Industrial Strength 8 channel SSR’s

[ttr502]

After an automations equipment upgrade where I work, I took one of the PLC digital output cards apart to see the circuit inside. They appeared to have the same basic circuit as the SSR’s I have seen here and on other sites, but are more “robust” of course being for an industrial application with status led’s, spike protection, and very underrated in regards to current. I rescued these from the landfill and after some surgery and testing, I was able to get the cards functioning quite nicely behind a Grinch controller. I also rescued the backplanes that these cards plug into that have screw terminals for connections. I plan on adding an RJ45 into the backplane as well to pass the signals from the grinch. This will allow me to quickly swap a bad module when needed. I had to remove the PLC addressing circuit and replace it with wire jumpers to the opto’s. I have accomplished all of these items and I am ready to modify more of the circuits for use this season.

Now this is my question: The way this circuit is laid out it only has one resistor on the 5v for the eight opto’s, around 400 ohms. All of the other SSR’s I have seen have a resistor for each opto. My electronics is a bit rusty, but isn’t these resistors just for limiting the current through the opto? I have repaired one output on one card that I smoked during testing by replacing the original opto with one of the MOC3023M opto’s I had bought for building normal SSR’s with no problems. I do not have any schematics or data sheets or information on the original opto’s or other components to work from. I have attached two pictures for reference. Is the one resistor configuration OK?

[ukewarrior]

Can you say what units you modified so I can look for the same in our scrap?

[ttr502]

PM sent with links for more information, but the cards were SIEMENS 6ES5451-8MD11. We use SIEMENS automation almost exclusively, so I could not say if other brands are similar and might work such as Allen-Bradley, ABB, GE, etc.

[awhaley]

Well, your SSRs look like a great find!

As to your question about resistors... Well first off maybe me bumping the thread here will help you get one of the experts to reply, but in the meantime, my understanding is....

The input pins of the opto are ACTUALLY the input pins of a tiny LED inside the opto, and so the normal rules about wiring up LEDs should apply to the opto as well. And the normal rule for wiring up LEDS is you SHOULD use one resistor per LED. A google search returned a dozen people telling me this is true, and very few telling me why. I've come up with a couple of answers myself, and I hope someone here can tell you if either of them are valid (lol.)

1, if there are manufacturing differences in the leds, one with lower internal resistance will allow the current to pass, so the others wouldn't light, because electricity takes the path of least resistance...

and

2. Voltage= Amps X Ohms.... that's the equation the cute online calculator uses to tell you what resistor to run... you have 8 volts, you want 3 volts at 20ma for your LED.... so you've got to drop 5 volts and limit current to 20ma.... and the math tells you that requires 250 ohms.... now you want to turn on TWO leds? That ideally would draw 40ma... which means instead of dividing 5/.02 I divide by .04 and get 125 ohms....

So if you trigger one opto on that circuit with 8 volts and the 125 ohm resistor, it gets 40ma supplied, and if you trigger the second opto it drops to 20, and if you trigger a third, you get 14, and so on and so forth. Is that correct, oh electronics gurus that i know are lurking out there?

If I AM right... then I assume your board has simply been designed with a wide enough tolerance to handle the changes in current. Perhaps the optos trigger reliably at 5ma, but can sustain up to 20, so you design the system to provide 20ma to 1 or 5ma to 4? Now, the single resistor would REALLY scare me when an opto was getting replaced with a different model like you mentioned, because if it has less internal resistance, it could take all the current and make your others refuse to work... or overload itself.

If it were me, I'd leave the boards the way they are, as long as they work, rather than redesign a known good system, but I'm going to keep using individual resistors on my parallel optos and leds when I design circuits, and I hope someday someone tells me the real reason why I'm doing that? Who knows, it still COULD be a myth perpetrated by the resistor factory to get you to design with more resistors!

Merry Christmas!

Art

[ttr502]

Art, all of your reasons and calculations sound good to me. I do not have the original schematics for the cards to work from, so a lot of things I have done have been trial and error. I have worked on the automations where these cards had been used since the mid 90's and have a good understanding of their original design capabilities. I did have to remove a portion of the original internal circuit to get these to work as SSR's. If you look close to the pictures, some cat5 wires are soldered in to replace what I believe to be the plc "addressing" circuit that was mounted at 90 degrees to the main board. The circuit I removed had one large IC and many other surface mount resistors, capacitors, and transistors, but it is hard for me to derive a schematic just by looking. It is possible that this circuit does have a resistor for each opto since 8 of the 18 leads on the circuit went directly to pin 2 of the optos. The others come from the edge of the card that plugs into the backplane where I also removed a similar looking circuit to add RJ45's that feed the card socket. By using the backplane, I can change the whole 8 channel SSR in seconds, plug and play so to speak. I think I will do some more experimenting with a card and use 8 resistors instead of the cat5 wires to see how that works. On one of the cards using 1 resistor, I ran a "burn-in" test where I had a looping sequence with on times varying from 25ms to 1s, and also varying the number of and combinations of different channels running for 8 hours. The only thing to get hot was the eight 100 count light strings! I also have ran four 100 count light strings on one channel continously for four hours, with the same results.