Problem with my multiple relay circuit

[travisc]

I'll try to explain my problem clearly. I have a circuit with a bunch
of reed relays in it to switch on and off different parts of my
system. One particular relay in my system essentially electrifies a
coil to create an electro-magnet. When I switch this relay on and
off, the magnet works correctly. However, when the relay is switched
off, there is something happening that i don't quite understand yet.
When I switch off this particular relay, every other relay in my
circuit is momentarily activated (< 1 second). I'm thinking this is
because the load on my relay of interest is drawing a lot of current
and when the relay is switched off, there is some sort of shock
provided to some part of my circuit, causing all my relays to briefly
switch on.

Anyway, I hope this problem explanation makes sense and someone can
give me a solution I haven't tried yet. If I haven't provided enough
detail about my problem, let me know and I will fill in the gaps.
Usually, I'm pretty good at figuring out stuff like this, but I
honestly don't understand this one. Help would be very much
appreciated! Thanks.

 

[shrtrnd]

Probably collapsing electromagnetic field from the relay shutting off, and however your
circuit is wired, the other relays get a bump from that pulse.
A varistor should absorb that spike.

 

[travisc]

Thanks for the suggestion, but please forgive my ignorance in the following question:

How would I go about correctly incorporating a varistor to solve my problem. The varistor is a new concept to me as it was never taught to me in school. I've done a little reading and it seems like you're onto something there, but I'm not sure how to pick the correct varistor characteristics (voltage rating, clamping voltage, peak surge current, etc.) for my circuit. My problematic relay is activated by a 5V signal. When switched on, it allows 8V to the electromagnet. The rest of my circuit is completely 5V. So, would I need a varistor with a voltage rating of 8V? Or is this where clamping voltage comes into play?

Then, where do I connect the varistor? Does it go in series with my electromagnet?

Like I said, please forgive my ignorance on the subject, I have absolutely no prior knowledge of varistors before today, which I find kind of strange, because it seems like there is a lot of information about varistors out there when I did my initial search.

Thanks again for your help.

 

[(*steve*)]

Show us your circuit diagram -- that will be a great help.

The varistor should be placed in parallel with the coil. The voltage rating for the varistor should be above the supply voltage (12V perhaps?)

At this voltage, you could also look at a transorb, or possibly even back to back zener diodes.

However the best solution is probably just a simple diode across the coil. It should be reverse biased in normal use. You normally use a rectifier diode (1A 50V should be fine).

 

[shrtrnd]

*steve*'s right.
A varistor is essentially (electrically) back to back diodes.
You should find plenty of information on the web concerning diodes across the coil of a relay, to prevent your problem.
If you're unfamiliar with a varistor, the diode solution will work.
I regularly put varistors across the coil of relays that see constant use, to protect the coil and extend relay life (the coil being essentially (electrically) a big inductor during the energize and de-energizing of the relay). Remember from school, the old 'ELI' the 'ICE' man? 'E' (voltage) leads 'I' (current), in 'L' (inductor). And 'I' current, leads 'E' (voltage), in a 'C' (capacitor). In a relay (it acts like an inductor when you first energize it), you get a big voltage spike initially, and the the current follows later.
I know your situation is an attempt to limit the spike when you de-energize the relay.
Just giving you an idea of what's happening with your circuit.
Good Luck.

 

[travisc]

Thank you all for your help. I have added a link to the section of my circuit that I am referring to. As you can see, pretty simple design. The "Shutter" is essentially an electromagnet, so it's a high current load. There are five other similar relays in my design that all control different things (signals, solenoids, etc.), but none of them cause this backlash problem. However, they all temporarily activate when the "Shutter" relay turns off.

Like I said before, I probably sound like I don't know what I'm doing, but this is my first time really designing circuitry for a mechanical system like this. All my previous electronic experience is signal processing, so good capacitor placement solves 90% of those problems, lol.

Thanks again for your help!

 

[travisc]

To follow up, this thread has helped me quite a bit. The solution I've found that works best for my circuit is a varistor placed in parallel with my shutter. I tried several different versions of a 'snubber', each one seemed to improve performance, but not totally remove my problem. However a 26V (DC) varistor cleaned things up nicely. Before this problem arose, I had never even heard of a varistor, but I am familiar with the theory of its operation.

Anyway, thanks again to all who helped with this one, I very much appreciate your input!

 

[Resqueline]

Thanks for sharing the final solution with us! We're happy to have been able to assist & guide you. Sunshine stories like this brightens our day and keeps us going!

You seem to know quite a bit about decoupling, but if the effect still lingers on occation then a picture of the layout might help us suggest something additionaly.

 

[(*steve*)]

I second Resqueline's comment.

Did you try just a diode across the coil? What benefits does the varistor have over that (I ask because in a typical low power DC circuit, it (a diode) is what is typically used).

I am curious as to the benefits you see of a varistor over that.

 

[shrtrnd]

Just for trivia, I use the varistor a lot to absorb spikes in relay circuits.
There is another curious phenomenon in relay circuits I've experienced problems with in
the past. Depending on the design of the circuit, I've seen relay 'chatter', where the relay will rapidly engage and disengage many times when energizing or de-energizing, (before it actually engages or disengages), instead of just latching and unlatching. The darned varistor eliminates that every time. It's a cheap fix for a lot of relay related problems.

 

[shrtrnd]

Trivia update. Travisc put this question on another site. Two guys fighting over whether a transorb or back to back zeners would be better than varistor. One claims varistors degrade over time, and back to back zeners are a better choice. I don't know what voltage this guy was at, but thought I should mention his observation. Said he worked for a company that dumped varistors as unreliable. (First one I installed 5 years ago, is still working today, ...so I don't know). Just info I read from somebody else regarding this question. MIGHT merit consideration?

 

[Resqueline]

Sure enough, I've seen varistors slowly change their characteristics over time if they take a beating, and I've seen failed varistors.
I've seen failed transorbs too. They would fail (short) suddenly as opposed to varistors that could give signs of failing if monitored.
I figure transorbs have a sharper characteristic than varistors. Whether that in itself is an advantage I don't know. It would give rise to some more harmonics I believe.

I'd like to point out to anyone reading this that using ordinary zeners alone as transient absorbers is a bad idea, they just can't stand it for very long.
You have to add an ordinary (switching) diode in series with them and a low impedance capacitor in parallell with them if they are to have a chance of surviving.

 

[(*steve*)]

Using the right part for the job is often a good idea

I am still wondering what the problem would be in using a simple diode? The circuit was using DC wasn't it?

 

[shrtrnd]

The back to back zener guy claimed his zeners never fail, ...
the old 'always or never' syndrome, that just doesn't hold water with electronic circuits.
Good to know about the varistors changing characteristics, I'll be watching for that.
I'm kind of thinking the zener claimant, was using varistors not rated for the circuit, if they
were as unreliable as he claimed. (Or maybe not used as intended in their circuit.)
If varistors aren't useful, why would they still be widely used?
His original circuit HAD a diode across the relay coil, ... but he didn't show the full
circuit, ... I was assuming he had something else in his design, that was a contributing
factor to his problem. Anyway, he's got a fix for now, and he has ideas from the other
site (transorb and back to back zeners), if he wants to try something else.
It's kinda tough to know for sure exactly what happened, without the full schematic.
Anyway, thanks Resqueline and *steve* for the info, .... I learned from this thread too.