PID Control

[dr.shulgin]

Hi!

I would like to know if the following task is resolvable without a LabView Realtime environment. A NI PXI System with embedded controller, NI PXI 4472 and NI PXI 6230 card ought to drive a bipolar power supply (BOP 20-20 Kepco) to hold the working point of a solar cell under modulated illumination. The solar cell is connected in serial with a shunt resistance to the power supply. As the illumination is modulated (20-200Hz) the current produced in the solar cell will also oscillate, resulting in an oscillating working point of the solar cell. The idea is to measure the voltage drop at the cell with a NI PXI 4472 card. The voltage drop at the cell is know compared with the desired voltage (working point), after a PID loop the NI PXI 6230 card gives an analog voltage output to the power supply. (Perhaps the measurement circuit is better to understand for the example of short circuit conditions. Under constant illumination, the cell delivers a constant current and there is a constant voltage drop at the shunt resistance, resulting in the effect that the cell is not at short circuit conditions. The power supply is know set to a constant voltage at which the cell is at short circuit conditions. No control would be necessary. Know a modulated illumination is added to the constant illumination, resulting in an oscillating current from the cell and and oscillating voltage drop over the shunt resistance. As incident light and current from a solar cell are only proportional under short circuit conditions and I would like to measure the current from the modulated illumination, I need a PID control which forces the power supply to maintain short circuit conditions at the cell. The same control should be possible for all the working points of the cell, for similar purposes.)

 As the currents are high (18A DC + 50mA  AC), and there are only real differential signals I choosed the equipment described above. Know I ask myself if a system without Realtime is fast enough to control the system with an adequate precision. The modulation frequency of the light is not faster than 200Hz, which is quite slow, but the system has also to solve the following tasks simultaneous. Digital LockIn operation of the two other channels (current from cell, and current from a monitor cell), analog utput of a TTL with 200Hz, output of a constant voltage and a watchdog for the I/0 lines.

Thank you for your responses...

[Wolfgang_Z.]

Hello dr.shulgin  

Real-time response is the ability to reliably and, without fail, respond to an event, or perform an operation, within a guaranteed time period.
A misconception about real time is that it means quick. More accurately, real time means in-time. In other words, a real-time system ensures that responses occur in time, or on time. With general purpose operating systems, you cannot ensure that your response occurs within any given time period, and calculations might finish much later or earlier than you expect them to. In Windows for example the processor loading can rise to 100% because of automatic starting of a antivirus software. Maybe then you are to late with the responce of you PID because LabVIEW can not calculate the result timely.

If you have to guarantee the result of your PID algorithm timely use LabVIEW Realtime. When you do not have a problem if you get the PID result somtimes a bit later you can use LabVIEW for Windows or a other not real-time OS. Note: Real-time don't mean quicker like not real-time.

Regards,

WolfgangZ