Incredible!!! Electromagnetic pulse turns ON my circuit

[Enrique15]

Greetings, partners from this forum.

I have this problem:

I built an alarm that is activated with a Push Button.

I have this Push Button in the bathroom, so a sick person can turn ON an alarm outside the bathroom, for the people outside get notice of when to take out the person from the bathroom.
In the bathroom, there's a floor fan (big one, very high speed), connected to an outlet installed there in the bathroom that is connected to the lamp (because originally there wasn't a specific and independent electrical outlet).

So here's the scenario:

If the person pushes the button, the alarm turns ON perfectly the time I determined.

But the strange and FUNNY thing is:
when I turn OFF the fan (or any movement of its "velocities" switch, for example, to slow down the fan or to speed up the fan), the alarm (that I repeat, is outside the bathroom) turns ON each time I move the fan's switch.

And the alarm, as you can see, made up of IC 555, I feed it with batteries (so the whole package, batteries, IC555, speakers, capacitors, resistors...) is outside the bathroom... IN the bathroom is only the two cables connected to the Push Button.

My question is:
What should I add to the circuit to prevent the alarm of turning ON every time I change the fan's speed ???
Because I have the push button, through the cables, connected, one side to GROUND of the battery, and the other side, to the Trigger Pin (Pin #2) of the IC555. A simple design as you can see.
The only thing I added to the whole thing is a capacitor 10uF between + and - of the battery (as a filter). But it doesn't help with the "strange turn ON".

Or should I add something to (or around) the cables inside the bathroom (or even the Push Button) to prevent them to induce voltage to the IC555 (thing I believe is what makes it to trigger and of course, activate the second IC555 that turns on the speaker).


Any help, advice, procedure..... I'll appeciate it.

Thanks for your time, folks !!!

 

[sigint]

You can use shielded wire to the push-button (shield connected to your circuit GND), or you can use a ferrite bead on the fan power leads - the first sugesstion is the most likely to work.

SiGiNT

 

[flatulent]

You can also put a capacitor across the other end of the wires from the push button.

 

[nandopg]

The suggestion to shield the wires is good, but don't forget that you have to shield the cable on both ends (inside and outside the bathroom) otherwise you won't have any effect.
Another suggestion is to decrease the impedance for the interference seen by the wires coming from the bathroom. This can be done maybe with a RC ladder made up by 22Ohms resistors and good quality 100nF caps to ground in series with each wire.
Another thing is to try another path to bring the cable from the bathroom to the exterior.

NandoPG

 

[Enrique15]

WAO. I didn't expect answers so quickly. Thank you very much for the help.

But to understand a little bit:

If instead of shielding the actual wires, or using new cables (shielded ones), I go for the method of "resistor and capacitor to ground":

1. One of the cables from the push button is already connected to ground of the battery. Then, should I put the resistor and capacitor only in the path of the other wire (the one connected to the trigger pin) ?? Or in both cables anyway ??

2. I trigger the IC555 by connecting the "trigger pin" directly to ground of the battery through the Push Button, so I can't connect a resistor in series with the wire. How then am I going to connect the RC ladder ??


3. And if I choose to install the alarm circuit inside the bathroom to avoid spending more money in more cable (shielded wire), the way to avoid this "Unexpected Turn On" is the same from what you've already told me ??? Or should I do something different ???


Thanks again for your time and support.
And excuse me if I bother asking too much. It's just that I'm not such an expert in circuit building as you are.

See you later. 8)

 

[Cortex]

pull-low the trigger pin using a resitor

it sould work

 

[Tornado]

Place A Low Pass Filter on the input of your circuit with some debouncing circuit.

I had this problem on an alarm circuit too.

Tornado

 

[Enrique15]

From Cortex:

pull-low the trigger pin using a resitor

I understand from this that the resistor is in series between the push button ant the trigger pin from the IC555. But if I put a resistor, maybe when pressing the button there's a chance that the IC555 won't see GROUND to trigger its output.
Well, I'll check it out and tell you my results.


And thanks Tornado for the suggestion. I'll try it out to see if there's any change.


As I'm new in this of building and testing circuits, it still looks strange and funny for me, seeing that just moving the speed switch of the fan, it can cause the alarm to turn ON. I know about electromagnetic interference, but I never thought it could be seeing in such a small level as to change speed of a FAN near the wires that connect to the alarm.

