2NO 2NC SSR

[RyanG]

I have been looking for something that could serve the purpose of a 2 normally closed / 2 normally open solid state relay for a while with little luck. I want to make one. The device I will be using has 2 separate power sources for the operation. One of them is a constant 12vdc and the signalling voltage will come from a 24vac source. Could I use the 12v as my power on my optotriacs' leds with the 24vac, through rectification and stepping down turning on or off transistors? Or would there be a better way? There would be 4 optos and 4 regular triacs with respective snubbers. I have drawn it up with a zener dropping the 24v after rectification to 12v and then feeding that into the base of some transistors with pull down resistors. I posted my original schematic of this a while ago, but there was little response or help. I know SSRs are better left to the companies that make them, but they don't really have what I am looking for.

[RyanG]

I drew up the schematic.

[Zero999]

It looks like it'll work but there are easier ways to do this, such as a CD4069.

R3 to R6 aren't needed. Q1 and Q2 are emitter followers, so the base current will already be limited by negative feedback. R9 & R10 already limit the base current to Q3 and Q4 and their base-emitter junctions will become reverse biased, when the relay is activated and they switch off.

[RyanG]

Will the 24vac need to have a smoothing cap? If I am going to remove the R3 - R6 is there a better value resistor for the pull downs? Just use 10k or something else instead? If I were to use a 4069 would that just replace the transistors and their resistors and feed it the +12vdc as Vdd and Vss is ground?

[RyanG]

Would the 4069 work like this?

[Paul Moir]

I think a 4069 is not a good idea.  It can only source or sink something like 1.5mA while the MOC3041 says it needs 15mA to guarantee a latched output.  A 4049 might be just able to do it with no current limit resistors but it's nothing I would want to rely on.
(FWIW, maybe with a different optoisolator and using 1 gate output per LED but this doesn't seem easier somehow.)

[Zero999]

I think a 4069 is not a good idea.  It can only source or sink something like 1.5mA while the MOC3041 says it needs 15mA to guarantee a latched output.  A 4049 might be just able to do it with no current limit resistors but it's nothing I would want to rely on.
(FWIW, maybe with a different optoisolator and using 1 gate output per LED but this doesn't seem easier somehow.)

You're right. The CD4009 should do.

Would the 4069 work like this?

Use a separate gate for each MOC3041.

[RyanG]

What is the max current source on the 4009? For the The number of pins would work out perfectly for what I need with 4 pins for the NC and 8 pins for the NO. Would the way I drew the circuit up with the 4069 work except with each opto on its own gate? And what resistors would be required?

[Zero999]

What is the max current source on the 4009? For the The number of pins would work out perfectly for what I need with 4 pins for the NC and 8 pins for the NO. Would the way I drew the circuit up with the 4069 work except with each opto on its own gate? And what resistors would be required?

Please read the datasheet and apply Ohm's law, then get back to us if you have any further questions.

You should also consider using the, MOC3043, rather than the MOC3041 because it has a lower trigger current and is the preferred device.

[RyanG]

I understand how to use Ohm's Law but I do not understand the datasheet completely. I've not really used an inverter before. I'm assuming the outputs use the VDD so I would use resistors according to +12V, and I am sure that if I use the MOC3043 that the 5ma required would be easily doable. A 2.2kohm resistor would bring it to around that 5ma. I just need a little help understanding this datasheet completely in the sink/source area based on my +12V for VDD and VCC.

[BrianHG]

When using any ICs, when driving OPTOs, it is better to tie the opto's anode to V12 & have you IC drive the cathode.  The logic IC sinks current better than it sources.  This means your NO will become NC and vice-versa...

Also, for better sink current, if you can change the 12v for 5v, I would use the 74HC14 as it will drive much more current into the OPTO Leds, & for the first input, I would diode protect it by feeding your source through a series 10k resistor + a BAT54S series pair schottky diodes.

[RyanG]

Should I just replace the 12v zener with a 5v and use the 74HC14? Should I use a 7805 on the +12vdc?

[RyanG]

Changed to include 74HC14. Is it correct? Also, would a 74HC04 do just as well? I have some of those around but no 74HC14.

[BrianHG]

Like this.  Note that the 12v always needs to be there unlike the control input which swaps your to relay outputs.  Remember, with the NC SSR, it needs power to it's optocoupler to be enabled.  That 12v can go anywhere from 7v to the highest voltage the 7805 will allow.

Changing the 2 10k resistors at the input will alter the input voltage sensitivity.  If you are feeding digital logic into the input, you don't need the resistors or BAT54S.

[RyanG]

So, this is the complete thing that I have. What modifications should I make to it? I didn't know how to incorporate the changes you wrote in the last one into this exactly.

[BrianHG]

D5 being 1 watt and R2 being 750 ohm is kind of overkill for a 74HC14 which has almost no input current.

I would use a 1/2watt or 1/4watt ZD for D5.  Maybe switch the voltage to 4.9v or 4.7v, just to be safe.
Change R2 to 10k and add another 10k in parallel with D5.  Otherwise, your SSR will take forever to switch off when you remove the 24v AC.
You may need to experiment with the value of C1.

I think C8 should be at least 10uf so that the 7805's output wont oscillate.
Also, you may use 78L05 if you wish since this circuit wont go anywhere near it's 100ma limit.
Everything else looks functional.

[RyanG]

Alright, I changed the 750 to a 10k and added the additional 10k. I did change the 7805 to a 78L05. I also changed C8 to 10uf. Now I have a few questions about the high voltage side. First, what wattage should the resistors be? Are the caps at 10nf and 1kv enough? And, the neutral of the 120v has to be on pin 1 of the triacs?

