Triac low output problem

Thread Starter

ceculbe

Joined May 1, 2012
3
I've got a strange (to me) problem with some interesting scope traces:

The circuit is two separate 3023 optoisolators driving two separate triacs, each one powering a large 3-phase contactor. One contactor is engaging fine. The other one is rattling badly because it isn't getting full power to pull in.
Each circuit looks like this:


The top photo shows trace from the good triac circuit, working fine. Bottom photo shows the bad one. It looks like the triac is dropping out right after hitting peak positive voltage.





- What's going on?
- More importantly, will replacing the triac fix this? A new board is going to cost me $480.

Thanks for any help. I'm a first time poster who is in over his head on a repair - really more of a mechanical guy than an electrical expert. (I'd be happy to trade mechanical advice!)
 

gerty

Joined Aug 30, 2007
1,305
Have you checked logic input signal?
If thats ok, it should be simple to change the 3023.
I would also check connections/traces from triac to terminal strip.
 

praondevou

Joined Jul 9, 2011
2,942
The triac should only turn off when the current passes through zero. Since you are driving a coil, maybe the current passes through zero where you see the voltage drop. To be sure you would need to measure load current too (on a scope).

two things to verify:

1. current through the photodiode (> 5mA)
2. Holding current required for the load triac. Which triac is it?
 

t_n_k

Joined Mar 6, 2009
5,455
As you can see from the attachment I can simulate something very much like the result you are seeing. The problem arises from too short a gate pulse. If you are applying a pulse (y/n ??) to the triac gate you may need to make it longer - say 90% of the AC cycle in duration.
 

Attachments

praondevou

Joined Jul 9, 2011
2,942
As you can see from the attachment I can simulate something very much like the result you are seeing. The problem arises from too short a gate pulse. If you are applying a pulse (y/n ??) to the triac gate you may need to make it longer - say 90% of the AC cycle in duration.
Is this the voltage measured over the triac or the load?
 

t_n_k

Joined Mar 6, 2009
5,455
Is this the voltage measured over the triac or the load?
The load voltage. The PSIM schematic used is attached. I guessed the circuit values. With a control duty cycle of 0.7 I obtain the waveform indicated earlier. With the duty cycle increased to 0.8 the problem vanishes and the load has the full sinusoid imposed. One would have to know the actual circuit parameters to correctly advise the OP.
 

Attachments

Last edited:

gerty

Joined Aug 30, 2007
1,305
. One would have to know the actual circuit parameters to correctly advise the OP.
That would be a great help. If all he is doing is turning on a contactor, he really doesn't need to pulse it at all.
 

praondevou

Joined Jul 9, 2011
2,942
The load voltage. The PSIM schematic used is attached. I guessed the circuit values. With a control duty cycle of 0.7 I obtain the waveform indicated earlier. With the duty cycle increased to 0.8 the problem vanishes and the load has the full sinusoid imposed. One would have to know the actual circuit parameters to correctly advise the OP.
Just for curiosity, can you put the load current in your picture?
 

Thread Starter

ceculbe

Joined May 1, 2012
3
Thanks for the help. The logic input is driven from a custom controller IC. Its outputs control two 120VAC heaters. The controller is reading voltage across two thermistors, reading some jumpers to determine the temperature setpoint, and then driving the optoisolator.

I didn't think to check the stability of the logic input to the optoisolator. I don't know why the controller would be pulsing that line, but I'll check it. Could ripple in the 5 VDC supply cause enough variation to throw off the triac but not the rest of the logic? The identical circuit for the other contactor works fine...

I ordered a new optoisolator, capacitor, and my best guess for the triac (its not marked), but I'll check logic input before swapping any of those.
 

praondevou

Joined Jul 9, 2011
2,942
Thanks for the help. The logic input is driven from a custom controller IC. Its outputs control two 120VAC heaters. The controller is reading voltage across two thermistors, reading some jumpers to determine the temperature setpoint, and then driving the optoisolator.

I didn't think to check the stability of the logic input to the optoisolator. I don't know why the controller would be pulsing that line, but I'll check it. Could ripple in the 5 VDC supply cause enough variation to throw off the triac but not the rest of the logic? The identical circuit for the other contactor works fine...

I ordered a new optoisolator, capacitor, and my best guess for the triac (its not marked), but I'll check logic input before swapping any of those.
The values of the components that determine the photodiode current and gate signal to the triac may just be at the limit. This is why one circuit can work and the other not. It doesn't need to be necessarily a problem with a defective component.

That's why I asked for the photodiode current and triac model.

Sure - here it is
Just as I thought, current passes through zero. But why is the simulation current having a DC-Offset?
 

t_n_k

Joined Mar 6, 2009
5,455
Just as I thought, current passes through zero. But why is the simulation current having a DC-Offset?
It depends at what point in the AC voltage cycle the triac is turned on as to whether you have a transient DC offset. Do you recollect your circuit theory transient analysis studies from the old days?
 

Thread Starter

ceculbe

Joined May 1, 2012
3
UPDATE:
Turns out it was the optoisolator. It was a SMD, so not so fun to replace on a crowded board, but it works now. Thanks for the help.
 
Top