I have some old pieces of equipment that has rather unusual (to me at least) control signals.
From what I have measured, it uses a 3 wire cable, with 3 different voltages that look kind of like they are using a token ring configuration to operate.

The signals on the wire look to be straight AC, by memory measuring 110 VAC, 50VAC and 0VAC. (I will have to check my notes at work to verify the voltages)
One of the pieces of equipment has failed, is VERY obsolete and no repair parts are available.
I am trying to figuring out how best to grab the signals from the incoming lines and feed them to a new PLC in a form it can recognize so I can rebuild the machine.
I had a thought of just using a voltage divider circuit and feeding the result into an analog input card, but I thought there must be a better way than that.
What would a less odd solution look like?

 

Do the signals need to rectified to DC for the PLC?
Optocouplers would seem like a good choice to isolate and reconfigure the signals as required.

 

Welcome to AAC.

Can you describe how the signals are used?
In what way are they similar to Token Ring?

As @sghioto mentions, optoisolators are one way that seems to be good for you (in the absence of enough information about the actual operation), Alternatively, if they are AC signals, transformers would also provide galvanic isolation and could be adjusted to match logical high for the PLC.

 

Some clarifications on how the machine signals work.

The signals travel on a 3 wire cable to input on the counting machine.
The machine sends signals out on a 3 wire cable.
All voltages found on the cables are AC. My readings where 120 VAC, 50VAC and 0VAC, or close to them - I'll check my notes when I'm back in the office next year.
The three voltage levels for signals come on the same wire, so three states to look at with 1 wire. I needed a way to separate the logic states off of the 1 signal wire into 3 readable input states. Of the 3 input wires, none are a neutral. The circuit appears to relies on the second cables output wires to act as the neutral side for the input AC signals.
The token ring similarity is that when they have finished counting and are ready to dump their items, they change the value of one of the signals from 120 to 50volt, but only if they are the first machine. Second machine still passes 120volt until it is ready, then if it is receiving 50 volt, it passes at 50 volt. This continues down the chain of machines until it reaches the master controller that sends another signal to advance. The exact logic I haven't gotten yet as I hadn't invested time into until I could be sure I had a method of capturing the signal logic to a PLC. The old controls are all transistors and relays and custom IC's.

 

The very first, and very important, question is what sort of PLC INPUT was the signal connected to??
A "counting Machine" is not much of a description at all. The description of the function is that it is a work-around for using more wires to deliver a digital state signal, and that it is not by any means representing a linearly changing condition. It may even have been from a "CYPAK" logic system, which became obsolete in about 1960.

I do not recall ever seeing a PLC analog input that would accept a large AC signal. Was the signal from an LVDT, (Linear Variable Differential Transformer) by any chance??

But given that the output appears to be one of three voltages, it becomes very obvious that whole lot more information will be needed to allow any useful advice to be provided.

 

If there is more than one "machine" in a system and that is reporting a status to some other system, then certainly there is a more modern way to handle the communications. No question about that. But we have no information at all about any of the machines or systems involved and that really does make offering any valid advice quite difficult.

 

In addition, there must be some portion of the machine that senses the condition to trigger the signal.
It seems to me now that it may be a good time to change the whole scheme to some sort of digital logic that is presently available, such as a simple PLC that can see all of the different condition signals and provide a correct response output for each one. And if a portion fails, be recent enough that a replacement part is available for purchase immediately.
So a question for Maintenance Steve is: Do you understand relay ladder logic??
If you do, then a simple PLC will be your solution.