Controlling a power supply with LabVIEW 8.2.

[archangel_hp]

Hello there. I'm triying to build a system to control a power supply. The power supply I'm using is LAMBDA ZUP60-14. The system is supposed to read a voltage with an agilent 34405A dmm. The problem is that because I don't know the transfer function of the power supply, I'm having problems making the right controller. Thanks for your time.

[Ray.R]

How are you interfacing to the power supply and DMM (to the instruments)? 

I think both Lambda & Agilent support GPIB and Serial (RS-232) interfaces.

How you communicate to / from the instrument will depend on the interface from your PC.

[archangel_hp]

Hi Joe. Yes, the interfaces are already done. I can current and the voltage from the labview and i can also get the meter readings. The problem is in making the program control the voltage. For example, I have a power supply connected to a load. I want the voltage in a load to be 10 volts and I set the voltage from the power supply to be 10 volts. However, there might be a voltage drop across the lines or something like that and I would be getting 9 volts instead of the 10 I want in the load. What I'm trying to do is a control system with LabVIEW that would rise the power supply voltage in order to make the voltage in the load (the one that is been measured by the dmm) 10 volts. Thanks for your time.

[smercurio_fc]

OK, this has nothing to do with the transfer function of the power supply. Just use the remote sense inputs of the power supply. That's what they're for. Look in the power supply's manual.

Note: If you really have a 1V drop to the load I would think you've got some other issues, like poor connections or poor cabling, or wire that's too low of a gauge.

[Ray.R]

As smercurio mentionned, it does not appear to be a LV programming issue.  A 1volt drop does appear like bad wiring, excessive loading, etc.  It's a very good suggestion to use the voltage sense on the power supply.

[archangel_hp]

Fellas, I don't mean to be rude, but the last message I posted was an example. Obviusly I can't have a 10% voltage drop on some connection cables, they would be really bad. However, in my actual set up I will have a small voltage drop on purpose, but it won't be constant, so I can't just suppose that it will be always .1 V and just have X voltage value + .1 in the power supply and then have the X voltage value at the load. Sometimes the voltage drop will be very small, sometimes it will be bigger, but what I need is a control system to always have X voltage in the load. Thank you.

[BillMe]

Most supplies have constant voltage (CV) and constant current (CC) modes. Your supply should be set to constant voltage (CV) mode since from your post it doesn't sound like it's being used as a constant current source. This takes care of any small variations in the load. And as stated in the other posts, if the supply supports remote sensing, convince your hardware designer to take advantage of it. Otherwise set the supply to the desired voltage at the load + any offset needed to overcome voltage drop between the supply output and the load. With CV mode, you don't have to constantly try and adjust the output voltage, something which to me sounds highly risky. One bug in your code could potentially fry your UUT.

[Brian_Beal]

I think that what you are trying to do is make Labview the controller, the DMM your feedback, and the power supply just an amplifier.  You should implement a PID loop in Labview to change the power supply voltage based on the feedback and error from setpoint.  If you don't have the PID toolkit, I believe there is a basic PID example that ships with Labview.

[archangel_hp]

Hi there. I tried to use the PID funcition but I'm not sure about which are the correct inputs. The box has a T_d, set point and process variable in the input. Which of these are the error of the closed loop system? Thank you.

[Brian_Beal]

I can't look at the VI right now, but the setpoint will be your desired voltage, process variable will be the actual voltage from the DMM, and I am not sure what T_d is.  There should be some tuning inputs somewhere.