Well... I set up the following circuit and it didn't work as I expected. At this point, I'm no longer interested in "Why didn't it work?" question, but on "Why does it work like this?". Below I will describe how I think it should happen when I press SW1 and then I will tell you what actually happened when I pressed the assembled circuit.

My interpretation:

After pressing SW1, the circuit will light the three LEDs first. For each LED lit, the associated transistor will open and this will “open” the associated relay. Then it will light up the three LEDs and so on. So, only two situations are possible for me: 1- the led will blink very fast until at least one of the leds turns off (stable situation); 2- Nothing will happen.

What really happens:
The three LEDs light up and stay still. No current flows through the transistors.

what am i missing?
PS1. I did the following experiment on this circuit. Add some really high resistance just before BAT1. The only difference here was that the lights emitted by the LEDs were dimmer.
PS2. To keep the diagram from getting too big, I didn't add the diode/resistance used to "discharge" the relay.

 

hi ajob,
Is the connection in the RED box intentional.?
E

 

Yes

 

hi ajob,
OK.
One problem is that the relays are in the Emitters of the transistors, which means if you consider the voltage drop via the 330R and LED's and the Vbe of the transistors, from the 4.5V source, the transistor are not being fully On.

E

 

What really happens:
The three LEDs light up and stay still. No current flows through the transistors.

Where did you get that mess?

This is how it would be drawn more conventionally:

You can see that the bases of the transistors are shorted together and to ground. Using humps and connection dots together went out of style around 40 years ago.

It's more conventional to drive loads high side with PNP transistors.

EDIT: Just noticed that you didn't have the negative terminals of the batteries connected. They have to be connected.

 

yes, instead of common collector, you should use common emitter configuration. just move relay coil from emitter circuit to collector circuit. also there should be flyback diode across each relay coil. and how is the circuit supposed to work? why are there three circuits paralleled?

 

hi ajob,
OK.
One problem is that the relays are in the Emitters of the transistors, which means if you consider the voltage drop via the 330R and LED's and the Vbe of the transistors, from the 4.5V source, the transistor are not being fully On.

E

Thanks. I connected it on the emitters because I was told that the relay would need some amplifier to actived. So, if I increase the voltage from battery BAT1 it should work has I tought? Or, to protect the LEDs, should I use other transistor to amplifier the voltage on the base?

 

Where did you get that mess?

This is how it would be drawn more conventionally:
View attachment 278687
You can see that the bases of the transistors are shorted together and to ground. Using humps and connection dots together went out of style around 40 years ago.

It's more conventional to drive loads high side with PNP transistors.

EDIT: Just noticed that you didn't have the negative terminals of the batteries connected. They have to be connected.

Thank you very much. Does it have any nice literature about this? I'm most of self taught.

 

hi ajob,
This Simulation that you have about 2V across the Relay coil.
E

Added showing effect of a 9V driving voltage, looks OK

 

yes, instead of common collector, you should use common emitter configuration. just move relay coil from emitter circuit to collector circuit. also there should be flyback diode across each relay coil. and how is the circuit supposed to work? why are there three circuits paralleled?

View attachment 278688

yes, instead of common collector, you should use common emitter configuration. just move relay coil from emitter circuit to collector circuit. also there should be flyback diode across each relay coil. and how is the circuit supposed to work? why are there three circuits paralleled?

View attachment 278688

Thanks. I will try this new configuration. The initial idea is see if the circuit is going to stabilize by leaving at least one LED off. I know I can use some logic circuit for this, but I would like to avoid it, since on the future I would like to do something similar but using 10 leds.

 

hi ajob,
This Simulation that you have about 2V across the Relay coil.
E

Added showing effect of a 9V driving voltage, looks OK

Thanks a Lot

 

Does it have any nice literature about this? I'm most of self taught.

After pressing SW1, the circuit will light the three LEDs first. For each LED lit, the associated transistor will open and this will “open” the associated relay. Then it will light up the three LEDs and so on.

There is no way that the circuit you drew can do this.

My advice is to ditch the relays and just use transistors and passive components to do what you want.

If you can explain clearly what you want, someone might be inclined to help.

If you're going to draw schematics, learn how to draw them so they're easier to read.

 

There is no way that the circuit you drew can do this.

My advice is to ditch the relays and just use transistors and passive components to do what you want.

If you can explain clearly what you want, someone might be inclined to help.

If you're going to draw schematics, learn how to draw them so they're easier to read.

I get it. Thanks for the advisers and sorry for the mess.
As for the idea, as I said before, I'm doing this out of curiosity to see how these things will work. But I can explain what I would like to understand at the end of this project. By "understanding" I mean knowing if it's possible, impossible, or if I still need to work hard.

Imagine we have 10 LEDs, say L1, ..., L10. Let S_1, ..., S_k \subseteq {L1, \ldots, L10} be fixed sets. In principle, these sets can be any size, but to make it easier, I only consider the case that they have size 3.

I would like to build a circuit with these 10 LEDs in such a way that once I turn the LEDs on, for any set S_i not all LEDs in S_i are either on or off.

For this circuit that we are working on here, I am considering only 3 LEDs and S_1 = {L1, L2, L3}.

I know we can use Logic Circuit for this. That's not what I wanted. I want to use the fact that the chain "always finds a way to unload" to force it to find the configuration if it exists. As in the diagram here.

 

It is difficult to understand what you are attempting to do.

Here is a simple circuit for turning on an LED.


When you close the switch the LED will turn on.

Now, can you describe in words what you are attempting to do?

 

Imagine we have 10 LEDs, say L1, ..., L10. Let S_1, ..., S_k \subseteq {L1, \ldots, L10} be fixed sets. In principle, these sets can be any size, but to make it easier, I only consider the case that they have size 3.