Such little "noise" caused in changing the position of the speed switch of the fan can make this interference or inductance in the wires ????? :?: 8O

The IC555 triggers because of a LOW voltage in its trigger pin. Does this strange TURN ON from the fan speed switching mean that the noise generates a voltage "drop" in the wire, so the triggers sees "low voltage" even though no one has push the button ???
Why isn't it a "positive voltage" caused for the electromagnetic pulse ??

Or maybe sometimes the noise causes positive voltage, because sometimes I switch the fan, the alarm doesn't turn on.... Could it be an explanation ??

It's like 60% (or maybe 70%) of the times I move the fan's switch that the alarm turns on by itself, and the rest of the times it doesn't turn on.
Could be because sometimes the noise makes low voltage in the wire, and other times makes high voltage in the wire ???
Or is just coincidence ? And should turn on always the alarm.
Well, I'm not so professional with this of electromagnetic fields generating out of nowhere and making circuits go crazy, jejeje. :lol:

Cause I always thought noise is a "variable of the system" if my circuit is plugged to the electrical connection. But if I have a totally independen circuit feed up by batteries, I never thought I could have this strange behavior with external noises.

Well, one learns something new every day.

See you later.

 

[nandopg]

Your last posts made a bit clear the situation. What you have is a pair of wires coming from the alarm. One end of these wires is connected to the ground (chassis) of the alarm and the other one, when the push button is not hit, is floating without any contact to the alarm.
If the above is right, my guess is that some transient due to the changing in the current through the fan´s motor is being coupled to the alarm through the ground connection, since the other wire remains disconnected at all to the unit.
My suggestion, learning a little more about the problem, is that you use a photocoupler for completely isolate the two long wires to the unit opening, therefore, the noisy grounding connection.

NandoPG

 

[Enrique15]

Thanks Nandopg for keep helping me.

And thanks to you I can notice I didn't make myself clear.


Here's the scenario:

In a room, place, or world, with no other power supply besides the alarm's battery, this schematic could fit any use of the IC555.

But in my case, added to the scene is a Fan, with no buttons, but a Rotating Wheel as a switch. And here the problem is the "Mysterious Power" :lol: that the Fan Switch projects every time I turn it around.

So, wanting not to boring you all with the same questions and problems, but trying myself to understand a little more the situation:

Nandopg, you said:

some transient due to the changing in the current through the fan´s motor is being coupled to the alarm through the ground connection

How can it affect if there's an add or drop of voltage in the ground, if the "voltage measured" that triggers the IC555 is in the PIN 2 (Trigger Pin) ??

and:

since the other wire remains disconnected at all to the unit.

well, is disconnected from PIN 2 through the Push Button, but part of the wire connected to the PIN 2 is in the bathroom (electromagnetic pulse area).

So, maybe there is where the voltage is induced to the PIN 2.
That's why I was asking if the induction is always of LOW voltages, or if it's random (some times positive voltage, sometimes negative voltage).


Hope I can understand this electric problem through your help.
And thanks everyone for your support.

See you later.

 

[nandopg]

First of all I think the induction is coming through the ground because this is the only electrical connection between the alarm box and the noisy environment. The other wire is isolated to the box by the push button.
What I think is happening is the transient is changing the ground reference in the form of random induced voltage pulses. When this pulses are coupled to the IC ground the behavior of the chip is totally unpredictable.
Your last post reinforce even more this hypothesis. I still stick by the opinion that if you isolate the wires, both, through a photo-coupler you will solve the problem.

NandoPG

 

[Cortex]

test this shema and post result

 

[Enrique15]

Exactly, Cortex. That's the way I have connected my circuit.

Like this:

In some electronic issues they tell about connecting Ground to the Trigger pin through a capacitor (a capacitor between the push button and the trigger pin).
But I tried that, and I found out that with the capacitor, the IC555 triggers only the first time I push the button, because the second and next times, I couldn't get the IC555 to trigger. And is that the capacitor keeps the charge, so I had to discharge it with a resistor in parallel with that capacitor.
A partner from this forum told me about keeping that added resistor in the circuit, so the capacitor always discharge, and so I could trigger the IC555 everytime.

