# Solid-state relay design

Discussion in 'General Electronics Chat' started by dewasiuk, Mar 10, 2012.

1. ### dewasiuk Thread Starter New Member

Feb 14, 2011
24
0
Using the below schematic, why was the value of the 22Ω resistor chosen?

My first instinct would be to think that it's for limiting current, however looking through the datasheet for the MOCxxxx isolators, I'm not sure which parameter would tell me how much current should be used and it seems to be a little more complicated to derive the output circuit formulas used in the design.

Also I could not find out whether the output current from the zero crossing circuit would be enough for the gate of the triac controlling the load.

2. ### ErnieM AAC Fanatic!

Apr 24, 2011
7,437
1,626
I think that value is more dependent on the gate requiremts of the Q4015L6 then the isolator. I'm not up on triacs so I can't explain it any better.

3. ### #12 Expert

Nov 30, 2010
16,703
7,341
At first glance, it seems absurd to use 22 ohms to limit the gate current of the Q4015 with 22 ohms. The gate certainly can't survive 5 amps from a 120 volt AC line. However, there is no current until the triac is triggered, and as soon as the triac is triggered, the voltage from line to MT1 collapses to the saturation voltage of the triac (1 volt or less). Therefore the 22 ohm resistor will be figured on the gate current (50 to 100 ma) when the line voltage is below 2 volts (Volts gate max).

You really have to pick the datasheets apart trying to find these numbers!

File size:
943.4 KB
Views:
20
File size:
148.3 KB
Views:
12
4. ### dewasiuk Thread Starter New Member

Feb 14, 2011
24
0
That's alongside the line voltage alternating from 0-120 in quadrant 1, I presume?

Last edited: Mar 12, 2012
5. ### #12 Expert

Nov 30, 2010
16,703
7,341
The point is, whenever the MOC is not conducting, the voltage doesn't matter to the 22 ohm resistor because no current is flowing, and, whenever the MOC does conduct, it doesn't take a millisecond for the Q4015 to fire and collapse the voltage from Line to MT1, at which point the 22 ohm resistor only has 1 volt across it. It doesn't matter which quadrant you are thinking about. It's the same in both quadrants.

Are we having a difficulty in comunication?
That's why there are dozens of helpers here. A different person can explain the same thing in different words until you understand. I invite another helper to re-word this if necessary to help dewasiuk.

6. ### dewasiuk Thread Starter New Member

Feb 14, 2011
24
0
Okay I understand it's the same in both quadrants, but now I'm confused about the voltages. Would I use KVL to find the current with Vline = I*R + Vmt or do I have to deal with paramaters of the teccor triac affecting the circuit?

7. ### #12 Expert

Nov 30, 2010
16,703
7,341
I'm so old that I don't know what KVL means. I probably use it all the time, but I don't know that name for it.

It comes down to the parameters of both the MOC and the triac.

Going microscopic: The MOC gets a signal on its LED and the zero crossing detector waits for zero voltage. Zero line voltage happens, and then, as the line voltage goes positive (or negative) current starts flowing through the 22 ohm resistor to feed the zero crossing detector to fire the triac inside the MOC. That triac gate current is the first current that has to get through the resistor and that (Volts gate total) (plus whatever voltage is required to get through the zero crossing detector) is the first that has to be surpassed. When the line voltage is high enough to excede the gate voltage and deliver enough current to start the triac avalanche in the MOC, that triac pours current into (or out of) the gate of the Teccor triac.

Now the line voltage has to go through the resistor and a conducting triac in the MOC to get to the Teccor triac gate. That gate is guaranteed to require no more than 2 volts and 50 milliamps. (The line voltage is still getting larger.) When those conditions happen, the Teccor triac goes conductive and the guaranteed maximum voltage across MT2 to MT1 is one volt.

Suddenly, there is 1 volt or less across the resistor and the gate to MT1 path in the MOC. That triac can't sustain and it opens, but that doesn't matter to the load because the Teccor triac is already slammed hard on.

