3-way slide switch (SP3T) to control model traffic signal

Thread Starter

PToms

Joined Dec 31, 2018
3
I have a SP3T switch with 4 pins or leads as shown in the attached datasheet. I am trying to figure out how to create a simple circuit that uses this 4-pin switch to switch between three LED's, as in a traffic signal light. I can't see how to make it work, given that the two middle pins on the switch are reused in the throws (not independent of one another). Is it possible to do this without logic circuits/relays/other switches?

Any help would be appreciated.

PT
 

Attachments

ElectricSpidey

Joined Dec 2, 2017
2,758
I would say no, because it looks like the center position is “off”.

You probably want a rotary SP3T.

One common and three switchable poles.
 

sghioto

Joined Dec 31, 2017
5,381
The only way I see it working with that slide switch is to add a transistor. Component values will depend on the working voltage and LED type.
SG

EEE SP3T  LED switch.PNG
 

ElectricSpidey

Joined Dec 2, 2017
2,758
If the center pins are shorted, as the OP seems to imply...that wont work.

The diagram also seems to indicate that the center pins are shorted.
 

Thread Starter

PToms

Joined Dec 31, 2018
3
I can confirm from my multimeter that the switch has three positions, and those three positions result in these connections between the four pins:
pos 1: pins 1-2
pos 2: pins 2-3
pos 3: pins 3-4

this batch of switches were sold to me under the pretext that they were "on-on-on" 3-way switches. At this point it would be really difficult to change to different switches (my mistake in not verifying their function beforehand), so I am trying to figure out if there's a way to use them as I had intended.
 

ElectricSpidey

Joined Dec 2, 2017
2,758
Of course the OP just tested otherwise...:rolleyes:

PToms....

If the center pins are not shorted as you have pointed out then the switches are indeed “on-on-on” and you can use them, but you will need additional circuitry.
 

sghioto

Joined Dec 31, 2017
5,381
PToms,
If you are sure the switch is a non shorting type then try my circuit in post #3. Even if it is a shorting type the switch can work using two transistors.
SG
 
Last edited:

MisterBill2

Joined Jan 23, 2018
18,179
It can work quite easily if you have two isolated voltage sources. And you do not need any extra diodes or transistors. Each LED will need to have it's own current limiting resistor, but since different colors have different voltage drops that should not be a challenge. With the terminals numbered as in the drawing, the red LED negative (cathode) connects to switch terminal #1, the positive(anode) connects to source #2 positive. Switch terminal #2 connects to source 2 negative, and switch terminal #2 also connects to the yellow LED positive (anode), while the cathode of the yellow LED connects to source #1 negative.
Switch terminal #3 connects to source #1 positive.Switch terminal #4 connects to the green LED positive end while the green LED negative terminal connects to source #1 negative. The only hard part is needing two isolated voltage sources. But if you are controlling a lot of traffic lights with a lot of these switches that should not be too bad.
 

Tonyr1084

Joined Sep 24, 2015
7,853
My first thought was that it couldn't be done. But after reading what @MisterBill2 said, I started to wonder HOW it was set up. This is what I gleaned from his description:

Each LED has its own resistor (LED's and Resistors not numbered or values given). Each circuit is its own circuit and will only work when the switch is in the appropriate position. Hat's off to MisterBill2. I don't think I would have come up with this solution.

[edit] As an afterthought, at all times one of these LED's will be on. You may need a master switch (DPST) to be able to shut the whole thing off.

01 Traffic Light.jpg
 
Last edited:

djsfantasi

Joined Apr 11, 2010
9,156
. . . Hat’s off to @MisterBill2. I don't think I would have come up with this solution.

[edit] As an afterthought, at all times one of these LED's will be on. You may need a master switch (DPST) to be able to shut the whole thing off.
This is logically the case with ANY SP3T switch. For “Off” to be a desired output state, you then have four possible outputs. Hence, a SP4T switch is required. However, given the TS application, I don’t see that four states are required. All traffic lights that I am aware of, there’s always one position/color lit. Hence, his application only needs three states.
 

MisterBill2

Joined Jan 23, 2018
18,179
Yes Tony, your drawing shows exactly what I was thinking. I considered ways to do it without two supplies but they all involved more components or other more complex voltage schemes. With some zener diodes and a higher voltage supply it could probably be done with a single supply. But the TS wanted simple.

With the switches described there may be an off position midway between the detented positions, which could serve as an OFF position. But consider that real traffic lights are always on one color or another.
 
Last edited:

Tonyr1084

Joined Sep 24, 2015
7,853
My thinking was to use comparators and voltage divider networks to achieve the switching. But like you said, these were the parameters the TS gave. So I had no idea how to solve it. I like your solution; works sweet.

True, traffic lights are always lit in one configuration or another. However, with modeling, which is where I think the TS wants to go with this, when the model is not in use, the lights need to be off. Otherwise batteries will go dead. And I'm only assuming batteries will be used.

