PIR switching relay problem

[garysmith]

Could anyone advise me please. I have a external PIR which I am using to switch on a small din mounted 230V relay (for the contacts of a PLC). The problem is, although the PIR is only on for a few seconds, as the time on is at its lowest setting, the relay still holds in. I suspect that this is to do with the fact that the load is both very small, and purely inductive. I seem to remember that if a test meter or other load was put in parallel with the coil terminals, the problem disappears ie the coil switches off.

I was thinking a suitable resistor as a discharge maybe?...

Thank you

 

[Harald Kapp]

A normal relay can't hold if there is no current through the coil.
If you turn of the current through the relay's coil, the stored energy within the coil will try to force the current flow to continue. But since there is no driving source nor a load on the coil, the voltage will rise to very high values possibly giving rise to a spark. But eventually the energy in the coil will be spent and the relay will turn off. It only takes longer, but not forever.
To reduce the voltage peak and the turn-off time, DC operated relays often use a freewheeling diode in parallel to the diode (but inverse to the controlling voltage).

Therefore I assume your problem has another reason.

When you say "230 V relay", do you mean that the coil is operated with 230V? Or is the relay operated by a small control voltage and only the contacts are rated for 230V? If the relay is AC controlled, a single freewheeling diode alone will not help. You would need an anti-series connection of two zener diodes or suppressor diodes with appropriate voltage rating (~250V).
Since you say the PIR is turned on only for short periods, a resistor may help, too, but a resistor consumes additional power during the turn-on time.

What do you mean by "the relay still holds"? Does it not turn off at all, or does it take only longer than expected? If it doesn't turn off at all, have a look at the specs. There is a relay type called bistable. This type will be turned on with one control signal, but needs a second control signal to turn off. This can be realized either by two separate coils or by using control signals with inverse polarity. Neither a resistor nor diodes will ameliorate your problem in this case.

If your relay is AC operated ANd it only takes longer for the relay to drop off than you expect, the reason may be buried in the construction of the relay. AC relays often have a built-in short circuit ring to generate a magnetic flux which is phase-shifted to the main flux. This auxiliary flux helps to bridge the time during which the main flux is zero due to the zero crossing of the AC control voltage or current. This reduces humming of the relay.

Harald

 

[duke37]

Relay can stick in if there is sufficient remanent magnetic field. DC driven relays sometimes have a little screw to keep the pole pieces slightly apart to ameliorate this situation. This can wear after many operations and may need adjusting.

I have never seen such a screw on an AC driven relay.

 

[garysmith]

Thank you for the replies.

Duke, it is a 230VAC small plug in type (OMRON) new relay.

Harald, I am using a 230V AC (coil) relay. The purpose is to supply a switching contact for a input to a PLC which is 24V. The PIR that is supplying the relay coil is a standard DIY type, and it did say not to use inductive loads as I recall (to protect the tiny PCB mounted relay I would presume).

I still think maybe a resistor across the 230V coil as a shunt may work, but what type or resistance I wonder? The fact that a small amount of power would be used (depending on the resistor) wouldn’t be a problem, as the coil will only be energised for a few seconds when the PIR is activated by passing traffic.

What kind/value of resistor would you recommend?

Thanks again.

 

[Harald Kapp]

I'd start with 100kOhm working downward if this is too high a value.
Take care not to overload the output of the PIR when the resistor comes near 47 kOhm.

Harald

 

[garysmith]

I'd start with 100kOhm working downward if this is too high a value.
Take care not to overload the output of the PIR when the resistor comes near 47 kOhm.

Harald

Hi, once again I haven’t explained too well. When I said standard PIR I meant the exterior type which are meant for security lighting, with a variable time output (light on), dawn/dusk etc. They will cater for up to 2000W, but in this case all I have on the load is the relay I mentioned. I will give it a go with the 100k ohm... And work down if necessary.

Thanks.

 

[duke37]

If the PIR can switch 2000W resistive, then it should not mind a small relay but to be sure, put a small light bulb across the relay to absorb any voltage spikes.

I do not understand why the relay is sticking in unless there is some current leakage in the PIR. A small load (the light) may help this.

 

[mechtronics]

are you just running the relay straight from the pir or have you any additional internal circuitry to control what is happening. ??

 

[mechtronics]

i am currently working on something very similar to it but you will need an additional bit of circuitry in order for it to act like that. i used a 741 op amp, a 7404 inverter and a 555 timer. my pir has two variable resistors on it. one for distance, and one for length of time it stays on i think the max time that you get out of the pir is about 3 minutes. but i have used a 555 timer to make mine stay on for longer basically when the pir comes on the 555 timer can make it stay on for anything up to half an hour or so.

 

[garysmith]

Still have issues

I had to leave the project for a while, but now on it again. I tried a 100k resistor without success. Just for the heck of it I rigged up a 60w lead lamp across the relay coil terminals, and that worked like a charm. When the PIR goes off, the relay drops out without issue, and lamp goes off). I tried my neon tester across the relay coil terminals (without the lamp connected), and when the PIR is open circuit there is a showing of about 50v, which is enough to make the relay light up and buzz. If I measuere the voltage without the relay in circuit, I think I recall that it showed around 200v (with PIR open circuit), but, this was with a neon tester. … I should really try again using a meter I guess.

So, it may be a quirk of the PIR that it passes a small leakage current if a small inductive load is connected. Strange!

I don’t want to have to “wire in” a lamp for two reasons. Firstly because where the board is sited, I don’t want a lamp that flashes on and off every time the external PIR is tripped, and secondly, if/when the lamp blows, it will mean that the PIR, and ultimately the whole PLC system which relies on it will be compromised.

So, I guess I am left to trying ever decreasing resistance using standard or wire wound resistors. Eve this is annoying in a way, as it is just another component to fit into the enclosure, and one that may require space and a heat sink, depending on how far I have to take it down.

I wish my understanding of electronics was better, as there must surely be a better way to do this?

I would try different PIR’s but that could be a common theme with them, also, the one I have was quite unique, in that it has a corner bracket as part of its make up. I also have a spare of the same, as the store had only two when I bought it.

If any of you clever sorts have any other ideas I might try I would welcome suggestions, else I will have to go down the trial and error route as stated above.

Thank you.