A single switch to control two circuits

tcmtech

Joined Nov 4, 2013
2,867
I bought what I did because: I will be running it off a dedicated 12vdc 1 amp wall wart, it's rugged, the relay can be easily swapped out by our client without touching wiring, it was free next day shipping on Amazon. It won't be less reliable, will it?
Actually the is a good chance it will be. Large power relays tend to have issues with their contacts oxidizing and losing their conductivity due to lack of sufficient current being switched and will fail to conduct the tiny current rather early. :(

Also they're frigging loud plus personally if I bought a product from you and saw such a setup as this with a huge relay running off a separate 12 volt source when a tiny 5 volt one that could have easily been powered by the units own 5 volt source was cobbled onto the system all just to switch a logic level signal I would never buy from you again just from the simple fact it shows you clearly don't know what you are working with and thusly likely know as little about everything else in the system.

Would you deal with a mechanic who put a tractor tire on your car because it 'fit the rim and he thought he got a great deal on shipping' even though he was told multiple times by multiple people how to do it the right way plus could have gotten the right tire for all around less had he done it? o_O :rolleyes:
 

Tonyr1084

Joined Sep 24, 2015
4,737
Peanut gallery here - If I understood you correctly, you have a device to control. Input 1 and Input 2. When no one is standing on the floor switch (tape switch I think you called it), power goes to one of the two inputs - depending on your need. When someone steps on the pad power switches from one input to the other. If so - here's my approach to this:

Floor Switch.png
 

Thread Starter

Thenextman

Joined Feb 3, 2017
42
Actually the is a good chance it will be. Large power relays tend to have issues with their contacts oxidizing and losing their conductivity due to lack of sufficient current being switched and will fail to conduct the tiny current rather early. :(

Also they're frigging loud plus personally if I bought a product from you and saw such a setup as this with a huge relay running off a separate 12 volt source when a tiny 5 volt one that could have easily been powered by the units own 5 volt source was cobbled onto the system all just to switch a logic level signal I would never buy from you again just from the simple fact it shows you clearly don't know what you are working with and thusly likely know as little about everything else in the system.

Would you deal with a mechanic who put a tractor tire on your car because it 'fit the rim and he thought he got a great deal on shipping' even though he was told multiple times by multiple people how to do it the right way plus could have gotten the right tire for all around less had he done it? o_O :rolleyes:

Hahahah. Well. I think that's a bit dramatic. And I think there is a higher level of fault prevention actually with this setup. The relay works great once I found out what standard terminal labelling was.

Furthermore, I don't have exposed pins / breadboard / enclosure / all that crap with this setup. I can screw thing right into the box - done. If the relay malfunctions, unplug it and pop in spare. No soldering for untrained techs, no bs.

You have to understand that this is built for use over a total of 9 days. It is a one off unit. And the budget captures ineffiencies that come with building such a setup (this is a very small piece of a larger puzzle) in 3 weeks. It seems unlikely that it will corrode in 30 days.

My client won't know what the heck this is, and if anyone else working in my field saw it, they'd probably say "why didn't he use an arduino?!"
 

Thread Starter

Thenextman

Joined Feb 3, 2017
42
Peanut gallery here - If I understood you correctly, you have a device to control. Input 1 and Input 2. When no one is standing on the floor switch (tape switch I think you called it), power goes to one of the two inputs - depending on your need. When someone steps on the pad power switches from one input to the other. If so - here's my approach to this:

View attachment 123105
Thanks Tony. I appreciate your answer. There is no shared power source. What you recommend is actually how this relay would typically be used for an automotive application from my understanding. My whole thing is I just wanted to close loop between 1 & Gnd on component or 2 & Gnd on component without feeding power to the component. It is meant to be used with a simple momentary switch usually.
 

tcmtech

Joined Nov 4, 2013
2,867
You have to understand that this is built for use over a total of 9 days. It is a one off unit. And the budget captures ineffiencies that come with building such a setup (this is a very small piece of a larger puzzle) in 3 weeks. It seems unlikely that it will corrode in 30 days.
It will be fine then.

I know others who use gross over kill switches for mundane applications which bizarrely unjustified reasoning for it. I have a friend who uses a 100 amp lever action service disconnect for his shop lighting which draws at best about 12 amps. Reason being the crappy 49 cent light switches he used when he updated the shop lighting would burn out all the time and he insisted that if they burn out a good $5 - $10 commercial grade light switch would do the same so putting in a $200 100 amp service disconnect box was the only thing that would work.