I messed around with some old dead CR2032 coin batteries. I wanted to deplete them before tossing them into a bin for recycling. So I slipped LED's on them to drain them. Those darn dead batteries burned the LED's for quite some time. For two weeks the LED's seemed to be quite bright. After four weeks they began to show signs of dimming. Six weeks they were still glowing and by eight weeks one of them could barely be seen as glowing while three others were still obviously lit. Took more than 10 weeks to completely drain all the batteries. And that was burning constantly with no resistance, lit day and night non-stop. So maybe fresh CR2450's will hold those traffic signal lights lit for quite a few months continuously without sign of dimming. (just guessing) The batteries in my car's "Tire Pressure Monitor Sensors" are CR2450's. I know this because I replaced a set on an 08 Tacoma. They lasted 8 years before beginning to fail. I still want to build a circuit that can read air pressure. I want to set up some sensors on the front of my Tacoma and see how much pressure and how much vacuum the truck is seeing at various speeds. Just a curiosity thing. Maybe in the coming year or two I might actually go after building such a device. It would be interesting to see how much wind pressure there is on the nose of my truck and how much that drops when behind a big rig tractor trailer.

Back to this thread: If the batteries are removable, then there would be no need for a switch to turn the traffic lights off. But I'm wondering how much cross-over there are on the contacts of the SP3T switch. It might be possible to set the switch into a dead spot so that no contact is made between any of the contacts. But then there's the possibility the switch could slip back into one of the detent positions and once again light an LED.
 

BobTPH

Joined Jun 5, 2013
8,814
I believe this works, but it relies on the fact that the 3V supply voltage is not enough to light up the red and yellow LEDs in series, which might not be true, they might light dimly depending on the parts used.

upload_2019-1-1_14-52-21.png

Edit: on second thought, the yellow and green should be the ones in series because they likely have a higher combined Vf.
 

MisterBill2

Joined Jan 23, 2018
18,179
I believe this works, but it relies on the fact that the 3V supply voltage is not enough to light up the red and yellow LEDs in series, which might not be true, they might light dimly depending on the parts used.

View attachment 166934

Edit: on second thought, the yellow and green should be the ones in series because they likely have a higher combined Vf.
I think that Bob has got it! Except that there would need to be a zener diode or two, and a higher voltage supply. As it stands, be advised that LEDs do illuminate a bit at much less than the rated forward voltage, at least a lot of them do. See Tony's post above. BUT my guess is that he grabbed new batteries instead of the package of dead ones.
 

Tonyr1084

Joined Sep 24, 2015
7,853
See Tony's post above. BUT my guess is that he grabbed new batteries instead of the package of dead ones.
I assure you, these were used batteries that would no longer perform their function. They were from garage door opener remotes, and they had been in my recycle bin for quite some time. It wasn't until I wanted to bring them in for recycling and wanted to be sure they were dead before I tossed them into a pile of other batteries so I started draining them via LED's. In fact, somewhere on this website I posted my tests. I'll see if I can find it again. If so I'll link it in, even though it's off topic.
 

Tonyr1084

Joined Sep 24, 2015
7,853
Found it! https://forum.allaboutcircuits.com/threads/coin-batteries-that-wont-die.129977/ After reviewing my post from December 2016 it turns out that my test lasted about one month. My memory isn't what it used to be, so in post #14 I claimed a little longer period than was factual. My apologies for my error. Still, using batteries that came out of service because of no longer being strong enough to transmit a signal to the Garage Door Opener (GDO) I haphazardly started my test.

If anyone wishes to comment on my test, please do so on the original post I made back in 2016. I do not want to hijack this thread, nor do I want to stray off topic. But just to reiterate, these CR2032's were used and too weak to perform their functions. And at some point in my test I switched from super bright white LED's to SB Red LED's. At this moment I have a CR2032 sitting on my bench. Came out of my doorbell transmitter. Too weak to ring my bell. Maybe this time I'll start a new test and report the findings. Till then, lets get back on topic please.

[edit]
Just started a new test using the doorbell battery that quit working. See linked post for more information.
 
Last edited:

MisterBill2

Joined Jan 23, 2018
18,179
Found it! https://forum.allaboutcircuits.com/threads/coin-batteries-that-wont-die.129977/ After reviewing my post from December 2016 it turns out that my test lasted about one month. My memory isn't what it used to be, so in post #14 I claimed a little longer period than was factual. My apologies for my error. Still, using batteries that came out of service because of no longer being strong enough to transmit a signal to the Garage Door Opener (GDO) I haphazardly started my test.

If anyone wishes to comment on my test, please do so on the original post I made back in 2016. I do not want to hijack this thread, nor do I want to stray off topic. But just to reiterate, these CR2032's were used and too weak to perform their functions. And at some point in my test I switched from super bright white LED's to SB Red LED's. At this moment I have a CR2032 sitting on my bench. Came out of my doorbell transmitter. Too weak to ring my bell. Maybe this time I'll start a new test and report the findings. Till then, lets get back on topic please.

[edit]
Just started a new test using the doorbell battery that quit working. See linked post for more information.
OK, Tony, and thanks for the clarification. The reason for the comment was because I have seen just exactly what I said, happen. Unknowing folks assuming that any battery not in the original package must be dead and worthless. Such folks are an endless source of grief and aggravation. Oh Well.
 

eetech00

Joined Jun 8, 2013
3,859
I believe this works, but it relies on the fact that the 3V supply voltage is not enough to light up the red and yellow LEDs in series, which might not be true, they might light dimly depending on the parts used.

View attachment 166934

Edit: on second thought, the yellow and green should be the ones in series because they likely have a higher combined Vf.
Hi

If brightness is an issue, a resistor might be used to shunt some current around the bright LED to make it dimmer.
Although some power would be wasted by the shunt resistor.

eT
 
Top