Hello

I am designing DC power supply by use of Zero-Cross Triac Driver Optocoupler. When testing it in LTspice with SCR and simple resistive load, all is OK.



As you can see on the picture below, current flows through R5 resistor only when current flows through led driver, as it suppose to be



But when we connect it to bridge rectifier, LTspice some weird thing happens



Half sine-wave pulse are missing. Does it means that SCR doesn't opens for no reason?.



Also adding capacitor for DC, (LED driver ->5 cycles pulse 500m period) looks like TRIAC doesn't think closing at all.





What could be the problem?
.

 

According to this paper there are two things that might be the cause:

1. The quantity of charge which remains when the current drops to zero
2. The slope of the reapplied voltage during turn-off.

https://www.st.com/resource/en/appl...c-behavior-at-turn-off-stmicroelectronics.pdf

 

There has been something very similar to this in the last few weeks on here,
was that yours as well ?

What voltage rated capacitors are you going to use to "smooth" the output voltage to DC ?

 

If I remember correctly, it's something on these lines.
Now that you have a smoothing capacitor across the load resistor, no current can flow in the bridge until the voltage across the bridge exceeds the voltage on the smoothing capacitor. That leaves the snubber capacitor charged up which defeats the zero-crossing detector circuit. I think (though it's some years since I last had trouble with this) that if you remove the snubber it ceases to be a problem.
It certainly doesn't affect the devices without the zero-crossing detector.
Preumably, you intend to regulate the output using the triac? In which case, you need the type without the zero-crossing detector so you can fire it random-phase.

 

Hello

I am designing DC power supply by use of Zero-Cross Triac Driver Optocoupler. When testing it in LTspice with SCR and simple resistive load, all is OK.

View attachment 257919

As you can see on the picture below, current flows through R5 resistor only when current flows through led driver, as it suppose to be

View attachment 257918

But when we connect it to bridge rectifier, LTspice some weird thing happens

View attachment 257920

Half sine-wave pulse are missing. Does it means that SCR doesn't opens for no reason?.

View attachment 257921

Also adding capacitor for DC, (LED driver ->5 cycles pulse 500m period) looks like TRIAC doesn't think closing at all.


View attachment 257928


What could be the problem?
.

For an application like this, use a bridge with 2 SCR's. They will perform better.

 

According to this paper there are two things that might be the cause:

1. The quantity of charge which remains when the current drops to zero
2. The slope of the reapplied voltage during turn-off.

https://www.st.com/resource/en/appl...c-behavior-at-turn-off-stmicroelectronics.pdf

Thank you. Interesting paper.

 

There has been something very similar to this in the last few weeks on here,
was that yours as well ?

What voltage rated capacitors are you going to use to "smooth" the output voltage to DC ?

No, that was not mine. Capacitor min 400V

 

If I remember correctly, it's something on these lines.
Now that you have a smoothing capacitor across the load resistor, no current can flow in the bridge until the voltage across the bridge exceeds the voltage on the smoothing capacitor. That leaves the snubber capacitor charged up which defeats the zero-crossing detector circuit. I think (though it's some years since I last had trouble with this) that if you remove the snubber it ceases to be a problem.
It certainly doesn't affect the devices without the zero-crossing detector.
Preumably, you intend to regulate the output using the triac? In which case, you need the type without the zero-crossing detector so you can fire it random-phase.

I think you are right, the problem is min current flowing through triac, which retain zero crossing triggering. If min current is provided, then circuit is working. I prefer Zero-crossing driver because it reduces di/dt dv/dt and EMI correspondingly.
Also without snubber cap, if there is a smoothing cap on an output, it defeats zero-crossing, in case if driver triac current is small.

 

For an application like this, use a bridge with 2 SCR's. They will perform better.

I will try. Thank you

 

Here's another idea if You just want a Solid-State-Switch ..........
It can easily be slowed-down, ( ramp-up, ramp-down ).
.

 

. I prefer Zero-crossing driver because it reduces di/dt dv/dt and EMI correspondingly.

But you can’t achieve regulation with a zero-crossing device.

 

But you can’t achieve regulation with a zero-crossing device.

.
Well, You can, sorta-kinda.
But it's not pretty, and impossible to filter.
.
.
.

 

But you can’t achieve regulation with a zero-crossing device.

Why? This is zero-crossing regulation. Duty cycle 11% and 80%.
How non zero-crossing regulation better?

 

The real problem is that the Thread-Starter has not stated the
intended purpose of "Regulating AC-Voltage".
A Triac is almost always a poor choice.
.
.
.

 

Why? This is zero-crossing regulation. Duty cycle 11% and 80%.
How non zero-crossing regulation better?

View attachment 258068

As @LowQCab says "It's not pretty".
That's a lot of ripple, and I bet there's a lot of surge current into that capacitor every time it switches back on.

 

Wonder if it might be time for the thread started to pop back in here

 

Here's another idea if You just want a Solid-State-Switch ..........
It can easily be slowed-down, ( ramp-up, ramp-down ).
.
View attachment 258041

I'll take note of that. Didn't heard about FDA215 before.

 

I'll take note of that. Didn't heard about FDA215 before.

.
It's not suitable for high-speed Switching with Big-FETs, but works great otherwise.
.
.
.

 

Problem solved. Found the reason why zero-crossing method was not working with smoothing capacitor and with little circuit modification it start working. But as it turned out later, it's not pretty to do like this )).
Thank you all for help.

 

Problem solved. Found the reason why zero-crossing method was not working with smoothing capacitor and with little circuit modification it start working. But as it turned out later, it's not pretty to do like this )).
Thank you all for help.

For others in the future,
Could you feed back to the community what the fix was please