Well the contacts hold up fine but the lever mechanisms for such disconnects were never designed for such frequent switching applications so his $200 disconnect boxes have the lever mechanisms wear out and break every 2 - 3 years and the whole box needs replacing because of it. Won't put a quality 20 amp rated commercial light switch in to save his life and to be honest he isn't the type that has $200 laying around to waist either. :rolleyes:
 

Thread Starter

Thenextman

Joined Feb 3, 2017
42
It will be fine then.

I know others who use gross over kill switches for mundane applications which bizarrely unjustified reasoning for it. I have a friend who uses a 100 amp lever action service disconnect for his shop lighting which draws at best about 12 amps. Reason being the crappy 49 cent light switches he used when he updated the shop lighting would burn out all the time and he insisted that if they burn out a good $5 - $10 commercial grade light switch would do the same so putting in a $200 100 amp service disconnect box was the only thing that would work.

Well the contacts hold up fine but the lever mechanisms for such disconnects were never designed for such frequent switching applications so his $200 disconnect boxes have the lever mechanisms wear out and break every 2 - 3 years and the whole box needs replacing because of it. Won't put a quality 20 amp rated commercial light switch in to save his life and to be honest he isn't the type that has $200 laying around to waist either. :rolleyes:
Hahaha at least it looks cool - right? I love big switches :p

Funny. You actually got me pretty worried about this, did a bunch of research, and (not that I questioned you ;) ) found that oxide buildup is a serious consideration in signal type circuits with relays.

This thing will be shipped in 3 weeks, then used for 9 days over a month long period - you think I'm safe? I can't find much info on minimum load on switch side.

I really wish I could find a plug in type relay that could handle this. There is a 12v (I already have the PSU - might as well use it haha) minimum 1 A type on digikey that is plug in with silver tin contacts. Not including harness though. https://www.digikey.ca/product-detail/en/Omron-Electronics-Inc-EMC-Div/G8HE-1C7T-R-DC12/Z5645-ND/2755412 Think it's worth the switch (pun) or will what I have be cool for a short period?
 

Tonyr1084

Joined Sep 24, 2015
4,737
Here's a simple mod to the drawing I posted earlier. I think this may be more in line with the application as you described it to me:

The power source - any power source, be it battery, wall wart, solar panel, - whatever - with a relay rated for the same voltage can do the switching you ask. No need for an automotive relay as they can manage 40 amps. Likely your circuit isn't drawing - well, I guess at this so don't take this literally, but if all it is used for is a signal switch, it may draw a mere few milliamps. Maybe as much as 500 mA, but I don't know. This circuit - if powered by a battery, assuming the tape switch may actually spend only a few hours a day actually closed, and a small relay that doesn't draw a lot of current, then a battery appropriately sized will probably do the job. That part of the engineering is going to be up to you to decide how to make this work. Simply put, step on the pad, switch ground from ( IN 1 ) to ( IN 2 ). Step off of the pad and it switches back to ( IN 1 ).

[edit]
If you don't want to use a common ground then connect the ground from the common on my drawing to the ground on the device and don't ground it to any kind of actual ground. Again, I'm not 100% familiar with what you're doing, the engineers cap is on your head.

Hope this helps.

Floor Switch.png
 

Thread Starter

Thenextman

Joined Feb 3, 2017
42
Hey all,

So this was launched, and in the field for 4 days before starting to act up on day 5 and 6. The client eventually had to bypass the video switcher. Quite sad for me hah.

Kramer, the people who make the video switcher, say that the problems might stem from the contact closure pins being constantly looped together, as opposed to momentarily.

The installation is across the country at the moment, so I cannot investigate in person. The ideal scenario is the send them a simple part that goes between the current SPDT relay and the video switcher that can make the contact between terminals momentary. No voltage can be fed to the switcher, however the component itself can be powered.

Any thoughts!?
 

Thread Starter

Thenextman

Joined Feb 3, 2017
42
Hey again,

I realize my post may have not been clear. Essentially, we have a component that is controlled by making a momentary connection between one of two terminals and ground. The terminals are labelled 1, G, and 2. If you take a piece of wire and connect 1 to G, "1" is selected on the device. If you then take that wire and connect 2 to G, 2 is selected.

The goal was to make a foot switch (a mat that acts as a simple push button - i.e. standing on it, connection is made, get off it, connection is broken) control the switching between "1" and "2". When no one was standing on the mat, "1" was selected, when someone stood on the mat, "2" was selected. @tcmtech brilliantly recommended a SPDT relay to accomplish this.

