RS-485 is a communications standard, so the idea is that any product which conforms to that standard can, in theory, communicate with other products that also conform to the same standard. Same with CAN (or Ethernet, or RS232, or USB, or whatever).
If you're testing for continuity and leakage, then much of what's contained within those standards won't concern you. All you're looking for by the sound of it is a way to check the integrity of the wiring used to connect equipment together, not necessarily the details of how the equipment actually communicates. Am I right?
An RS-485 chip will be one which specifically implements some or all of the circuitry needed to talk with other equipment according to the RS-485 standard. Take a look, for example, at these:
http://www.exar.com/connectivity/transceiver/rs485These chips contain line drivers and receivers, which physically connect to the wires and drive electrical signals onto them. The actual data rate on the wire, and the specific pattern of 1's and 0's used, is determined by the microprocessor or other controller that they're used with.
Physically, there will usually be a pair of wires used to transmit data, and one pair used to receive. A logic '1' is normally represented by one of the wires in each pair being at +5V and the other at GND, and a logic '0' being the other way round. The transmit pair of one device is connected to the receive pair of the other. If the product you're testing has a 3.3V interface then the logic levels on the wire might be +3.3V and GND instead, which is probably OK in many cases.
I believe it's also possible to have multiple receivers driven from a single transmitter.
With CAN, there's only a single pair, which is terminated with a resistor at both ends. Normally the two wires sit at about the same voltage (around 2V or thereabouts, though the spec allows this to vary quite a lot). This is called the "recessive" state. When a device wants to transmit, it drives the CAN+ wire to a higher voltage and the CAN- wire to a lower voltage; this is called the "dominant" state. Data is transmitted by rapidly switching from one state to the other, and every device on the bus will receive the same information. Each message has includes a header which states which device the message is intended for, and every device on the bus must therefore have a unique address.
Does that help at all? Hope so!