5 replies to this topic

[Cabadam]

I was looking at solid state relays, such as: http://www.sparkfun.com/products/10636
They have an example layout that includes a number of additional components including 3 other resistors and a transistor. (I'm excluding the LED and its resistor as that is understood and useful).

Take that, and then compare to a diagram such as: http://www.flickr.co...57603835280870/

It is based on Arduino, and basically just needs 1 resistor on the signal line.

Take then a TRIAC such as: http://search.digike...L401E3-ND/13106

The specs on that TRIAC seem to match up, needs 3mA to trigger the gate - well within the range of the Netduino digital outputs. 1.3V on the signal line ** unclear to me if it needs at least this or if something higher (3.3V) would break it??

This looks "that simple" - which tells me I'm probably missing something. Am I?

Thanks,

[Mario Vernari]

The main problem of the TRIACs is that they need a "dual-polarity" trigger. It means that you have to inject a positive current for triggering the TRIAC on for the positive wave, and another injection of current (negative, this time), for the opposite wave. Due to this complexity, it's better to achieve the bi-trigger directly from the target wave (e.g. the mains), which is already both negative and positive. Moreover, since most of the times the voltage involved is dangerous, an opto-isolation is absolutely required for safety. Although 24 VAC is a safe voltage to play with (unless you put the wires on your tongue), is much worthwhile using a SSR. Quick-'n-dirty: safety guaranteed, economy, reliability. Hope it helps. Cheers

[ErikN]

SparkFun has a couple EL Wire controller boards which use Triacs to switch high frequency AC from an inverter to the wire. One model has a microcontroller on board and the other is meant to be used as a shield.

I've never looked at the schematics to see how this differs from what I know about SSRs so this is presented as informational and entertainment purposes only. If you learn something it's your own fault.

Enjoy!

[Cabadam]

SparkFun has a couple EL Wire controller boards which use Triacs to switch high frequency AC from an inverter to the wire. One model has a microcontroller on board and the other is meant to be used as a shield.

I've never looked at the schematics to see how this differs from what I know about SSRs so this is presented as informational and entertainment purposes only. If you learn something it's your own fault.

Enjoy!

See, the diagram for that shield looks REALLY simple, which is what got my attention about TRIACs. If you remove the fact that there is an inverter involved (to go from DC back to AC), and remove the splitting for having both the plug and through holes, there isn't much left. Digital I/O -> Resistor -> TRIAC gate. Much simpler than the schematic suggested on the solid state relay page. Is that relay one just overly complex, or is it all actually necessary?

[Mario Vernari]

See, the diagram for that shield looks REALLY simple, which is what got my attention about TRIACs. If you remove the fact that there is an inverter involved (to go from DC back to AC), and remove the splitting for having both the plug and through holes, there isn't much left. Digital I/O -> Resistor -> TRIAC gate. Much simpler than the schematic suggested on the solid state relay page. Is that relay one just overly complex, or is it all actually necessary?

Probably I was sleeping when I was writing my post above: where am I invented 24VAC???

You should clarify better what's your target. If you mean driving mains, then it's not the TRIAC the problem, but the isolation of the circuit. If you take a look at the shield schematic, for instance, you'll see that there's no isolation between the MCU and the TRIAC leads (i.e. mains).
We might forget that never ever a similar circuit would be approved by the safety laws, but that is one of the simplest way to damage PC or else if the ground wiring isn't properly connected. I'd also add that it could be dangerous for the human health.

If you are dealing with voltages, AC or DC, above 30V, I wouldn't think any second more on choosing a SSR over any hand-made surrogate. A SSR is about $5, not $50: what's the problem?

[Cabadam]

Probably I was sleeping when I was writing my post above: where am I invented 24VAC???

You should clarify better what's your target. If you mean driving mains, then it's not the TRIAC the problem, but the isolation of the circuit. If you take a look at the shield schematic, for instance, you'll see that there's no isolation between the MCU and the TRIAC leads (i.e. mains).
We might forget that never ever a similar circuit would be approved by the safety laws, but that is one of the simplest way to damage PC or else if the ground wiring isn't properly connected. I'd also add that it could be dangerous for the human health.

If you are dealing with voltages, AC or DC, above 30V, I wouldn't think any second more on choosing a SSR over any hand-made surrogate. A SSR is about $5, not $50: what's the problem?

No no - 24VAC is correct (subject line is where you invented it ) I have no desire to play with 'main' power with my level of experience. My comment regarding the schematic ErikN referenced is simply how many fewer components are involved in that method vs the SSR schematic.

My ultimate motive here is not cost related at all. Rather, I'm hoping to reduce the form factor of my project a little bit. My current hardware I'm testing with are standard electromagnetic relays. I have an 8 relay board and a 2 relay board for 10 total relays.

As I'm looking how to enclose this and such, I first noted that there was a lot of wasted space in this approach. One of the primary reasons being all the screw terminals. NO, NC, Load for each relay. I don't need NO and NC - I only have one 24V power source, not individual sources for each relay. So that's 10x2 = 20 terminals reduced to one, plus the individual Load terminals.

The next question was what to put behind those terminals. I first went to solid state relays because they are slightly smaller in size, but like regular relays, need a number of extra components. While the original retail boards I have are mostly surface mounted components, my work certainly will not be, so I would like to reduce the component count if possible. As I continued to search around, I came across a couple examples people using TRIAC. And that is basically what brought me to this thread.

One other motivation for me to look at TRIACs was when I came across this project: OpenSprinkler. The schematic for it is available, and a closer inspection shows (in the bottom right) that the design is using a shift register connected to a series of TRIACs to control each of the 24VAC outputs. I'm not sure I am actually interested in the shift register part (although it is an interesting way to getting rid of 10 distinct signal lines out of the Netduino, so who knows...). Looking at the photos of their board, the amount of space taken up by the triac/terminal area is much smaller than what I have now.

Hopefully that gives some insight into what I'm actually trying to accomplish. Based on your initial post, I don't think any of these referenced systems have the isolation you referred to? I'm also not sure what they are doing to achieve the negative current? Disregarding the shift register for the moment, would negative current simply be a matter of turning the Digital I/O pin into an "input" rather than output so that current flows the opposite direction?

Thanks