Triac versus Relay output card

Just wanted to know what are people's thought on a triac output versus relay output cards? I'm working with the Allen Bradley SLC line, 1747-OA16 and 1747-OW16 output cards. I've used the OA16 before to control a solenoid manifold which operates high pressure valves. I snub each coil of course, for arc-quenching. I need to build another PLC controlled solenoid rack and it has worked in the past (still works today after years of service) using a 120V TRIAC output card, but it doesn't mean it is right, or the best way to do it.

I'm just wondering when it is better to use a triac or relay output card. What are the pros and cons? Does load current determine one or the other?

Relays have a finite life. Triacs, and other solid state outputs, have infinite life when done right as you seem to have experienced.

Relays generally handle more load.

Triacs "leak". Triacs, and most other solid state outputs leak enough current to fool a high impedance voltmeter in an open circuit condition when in the off state. Also, the off state leakage may cause very low power field devices to get "stuck" on. Other symptoms may include flickering, humming, or lamps that glow dimly.

Relays are "dry contacts" so they allow isolation from otherwise independent voltages and circuits. When they are off, they are all the way off. They don't "leak".

That is the "nutshell" version in random order.

That's a pretty good "nutshell" explanation. I like it. Thanks!

Every panel I build uses solid state outputs. The only two things I drive with them are slimline interface relays, or 22mm indicators on the panel door. I never drive field items or contactors straight from the outputs.

This way I get the more or less "infinite" life from my outputs that OkiePC suggests above, and then the relays, well, it's a whole lot cheaper and easier to replace a relay now and again than an output card.

The only two things I drive with them are slimline interface relays, or 22mm indicators on the panel door.

Click to expand...

I kind of like the idea of slimline relays. I just looked them up. Altech Corp were the first kind that popped up doing a G-search. Omega makes the same kind too. I've never used them before. I've had success with just using the -OA16 output card, each output fused, directly wired to the solenoid I needed to actuate. They are pretty small solenoids; probably the size of a standard SPDT automotive relay. I was always hesitant of using more relays to a setup that already works, adding more components that can fail and taking up a lot more real estate. Even with adding small ice-cube style relays the amount of room that 7 or 8 solenoid bank control would be significant. But you've definitely got me thinking... do these slimline relays have any plug in arch-suppressors you can add to them? Or, are they even necessary? The solid state output from the card is still essentially powering a coil within the slimline relay (even if it is a very tiny coil) and still susceptible to voltage spikes, right?

I attached a pdf file of the valves I control to get an idea with what I'm talking about.

It just doesn't say the current draw or inrush of the coil. I may have to do some in-use current tests. hmmm...

All solid state outputs here, mostly 24VDC, but still a few legacy 120VAC ones, and unless the load exceeds the output capacity, we drive field devices directly.
In 10 years, I can count the actual number of failed outputs on one hand, with fingers left over. In fact, every one was on an older system, and was an SLC-500 AC output driving an oversized contactor.

I do go ahead and use these nice isolation relays but only if I am driving a load the output can't handle, need voltage translation (newer designs stick with all 24VDC outputs, no 120 on the PLC rack), or need actual isolation (contact closure to energize a drive for example).

DC does not leak, however AC does, so if loads are small <20 mA you will need a isolationrelay.
On DC a simple diode is good as snubber, in AC a resistor and capacitor or a VOR is needed, these overvoltage protection has or a limited life or makes leakage current.