The layout was simple, Normally Closed and Comm on the relay were hooked up to "1" and G on the component, and Normally Open on the relay was connected to "2" on the component. When you stood on the mat, it would complete the connection between a 12V PSU and the coil in the relay, effectively selecting "2" on the component by creating a closed loop between G and 2. When they got off, the coil lost power, switching back to "1".

The system worked well when we tested it, however, after a few days on the road, it started to malfunction. When people would stand on the mat for a long time, it would make the input "Time Out" as they put it.

The manufacturer of the component says that typically, you just want to create a momentary connection to initiate the switch. They think that because it is being held, it is causing an issue. (Annoying, because I described this setup to them in detail and they said it would work well)

So, now the unit is across the country, and we need to make it work again. The ideal scenario is to provide something they connect between the component and the SPDT relay to translate the constant connection to a momentary connection. No voltage can be fed to the component, however, the addition to the circuit can be powered by a transformer (available 120vac power in the installation).

Please see diagram for more info.

Thanks so much all!
 

Attachments

Tonyr1084

Joined Sep 24, 2015
4,737
After reading your description, I come away with "there's a problem if someone holds the switch for an extended period of time" (i.e. standing on the mat for hours {per se}). What I'm not sure of is how they want to switch this device. I'm wondering if they want a momentary G1 pulse to set it to G1, then a momentary pulse to set it to G2. Apparently they don't like having either G1 or G2 constantly held. Let us know if this is an accurate understanding of the new goal. I'm CERTAIN someone here can find a solution that you can simply build and send to them.
 

Thread Starter

Thenextman

Joined Feb 3, 2017
42
After reading your description, I come away with "there's a problem if someone holds the switch for an extended period of time" (i.e. standing on the mat for hours {per se}). What I'm not sure of is how they want to switch this device. I'm wondering if they want a momentary G1 pulse to set it to G1, then a momentary pulse to set it to G2. Apparently they don't like having either G1 or G2 constantly held. Let us know if this is an accurate understanding of the new goal. I'm CERTAIN someone here can find a solution that you can simply build and send to them.
Hi Tony,

A momentary connection between 1 and G or 2 and G is what will trigger the component to select 1 or 2. We want 1 to always be selected, unless someone is holding down a button, then 2 should be selected. When they release the button, 1 needs to be selected again.

Our current setup makes constant connection between 1 and G when no one is holding down the button, and a constant connection between 2 and G when someone is holding down the button.

The goal is to provide a component that can be installed by non-technical staff to make a constant connection between two terminals translate to a momentary connection (no pulse, no current, just a momentary connection as with a push button).

Hope there's a simple fix! I'm leaning towards Arduino at this point hah.

Thanks!
 

Tonyr1084

Joined Sep 24, 2015
4,737
Well, as usual, the peanut gallery has come up with SOMETHING:

The drawing below [EDIT - Drawing REMOVED. See next post for corrected drawing] shows the power source and the floor switch. When someone steps onto the floor switch it triggers a OneShot that is timed to put out a pulse long enough for the unit under control (the UUC) can switch to input 1 (or 2 or vice versa). As long as someone stays on the pad the unit remains in mode 1. When someone walks off the pad the second OneShot sends a pulse, also long enough to switch the UUC to the other state (whichever you choose).

There IS the possibility that someone steps on and off so fast that the OneShot doesn't have enough time to reset, but it might not be an issue for the UUC. Especially if the second OneShot has a slightly longer duty cycle. My drawing shows 800 mS duty cycle, but the UUC may need an even shorter pulse. So the first pulse can be whatever the UUC will tolerate well and the second pulse can be 200 mS longer (or anywhere longer than the first pulse).

I used two Opto Isolators in my design simply because they can give you the connection you want between 1G and 2G. Also notice in the drawing that the Opto's are wired "Emitter to ground". Since the "G" terminal may represent ground, you want to strike 1 or 2 to ground to activate that function. IF this is wrong, having two wires go to the G makes it possible to swap each connection as may be needed in order to satisfy the directional sensitivity the UUC may have. Also, if you can get specs on the UUC determine what current is going between 1 & G or 2 & G. If the current is too high you'll want to put a resistor in the circuit. But since I don't know anything about the UUC I can't make any recommendations.