But then I thought, WHY ADD A RESISTOR AND A CAPACITOR? IT'S BETTER TO JUST CONNECT DIRECTLY TO GROUND.

But of course, with the resistor between Vcc and the trigger pin, so it sees a Positive voltage, until I push the button to connect the Pin 2 with ground.


For NandoPG:
If I connect the photo-coupler, I assume I have to add a resistor to the circuit, to connect Vcc to the anode of the diode inside the coupler. What value will be good?

And by the way: how is it possible to the "electromagnetic pulse" from switchin the fan, to induce a voltage in the circuit (through the ground cable), if one of the sides of this wire is isolated (while no one pushes the button)? I mean, how can a voltage be INDUCED if one of the sides of the circuit is not connected to anything (so, I don't have a closed circuit), and so there can't be any current ???

If someone can clear me this doubt, I'll be very grateful.

See you later. 8)

 

[nandopg]

When you have a chance just try to diconnect the wire that is tied to the ground of the box and try to provoke the interference. Please, let me know the outcome for this experiment.

As to the photo-coupler, the very best is to use an idependent batery to feed the led.

NandoPG

 

[Enrique15]

OK. Sorry for the delay. I was all day testing my alarm circuit to find out what was happening.

First of all, I need to explain something:
I was planning to install de alarm circuit inside the bathroom and only have the speakers in the outside rooms. And so, I bought cable that is called "for speakers" (maleable, with transparent cover). Well, the point is, for testing, I decided to install the circuit outside (so it could be easier for me to check it out) and I installed just the push button inside the bathroom. And so, I connected the wires "for speakers" to the circuit: to ground and the IC555 Trigger pin.

All I tell here is because I feel the problem is in the wire:

1. Today I took the circuit into the bathroom, disconnected the push button from the cable "for speakers", and connected the button with simple wires (those to make tests in protoboards) to the circuit. I connected the batteries, tested the alarm, everything ok, and later, switching the fan: and nothing happened. The alarm didn't turn ON because of the fan switching.

2. I connected again the button, through the wires I used previously, to the wires "for speakers". And again I took the circuit outside the bathroom, and connected it to the other sides of the "speaker wire". After I connected the battery, this is what happened:
--> I connected one wire to Ground. And after I connected the other wire to PIN 2 of the IC555, the alarm started by itself.
--> If I did the same in the oposite way (first one wire to Pin 2 and next the other wire to Ground), the same: The alarm turno ON.

3. After the alarm stopped, I kept the circuit connected, and I went to the bathroom. And did again the fan switching, and again I get to turn ON the alarm.

4. Returning to the circuit, I disconnected the wire in Pin 2, kept the other attached to Ground, and return to switch the fan. NOTHING HAPPENED.

5. Again in the circuit, I connected the wire to Pin 2 and disconnected the Ground one. And switched the fan again. This time the alarm TURNED ON.


What I see here, and after some more tests I made, most of the times there's a Positive Voltage that turns ON the alarm when connecting a wire to the PIN 2 from the "Master IC555" (the one that controls the other IC555 which drives the speaker).

And that's weird, because after many days of testing each part of the circuit (one at a time), they worked all good. And even two days ago, when I connected all them together, the circuit worked well (NEVER TURNED ON UNTIL I PUSHED THE BUTTON). But after yesterday, with all the problem of the "electromagnetic pulse" and stuff.... now that I test the circuit, happens what I mentioned before:

The first IC555 gives me a Positive Output that drives the alarm just by touching the Pin 2 with a wire side.

First I thought it was the IC that was damaged. But if I test again each IC555, they all work good. It's just when I connect everything together, that the circuit comes crazy.

I'm getting a Headache with this problem. :evil: The circuit was supposed to be simple. I planned it to be simple. Not to be giving me this strange problems.
I just love ELECTRONICS for the weird behaviours they can get some times.


Any help ???

P.D. Are the batteries part of this non-sense.
I see they can't handle the load from the speaker (it's supposed to load near 350mA when active). Because the IC555 loads are supposed to be little. Even if I keep the batteries connected all day long to the circuit.
Because after two days of use (the complete alarm, not just testings as I did before), yesterday the volume of the sound started to drop.
After I disconnected the batteries and did the tests today to post this message, now that I connect the alarm, is sounds loud again. But I think after a few hours the volume will fall again.
Is sad because I thought that, with the speakers only working a few times per day, the batteries were going to last me longer.

