Triac & Solid State Relays

CPUMAN

Ok I have multiple questions. So here goes...

1.· Triac vs SSR? Why use one over the other? Pros & Cons?

2.· Can SSR's take the abuse of frequent switching? Says rates of once per second or maybe even once per half a second? What about Triac's?

3.· How would I interface either device to a stamp?
I've seem some circuits with Triac's that use a...
Output ---> Optical Isolator ---> Transistor ---> Triac
connection scheme.· Would this be the best way to connect a triac?

I've heard the SSR's are already optically isolated so·SSR's can be interfaced directly to the stamp, is this true?

... and so ends the 10 million questions... ok not quite but there's plenty [noparse]:)[/noparse] If your wonder what I'm building, it's a light controller that will control 110V AC lights that are up to 100 Watts (though most likely will be 50 Watts or less). Think of a string of chasing xmas lights to get an idea of what I'm doing.

Any help is greatly appreciated...

Chris

dbpage

Here's an attempt to answer your questions.
1. With an SSR you get the package deal, and with a triac, you get the do-it-yourself deal. Get the one that matches your requirements.
2. SSRs can be switched fast, but latency becomes an larger issue as the switching speed increases. I would guess that latency is not a problem in your application.
3. The best way to connect a triac depends on your requirements. I recommend optoisolation when loads exceed the STAMP Vcc voltage.
4. Different SSRs have different input impedances. Some SSRs can be interfaced directly to the STAMP, and the number of SSRs that a STAMP can drive depends on the specs. If the specs are not known, use a transistor to drive the SSR.
For your application, I would recommend optical isolation. It's up to you to make the buy vs build decision.

dbpage

Clarification:
I was so focused on optical isolation that I may have missed an important point. The output of an SSR is NOT necessarily triac, it may be a triac, SCR or transistor. The output may be suitable for resistive, inductive, capacitive, AC or DC loads, depending on the SSR specs. Ya gotta know the specs!

CPUMAN

Ah ok that makes sense. I've already picked up some 110V AC with a 3-32V DC control SSRs.· The main reason I was asking about how difficult the triacs would be is because these SSRs weren't all that cheap, though I’m going to look on eBay for some.· My lighting controller will have 16 channels, so I need to keep individual component costs down and preferably keep the amount of work to build each channel controller to a minimum as well.
·
You said the SSRs could be switched fast but latency would become an issue at higher speeds.· What type of speeds would·the latency start becoming an issues?· As I have no plans to exceed 10Hz considering for most light·controllers that’s·fairly rapid.
·
Thanks,
Chris

dbpage

You might notice latency when switching at speeds less than one-half cycle at power line frequency. Your SSR should turn on immediately, and turn off at the next zero crossing. Your SSR latency should be shorter than the filament heat/cool cycle of a light bulb.

tractionfan

Here's a related question I was going to ask before I ran into this thread: I want to control 110VAC lamp dimming by firing a Triac at varying points in the AC half cycle. I was wondering about the SSR's ability to work at that speed. I also would need a way to detect the zero-crossing point so I could "count" an approriate delay before firing... all within the half-cycle of the 60hz AC line.

But... maybe I don't really need to do that considering the long lag reaction time of filament bulbs. Rather I could just provide some on-off output with varying degrees of on and off but with a cycle much greater than the 120 hz of 1/2 cycle AC.

Any thought from anyone?