Also attached is the most up to date schematic. I still have to add screw terminals for the mains to the triacs.

[BrianHG]

Before we go any further, I assume you 24vac & 12vdc come from isolated sources, I don't want the 5v regulator to recieve a nasty over-voltage surge if the 24vac pushes your GND down by another 24v or more.  If you want to be super-safe, we should replace your input stage, diodes and cap, with an AC optocoupler.  This will guarantee safe functionality of your circuit under any random circumstances.  (It will also make me feel better too since I don't know your complete setup circumstances).  For your 24vac input, use the FOD814, see the data sheet: https://www.fairchildsemi.com/datasheets/FO/FOD814A.pdf
Just feed the 24vac at the input side through a series resistor.  No need for any diodes there.  On the output side, just use a pullup resistor  & 1uf cap feeding your first input on the 74HC14.

First let's clean up the everything to the left of the schematic.  All that tracing around the HC14 is giving me a headache following it...

For the IC, in your cad, you can mirror it so that the inputs are on the left and the outputs on the right, or, rotate it 180 degrees.  This will unscramble all your wires going to the optocouplers and you will should catch a wiring mistake you made when fixing this.  Also, for VCC5 and GND, don't be afraid to use labels instead of a wire trace going all around everywhere just to maintain that wired look.

1.  The resistors on the left should be at least 1/8th watt.
2.  Except if you want to keep the LED powered from the 24v, make sure R1 is 1/4 watt.
3.  The 10uf and 1uf are usually polarized.

On the HV side:
Take a look at my attached screenshot...

[RyanG]

Alright, thanks for the continued help on this Brian. Sorry about the traces before. I knew that I could rotate the HC14 but didn't really think about it until I had most of the traces on. Should the 1uf cap be on the collector or emitter of the optocoupler? I was thinking the emitter side because of the AC signal, but I put it on the collector side because it was easier in the schematic. I have a bunch of 1/4 watt resistors of a wide range of values so I'll just use those across the low voltage side. For the high voltage side, the picture you posted said to use higher value resistors on the snubbers for inductive loads. This is intended to run AC motors, so should I increase the value of the 47 ohm resistors? Also, what wattage value resistors need to be used on the high voltage side? Looking at some of the other solid state stuff we have around at work, it seems they don't use anything higher than 1/4 watt or maybe 1/2 watt on the triacs/snubbers.

[Zero999]

There should be a pull-down resistor on the emitter of the opto-coupler. The collector can go directly to VCC.

LED1 will require some form of reverse polarity protection, as the maximum rated reverse voltage is typically 5V.

Probably not what you're looking for but for the normally closed contacts on the SSR, you could use an ordinary opto-coupler with a bridge rectifier and BJT to drive the TRIAC. It won't have any zero crossing but it saves having an auxiliary power supply. The normally open contacts can just use the usual MOC3043 circuit.

[BrianHG]

Here you go, I fixed the 2 tiny errors.
As for the SSR-HV side, if you are using a high inductive load, switch the snubber resistor as describes in the datasheet.
1/4 watt is ok.

If you want an led right on the 24vac input, place 2 in parallel with opposite polarity through a series resistor right there.  Do not connect the other side of these LEDs to your GND, this circuit was designed to totally isolate yourinput 24vac from your logic 5v circuit from all other power sources making it safe to work with any combination of power sources which you might not know where they are coming from or how they are earthed.  IE, the 24vac can come from a furnace thermometer control box, or a door-bell button power supply while the 12v/5v can be from a PC power supply while the devices on the TRIACS can be connected to the 120v AC mains without any interference or damage.

[RyanG]

Having the LED there is fine. It is just an indicator of the relay being activated. The 24vac and the 12v are already coming in from the machine I want to place this in, which is why I selected those voltages to work with. I'll up the resistors on the snubbers, and then I think everything is good. Correct?

[BrianHG]

Everything looks OK with those 2 final changes.
Would anyone else out there following this thread want to comment?

[RyanG]

Now, I'm trying to look at all the machine information I have to figure out which wires will have to go where on the board. I color coded the wires on the schematic to make it easier to see them. Sorry about the not so good picture quality. It's an original schematic from about 1972.

[RyanG]

I did some figuring out, and I think I got the correct setup for the triacs. The triacs' A2s would be NO7, NO9, NC10, NC3. The A1s would be NO2, NO4, NC5, NC8.(Those are from my 1st picture in my last post) Figured I would show you guys what I am trying to do here if you are interested.

[RyanG]

So, I finally got the parts to put this all together. It doesn't work. I'm thinking it has to do with the FOD814. I am using the last schematic that BrianHG posted. I think that the C and E of the 5V side are backwards. Is that all that it could be? The NC and NO sides work with just the 12V line connected, but when I connect the 24VAC it doesn't light the LED or swap the closed to open and vice versa.

[BrianHG]

Yup, the C&E are backwards.  If the 74HC14 is powered and wired correctly, applying 24v should turn it on.
Also check pin1 of the 74HC14, it should be high/5v, then, when you apply 24v ac, or even dc at the input, it should go to GND/0v.

[RyanG]

Thank you. I tried swapping it earlier and it did work. 

[Zero999]

Thank you. I tried swapping it earlier and it did work. 

Good, I'm glad it works but is it safe?

Have you left sufficient creepage and clearances between the mains and 24V sides?

Is it build on a breadboard, stripboard or proper PCB? What's the minimum distance between anything on the mains side of the opto-coupler and 24V side?

[RyanG]

It's on a PCB with some stuff jumped around for the optocoupler. My creepage and clearance should be good. I used one of the calculators on a website.