See you later.

 

[Enrique15]

Refering to the previous answer:

Why is that "speaker" wires induce this voltage to the circuit, while having the circuit near the Fan doesn't trigger the circuit ???

The distance is relevant here ??? Because I always thought that if having a circuit far away because of wire connection, will drop voltage to the circuit (and so, less induced voltage).
I have 20 feet of wire between the circuit and the push button inside the bathroom. Is too little distance to have significant effects ???
Why then the alarm turns ON by itself when connected to the "speaker type" wires ?


Thanks for any ideas.

 

[nandopg]

Dear Enrique15,

OK, so the transient doesn't´t look like to be induced through the ground and even when you disconnect the ground side of the wire keeping the other one connected you got the interference.

Let me tell you what my theory is:

1- You have by one side the fan´s motor fed by an AC low frequency voltage. So, every time you change the fan´s speed you provoke a current changing over the motor and this current change release through the air an electromagnetic field whose alternating is much smaller than the 50 or 60Hz that is feeding the fan´s motor.

2- By the other side you have a pretty damn big piece of wire opened through the free air, connected to the pin 2 of the IC. Electrically this piece of wire could be defining a secondary of a transformer whose primary is defined by the fan´s motor.

3- The return of the transformer´s secondary (the `wire for speakers`) it is not being made by anywhere in your circuit but somewhere in the room. This is typical in very low frequency propagation and interference, and this looks like to be true as you don´t need the ground terminal be connected to the circuit to see the interference effect.

Concluding: In this case I don´t know if any attempt to solve the problem hanging stuff from the IC´s pin 2 to ground will effectively work. I would suggest you a couple of things to try:

1- Try to connect the ground of the alarm´s circuit to somewhere that can work as a ground at the outside.

2- Try to feed the alarm with a bench power supply instead of a battery

3- Try to change the physical location of the circuit to see if you have any change in the situation.

4- If you have an oscilloscope, try to see the pulse that is applied to the pin 2 during the transient.

5- By the end just for a sake of science replace the wire for speakers you are using by a shielded cable. Try to ground just one side, both sides or any side of the ground conductor to see what is the effect.

NandoPG

 

[sigint]

Try reducing the resistance on the trigger wire to vcc from 22K to 10K or take your wires made for speakers, i assume it is 2 wires molded together and twist them, (this will create a little bit of shielding), 2 or 3 twists per inch, you'll probably
have to add wire since this will shorten them, or bite the bullet, and opto-isolate it, (shielded wire would probably be cheaper), since the motor control in the fan is probably PWM controlled it can produce either positive or negative spurious spikes.

SiGiNT

 

[10kangstroms]

I have to confess I didn't take the time to carefully read this entire thread, but...
did you try a 100nF cap from pin 2 to GND, right at the 555? That should work. You might want to also add a 1k resistor in series between the wire and pin 2, to damp any ringing that might occur.

 

[Enrique15]

Hello again.

At last I got the time to check my circuit, to test some of your suggestions.

Because I don't have much money to spend (I'm not working yet), I try to do whatever is the cheapest way.

And after all, the easiest and fastest way was:

ADDING A CAPACITOR BETWEEN PIN 2 AND GROUND. The easiest and most well-known method. The BASIC.

This is my first "100% made-by-myself circuit" since in college I didn't have the chance or money to do that. It was always a "five or ten member group" to build a circuit, so this is my first time doing everything alone.
But even though I built it with basic elements, I had this kind of "dumb" troubles. So I apologize for being too much ignorante in some things.
I need practice.

So, for the record, here's my final circuit:

Some aditional troubles I mentioned previously, I think they were caused for the low batteries. Because now that I'm trying NEW ones, the alarm is working ok.
I'll test it a few days to see how much will this batteries last.
In the future I'll practice building a Power Supply to feed it better. But that will be another story.

THANKS FOR ALL OF YOU THAT GAVE ME A LITTLE OF YOUR TIME AND KNOWLEDGE.

Hope that someday I can help you out in something too.

See you soon.