3 series shunt regulators

[ultrasoundguy2]

I am making a long cable with many shunt regulators, only 3 of which are shown in the attached figure. The shunt regulators help prevent the whole cable from working if there is a failure. The long cable is similar to Christmas light scenario where if one load or shunt regulator diode may go out, the whole thing will stop working. For this application, there maybe 100 or more shunt regulators each supplying power to a unique device.

The way we are protecting the entire cable from failure of an individual diode is with a crowbar circuit in each shunt regulator. Ive pretty much copied this off Wikipedia.

If a diode is removed from one of the lower shunt regulators (voltage can get higher than the diode), the voltage should get higher, thus causing the shunt voltage regulator (TL431) to draw more current. This is turn makes the voltage higher than turns on the gate for the TRIAC (MAC15). When the TRIAC conducts, this allows a short circuit to happen which bypasses the load and or open circuit. This should be simple, right?


There is also supposed to be a parallel ballast resistor that is parallel to the load. The way to calculate it is Rballast = 6.8(diode drop) / (total current – current in load - shunt current). This comes from an established design (although I don't know the schematic). The values Im getting are about 25 ohms. The goal of the ballast resistor is that “dissipate any power supply current not used by the load”. When I add in the ballast resistors, Im having problems getting the load to switch at all, I think all the current pretty much is going through it.


Im having a few different problems.. One is that I can TRIAC to consistently turn on. In post2.png, I can get the TRIACs to turn on but not with the ballast resistors present.


A different problem is that I keep getting convergence errors. I keep setting the tolerances to be low and still get them , ie set voltage precsion to be from pA to A and still get them. Its crazy. This is with orcad capture.

Anyone with any experience in this or see anything that Im missing, its really appreciated.

 

[ultrasoundguy2]

http://files.engineering.com/getfile.aspx?folder=5b32c273-a224-4ea9-984b-c19d00aa3b12&file=post.png

http://files.engineering.com/getfile.aspx?folder=b3a7b490-6a06-4fc3-848a-14c1ad7083c7&file=post2.png

 

[ultrasoundguy2]

software is orcad capture 16.0

 

[LiteYear]

No experience with Orcad, but it does sound like your ballast resistor defeats the purpose of the shunt regulator. The shunt regulator works by detecting when the load becomes high resistance. When it does so, the voltage across the shunt regulator goes high and the shunt crowbars to short out the load. The ballast resistor on the other hand, will keep the voltage across the load from rising too high, since even if the load goes open circuit, the ballast resistor will look like a load. In fact, looking at your numbers, at 400x the resistance your load is insignificant compared to your ballast.

So to answer your first question, I'm not surprised the ballast resistors stop your shunt regulator from working. I don't know about the convergence errors.