Depending on what country you live in, this happens 100 or 120 times per second. It's a bumpy graph as a gate voltage has to be exceeded and a triac comes on then another gate voltage has to be exceeded and another triac comes on.

I'd be tempted to use an oscilloscope and reduce the resistance until I get the fastest, cleanest turn on. The Teccor triac has a gate power limit of 1/2 watt, but that is going to be "not a problem" because the MOC can't feed any gate current after the Teccor goes into conduction. i.e. most of the time there is no gate drive to the Teccor because it has collapsed its own supply. The Teccor just conducts until the line voltage gets very close to zero again, the current falls below the minimum sustaining current of the Teccor, and the whole thing starts again with the zero crossing detector waiting for zero to happen and the next half cycle to gain voltage magnitude.

To get the spec sheet declared 50 ma into the gate of the Teccor requires 1.1 volt across the 22 ohm resistor. In that perfect instant the line voltage has to be 1.1 plus the volts across the MOC triac plus the Vgt of the Teccor. Now you suddenly realize that the value of resistance is a way to decide how long after zero crossing it takes for the Teccor to come on. It has almost nothing to do with what you started out thinking!

Maybe that's why I don't know what KVL means?
I only know that the resistance can't be zero and it can't be more than 120 Vpeak over .05 amps.
(That's 3394 ohms.)

Somewhere in between gets a turn on only a volt or so above the sum of the diode junctions break-over voltages, and 22 ohms seems OK to me. If you want better than that, I have to turn this over to another helper. That's all I have.

Last edited: Mar 12, 2012
8. ### dewasiuk Thread Starter New Member

Feb 14, 2011
24
0
KVL = Kirchoff's voltage law

I have a general idea but it's difficult to follow without diagrams, etc. The thing that still confuses me is that according the qxxxx datasheet, it will put a 1.6 volt potential across the resistor and the triac in the MOC, yet that triac in the MOC drops around 1.8V which is more than what's being supplied in that instance of time. That violates KVL.

Also it would be nice if the datasheet actually showed what the zero crossing circuit looks like.

Last edited: Mar 12, 2012
9. ### #12 Expert

Nov 30, 2010
16,703
7,341

As for the next part, the Teccor triac does have a voltage from MT1 to MT2 of 1.6 volts when it is passing 21. something or other amps (according to the datasheet). Since that is not enough voltage to keep the triac in the optoisolator on, it turns off immediately after the Teccor triac comes on.

quote: "Suddenly there is 1 volt or less..."
I guess I could have said 1.6 volts or less but I was looking at a different part of the spec sheet.

This isn't simultaneous. It's happening across time.
Is that the "diagram" you want?

ps, I've seen the drawing of the insides of a zero crossing detector. It's cool and it works but it has a voltage cost of 1 diode drop, I think.

sound.westhost.com/appnotes/an005.htm
This isn't the guts of an IC that I saw but it shows the principle. A transistor is operated as a saturated switch until the base drive is missing, then the stored voltage in a capacitor is allow to show up at the collector.

Last edited: Mar 12, 2012
10. ### dewasiuk Thread Starter New Member

Feb 14, 2011
24
0
I meant diagrams on top of my original diagram(current and voltage labelings, etc.). But I did actually set up the circuit and scoped it myself and used a pot for the resistor and could see how changing it affected the firing time, rather than worrying too much about the current. Thanks

11. ### #12 Expert

Nov 30, 2010
16,703
7,341
Excellent! A scope is the only way I would even attempt to see a diagram of the voltage changes. This is because the datasheet gives guaranteed maximum voltages and most devices work better than the guaranteed limits. The switching voltages and saturation voltages will almost certainly be smaller than the specs.

I would also suggest that you put 1 ohm or so between Pin 4 of the MOC and MT1 of the Teccor so you can see that the time duration of the current flow to the Teccor gate is very small. This will demonstrate that the triac in the MOC stops conducting in a matter of microseconds.