about the rectifier diode being buletproof, but we just don't want to take any extra chances of shorting out that power.

=====================================
(2B)+(2B)' ?

 

[electricpete]

Let me think about that comparison (my resister vs their zener) a little closer... in figure 3 at moment of switching the coil voltage changes from +100% to -200%. So I assume the zener breakdown voltage much be around twice the supply voltage. Is it too simplistic to think that quates to an initial resistance of Rzener_eq = 200%voltage/100%current = twice coil resistance?

I'll answer that. Too simplistic. A a closer look at the graph shows that for the 2 millisecond period after interruption the voltage is relatively flat at -200% while the current is dramatically decreasing from 100% toward zero. So I guess the equivalent resistance of the zener averages much higher than suggested by my 200% number above…. and if quick deenergization is the goal it does much better in that particular respect than suggested by my earlier discussion.

=====================================
(2B)+(2B)' ?

 

[Skogsgurra]

There are two entirely different physical effects at play in zeners and varistors.

The zener breakdown is an avalanche effect and there is nothing being 'worn'. The varistor works with grain boundaries that conduct when the voltage between them excedds a rather poorly defined, but still characteristic, limit. Each time that happens, some of the existing grain boundaries 'wear out' and after a number of pulses, there will be a (mostly) short. The number of pulses depends on the energy contained in every pulse. Not linearly, but monotone.

Doing mesurements. There seems to be a dependency between current rate-of-fall and contact opening speed. My first measurements are quite crude. I have set up a relay where the diode can be switched in and out. There is also a resistor that can be switched from 0 ohms to megohms. I measure the contact speed using the time it takes for the C to move from NO to NC.

They delay is, of course, very dependent on diode/resistor and there is a small variation (around 1 ms) between diode withouut resistor and no diode. The difference in travelling time (equals speed) is difficult to see when diode+500 ohms and no diode are used.

More tomorrow.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[electricpete]

Thanks Gunnar. I will try to erase from my brain the idea that a varistor is like a back-to-back zener.

=====================================
(2B)+(2B)' ?

 

[btrueblood]

Good description of zeners vs. MOV, and even gooder data on the transfer speed, Gunnar.

Oh, and I forgot one other reason that probably explains why some like zeners: if there are spike-sensitive electronics attached somewhere in the path of the coil (e.g. driver FET's instead of relay contacts), the negative-going pulse from the coil flyback could zot them; the zeners give a limited voltage spike, whereas a resistor controls the spike magnitude less closely. If that makes sense. Some people add blocking diodes to prevent that from occurring, but that adds a voltage drop too.

 

[Skogsgurra]

Done some measurements. You are right BTB. Read all about it in the attachment.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[sibeen]

Gunnar,

On the oscillograms you've labelled them:

Oscillogram 1. Free wheeling diode without series resistor.

Oscillogram 2. Free wheeling diode with 4.7K series resistor.

Oscillogram 3. Free wheeling diode with 4.7K series resistor.

Should that be:

1. No snubber
2. Diode plus 4.7k
3. Diode only

?

BTW, very interesting

 

[electricpete]

I think it's
1. Diode only
2. Diode plus 4.7k
3. No snubber


=====================================
(2B)+(2B)' ?

 

[Skogsgurra]

There were a few more errors in the comments. Also, the discussion presented the recordings in reverse order, which may have added to the confusion. I have corrected it in the attached pdf and hope that it is more clear now.

For those that think that the word 'snubber' should be reserved for an RC unit, I agree, but have used it in a broader sense. In this case for a single free-wheeling diode as well as for a diode/resistor combination. Anything that 'snubs' is a 'snubber' - isn't it?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Skogsgurra]

Sorry was busy correcting the text. Yes, you are right, Pete.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[electricpete]

That's great data Gunnar. A good idea to use the change of 2 contact states that way as indication of speed.

I can imagine there is a wiggle as the magnetic circuit breaks and there is a reaction trying to keep the flux constant. It is basically the same principle as the voltage that lights the LED when I break the hinge off of the magnet in the video below (except that instead of me supplying the force to break the hinge open, the spring is supplying the force):

http://www.youtube.com/watch?v=HI5ERdvroec&playnext=1&list=PLA7FF4251D29DB01B

I'm not sure I understand if and why we'd expect the electrical contact to break exactly at the local extreme value of that wiggle.. or maybe just a coincidence of timing. Not a big deal, just a curiosity.

* It is interesting to compare graph 3 (2 megaohm resistor) to graph 2 (4k resistor plus diode) after graph 3 settles down. At first glance, we'd expect the L/R time constant graph 3 to be on the order of 500 (2 Meg/4k) times shorter (faster) than graph 2, but it's not.. it looks like not even twice as fast. I can't explain that. Now another question... if we look for a "dropout voltage" we see it is roughly the same on slides 2 and 3.... but that means the current in slide 3 (2 meg resistor) was roughly 500 times lower than slide 2 at time of dropout. That's a weird thing... suggests maybe there are some complicated dynamic effects to consider.... or... not sure what it means.


=====================================
(2B)+(2B)' ?

 

[sibeen]

Gunnar,

I'm quite surprised by the amount of inductive kickback shown by the coil when the resistor is included in the circuit.

Is it enough to damage any relay driving device?