As for the OneShots - I drew two 555 timers. Possible you can use a single 556 (double 555) timer to reduce chip count. And the Buffer/Inverter can be whatever. As long as it's an inverting buffer of the correct voltage. You may even opt to build an inverter out of a transistor instead of putting a six pack of buffer/inverters on the board with 5 of them just laying there doing nothing. Whatever you choose, I think this may be one remedy to your issue, now that I understand it better.
 
Last edited:

Tonyr1084

Joined Sep 24, 2015
4,737
Just noticed, the Buffer/Inverter input should be tied to ground through a resistor to keep it from false triggering or oscillation.

[edit] Re-read your last post. Looks like I got 1 & 2 reversed. Simple enough to rewire at the connector. U2 should be connected to 1G, not 2G as I initially drew.

[edit #2]

New CORRECTED drawing:

Oneshot Pad Switch.png
 
Last edited:

Thread Starter

Thenextman

Joined Feb 3, 2017
42
Just noticed, the Buffer/Inverter input should be tied to ground through a resistor to keep it from false triggering or oscillation.

[edit] Re-read your last post. Looks like I got 1 & 2 reversed. Simple enough to rewire at the connector. U2 should be connected to 1G, not 2G as I initially drew.

[edit #2]

New CORRECTED drawing:

View attachment 126523
Looks right! Hopefully, it does not come to this hah.
 

Tonyr1084

Joined Sep 24, 2015
4,737
Why not? A 556 double timer is a single chip. The two Opto-Isolators - each a chip in itself, another chip (inverter/buffer) brings this to a four chip count. OR you can (as said before) leave out the buffer in favor of a single inverting transistor arrangement. Some caps and some resistors and maybe two 10 turn trimmer pots to tune each OneShot so that the one that restores to the original config, has a slightly longer high output.

Going with Arduino means creating a program as well as the Arduino. You're still going to need some kind of switch, be it a relay or an Opto. Component count may be slightly lower but cost will be much higher. Then there's the issue of programming. And what happens if the Ard' loses its program or the program becomes corrupt? What if something happens and you blow out the circuit? Replacing the Ard' will be more expensive than replacing a 555 (or 556).

For all my lack of skill and knowledge of engineering such a thing I still think the chips are the better way to go. Put them in sockets so they can be replaced without need for de-soldering anything and I think you're good to go.

If the system doesn't switch from 1G to 2G then reverse the two wires on 2 & G. If it doesn't switch back then switch the two wires on 1 & G. OR if you want to have to not care which wires go where then instead of using the Opto-Isolator go with an Opto-Triac. That way it won't care which direction current is going, or even if it's an AC signal, OR what frequency it may be. As long as the triac can switch fast enough to meet the AC signal. All "IF's". But for the life of me I can't see any reason why this wouldn't work. All you will have to do is design your OneShots - and that part is easy enough. It will depend on your supply voltages and voltage and current capabilities of each component.
 
Last edited:

Thread Starter

Thenextman

Joined Feb 3, 2017
42
Here's a transistor inverter - very simple.

View attachment 126554
Haha. Funny how the simplest thing is so beyond me. I guess the advantage of the arduino is that I get an UNO with a project box, put three terminals on each side and tell them "put this between the relay and the component". Simple for them to install.

The component manufacturer is now saying that it should work with constant / latching switch so thinking we just replace the current relay and component.

Will see next week.

Thanks Tony!
 

Thread Starter

Thenextman

Joined Feb 3, 2017
42
Hello again!

So, the manufacturer of the component has said that latching bs momentary contact shouldn't be a problem.

I am starting to think it may be my relay causing the issues.

I did not use a diode with the SPDT relay. @tcmtech and @Tonyr1084 - what type of diode would I need and where would you put it?!

Do you think this would cause a malfunction over such a short time?

Thanks!
 

Tonyr1084

Joined Sep 24, 2015
4,737
In posts #42 & #47 I didi not draw a flyback diode. Probably should have, as the flyback from the coil can eventually burn the switch up. Sorry, my fault if the tape switch has gone bad.

A flyback is advisable especially when you have sensitive electronics to protect. They're always put in a way to oppose the current from the source. I've included two of your drawings showing the diode opposing the flow from positive to ground. But also looking at your drawings, if you are using batteries, the coil itself will continue to draw current as long as someone is standing on the pad. The batteries will go dead rather quickly. If you're using a line powered supply then you don't have to worry about draining a battery.

Here are your drawings with my modifications. I believe this will answer your question. I'm just hoping the tape switch isn't the problem. As far as I can see there shouldn't be any problems at all, other than the tape switch. A flyback diode will protect the tape switch.

Your drawings sir:
 

Attachments

Top