Labview PID controller PWM

[´MikeSv]

Hello everyone.

Iam working on a Labview temperature control program which makes use of a PDI controller, a heater and a thermocouple reading the temperature.

Iam able to read the sensor and I think I have managed to implement the PID as well.

What I need to do now is to convert the output of the PID into an appropriate Duty Cycle in % to energize/de-energize a Solid State Relay for the heater supply voltage.

To generate a square wave I thought using the Square Waveform.VI. (see image below)

The output for the PID is in the range of 0...5V so I guess I have to scale this output in some way?

Thanks in advance  for any help,

best regards,

Michael

[nathand · Accepted solution]

I haven't seen that Simple PID VI in a while, and if you have access to the PID toolkit (now included with non-Base versions of LabVIEW) you should use that one instead. In either case, though, the output limits are set by the inputs to the VI - in your case, the "lower limit" and "upper limit" controls. Change those so they're 0 and 100 and then use the output as the duty cycle.

[aputman · Accepted solution]

Is it a linear relation between 0-5V and 0-100% duty cycle?  If so, it should be pretty simple to do a linear conversion. 

For example, if you get 3.7V from the PID, solve the equation:

3.7V / X % = 5V / 100%

X = 74% Duty cycle

If you have the PID toolkit from NI, I think this functionality is built-in to the PID.vi.  You simply specify the output range and the VI will handle the rest. 

[´MikeSv]

Oh man....

Sure it is linear. It's too late at my place.

Thank you!

Regards,

Michael

[´MikeSv]

Thanks for your reply. I actually have access to the toolkit.

I think I found the old PID VI in some exemple code.

I didnt know it was that easy to scale (If this is the right word here) the output of the PID that easy.

regards,

Michael

[´MikeSv]

Thanks for your reply. I actually have access to the toolkit.
I think I found the old PID VI in some exemple code.

I didnt know it was that easy to scale (If this is the right word here) the output of the PID that easy.

regards,

Michael

[nathand]

---

@´MikeSv wrote:

I didnt know it was that easy to scale (If this is the right word here) the output of the PID that easy.

You're not scaling anything here, it's just a limit on the output. The proportional gain scales the output to the input. So, you could use the 0-5 output and scale it to 0-100 as aputnam suggested, or you could take the direct route of changing the output limits, in which case you would need a proportional gain (Kp) 20x larger (to turn 5 into 100) for the same results. Mathematically these are identical.

[´MikeSv]

Hi again.

Thanks for the explaination! 

Regards,

Michael