Electronic Circuits and Projects Forum

OK so the deal is that I'm about 99% finished with this design for a 3Φ, shunt type, TCR Static Var Compensator. The theory here is that I've got an inductive load; placed in parallel with that I've got a compensating unit. Inside the unit I've got a compensating capacitor in parallel with an inductor. The inductor does not conduct all the time, it is controlled by a TRIAC, which receives it's gating signal from an opto-isolator. The gate pulses cause the inductor to conduct for only portions of each cycle, thereby absorbing a certain amount of complex power back off of the compensating capacitor, causing a tunable, net output from the unit that is easily varied for a range of load types and power factor values. I'm running the whole show with an Arduino, based on a six case phase shifted firing sequence, which is chosen depending on the amount of delay time required for a gate signal to phase A.

Logic testing confirms that the whole shebang, up to the output from the isolators, is working like a dream. I'm developing my firing angle correctly, the pulse outputs from the opto-isolators are correct and coherent with those generated by the controller. The controller is correctly timed to the mains voltage. It reads a zero crossing, and then implements the correct delay to fire the TRIAC.

This is where the problem happens: with no connections to the output from the opto-isolator, the output voltage can be seen on the scope to be correct. However, once I connect the opto-isolator (as shown in the attached diagram) everything goes to hell. The scope reads hideous, broken, 20V peak sign waves across the gate terminal and absolutely nothing seems to be happening as far as conduction through the TRIAC. I tested this by using a light as the load on the TRIAC branch. Not a thing. If anyone has any ideas or could point me to a circuit type I should be using, I would really appreciate it.

As an addendum. I have already tried a different version of this circuit that involves running the gate lead from the collector of the isolator, as opposed to the emitter, but this situation led to an extensive DC bleed onto the gate of the TRIAC, so I abandoned that format. The circuit was the same in that case as in my diagram, except that the resistor on the isolator was in the collector branch of the isolator and the TRIAC gate was connected directly to the collector.

Thanks ahead and happy Spring Holiday