I would like to build a circuit with these 10 LEDs in such a way that once I turn the LEDs on, for any set S_i not all LEDs in S_i are either on or off.

That's clear as mud. Try explaining without trying to use set notation that you don't understand how to use.

Do you want only one LED to be on and for them to turn on sequentially?

 

Okay, you have 3 sets of 3 LEDs. What happens when you turn it on?

 

For now, I just want to understand the circuit from the beginning of this thread. And thanks to ericgibbs and panic mode, I think I got it. I'm going to assemble the circuit described by them to see if it works.

For the next step, I can only tell you what I want to achieve, not "how I'm going to do it". The reason is because I don't know how.

The question I'm answering with the circuit in this thread is "Is there any configuration of three LEDs so that they're not all on or off?". I know the answer is trivial, but if I can't make a circuit that gives me this trivial answer, how am I going to do it for the big ones?

The next step is to try it out with 6 LEDs. Suppose each LED is labeled as below
LED 1: AB
LED 2: AC
LED 3: AD
LED 4: BC
LED 5: DB
LED 6: CD

so I want to build a circuit with the above LEDs that once I turn it on, it "finds" a configuration where given a letter, say A, neither the three LEDs with A on its label are on nor off. I would like the circuit to find this setting for me. An example of this configuration is LED1, LED2 and LED5 on and the others off.

Soon I assemble the circuit I come back .

 

"Is there any configuration of three LEDs so that they're not all on or off?"

neither the three LEDs with A on its label are on nor off

I don't want to seem unsympathetic to your efforts, but I cannot make sense of EITHER of the above statements. "LEDs are neither on nor off"????

A clear explanation of the goals of your project is a necessary first step, for your own sake but even more so when you are soliciting assistance.

 

Well... I set up the following circuit and it didn't work as I expected. At this point, I'm no longer interested in "Why didn't it work?" question, but on "Why does it work like this?". Below I will describe how I think it should happen when I press SW1 and then I will tell you what actually happened when I pressed the assembled circuit.

What really happens:
The three LEDs light up and stay still. No current flows through the transistors.

This is what I would expect to happen. When you press SW1 the current flows through the three LEDs (and there resistors) and then through the three relay terminals that are in their NC (normally closed) position.

No current will flow into the base of any of the transistors because there is no closed circuit for any base current to return to BAT1. Without any base current, all of the transistors will be in cutoff and so all of the relays will be left unenergized and the relay switches will remain in their NC position.

My interpretation:

After pressing SW1, the circuit will light the three LEDs first. For each LED lit, the associated transistor will open and this will “open” the associated relay.

There is no association between an LED and a transistor. The cathodes of all three LEDs, the bases of all three transistors, and the common contact of all three relay switches are hard-tied to the same node.

Then it will light up the three LEDs and so on.

This makes no sense. You just got done saying that after pressing SW1 that all three LEDs will light up. So what does it mean to then say that something will then light up the three LEDs? And what does "and so on" mean.

So, only two situations are possible for me: 1- the led will blink very fast until at least one of the leds turns off (stable situation); 2- Nothing will happen.

The LED branches are in parallel, so either they are all three on or they are all three off. If you tie the negative sides of the two batteries together, you will have a path for base current to flow, but you may not be able to open any of the relays. What is the coil resistance of the relays? How much current is needed to energize them?

With at least one of the relays in the closed position, the bases of all three transistors will be tied to ground. So you won't be able to turn on any of the transistors at all. The circuit will be happy to sit there with about 7.5 mA flowing through each LED.

what am i missing?
PS1. I did the following experiment on this circuit. Add some really high resistance just before BAT1. The only difference here was that the lights emitted by the LEDs were dimmer.
PS2. To keep the diagram from getting too big, I didn't add the diode/resistance used to "discharge" the relay.

It's good that you recognize that you need those anti-kickback diodes on each relay. Without them, those transistors aren't long for the world.

 

For now, I just want to understand the circuit from the beginning of this thread. And thanks to ericgibbs and panic mode, I think I got it. I'm going to assemble the circuit described by them to see if it works.

For the next step, I can only tell you what I want to achieve, not "how I'm going to do it". The reason is because I don't know how.

The question I'm answering with the circuit in this thread is "Is there any configuration of three LEDs so that they're not all on or off?". I know the answer is trivial, but if I can't make a circuit that gives me this trivial answer, how am I going to do it for the big ones?

The next step is to try it out with 6 LEDs. Suppose each LED is labeled as below
LED 1: AB
LED 2: AC
LED 3: AD
LED 4: BC
LED 5: DB
LED 6: CD

so I want to build a circuit with the above LEDs that once I turn it on, it "finds" a configuration where given a letter, say A, neither the three LEDs with A on its label are on nor off. I would like the circuit to find this setting for me. An example of this configuration is LED1, LED2 and LED5 on and the others off.

Soon I assemble the circuit I come back .

I think, between this and your earlier post talking about sets, I have an idea of what you want, but it's unclear how you expect a circuit to be able to find the configuration that you seek.

But, first, let's see if I can phase your goal in simpler terms.

You have a bunch of LEDs and each LED belongs to one or more sets.

When the switch it turned on, you want the LEDs to be turned on or off in such a way that, for each set, at least one of the LEDs in the set is on and at least one of the LEDs in the set if off. As long as that condition is satisfied, the state of any other LEDs in that set doesn't matter.

Does that sound correct?

If so, how does the circuit know which sets you have assigned each LED to? Is this somehow hardwired?

In your example above, why couldn't LED1 and LED6 be on and the others off? Or just LED2 and LED5?

What is your circuit supposed to do if there are multiple solutions? What is it supposed to do if there are none?