PHYS1630: Thermal Control Systems
Lab #1: PID (2018F)

Introduction

In this lab you will tune a PID temperature controller

Pre-Lab (Do this before your assigned lab period)

Read through the Wiki to familiarize yourself with the operation of a PID controller as well as the manual tuning method.

Read through the Quick Start Guide to the LDT-5910C controller describing how to set parameters.

Background

Building a stable DPSS laser requires a lot more than simply shining a large 808nm diode pump laser at a small vanadate crystal and hoping for green light :). Consider the Crystal Laser DPSS system, a typical small DPSS:

DPSS Components Seen here unmodified is the Crystal Laser Green DPSS system employed in this lab. Radiation from a powerful 808nm laser diode, mounted on a thermoelectric cooler, is shaped and focussed by a two lens and two prisms. The focussed beam pumps a tiny crystal of vanadate (Nd:YVO4) with an attached KTP frequency-doubling crystal. The entire vanadate/KTP crystal is mounted on a second thermoelectric cooler.

The green output beam then passes through a beamsplitter which diverts a small portion of the output towards a photodiode used to monitor the output power - the rest passes through a filter which removes residual IR (both 808nm pump radiation and 1064nm laser radiation) to become the output beam.

The pump laser is mounted on a thermoelectric cooler for two reasons: one is to simply sink excess heat produced during the operation of the laser itself and the second is to thermally-stabilize the diode to maintain wavelength stability ... if the diode wavelength is allowed to drift during operation, output power will also drift.

DPSS Experiment - Thermocoolers As well as the pump diode, the vanadate/KTP hybrid crystal is also mounted on a thermoelectric crystal for stability - the necessity for accurate temperature control of a Second Harmonic generator (SHG) crystal.

In order to tune a system such as this, the temperature must be controlled. One must add some sort of temperature sensor to each thermoelectrically-cooled block - the diode and the crystal - in order to monitor each during tuning. In the case of this lab, factory-installed thermistors are used to monitor each as feedback for the temperature controllers. A second thermistor was added to each controlled element allowing montoring of actual; temperature.


Equipment

The experiment consists of the temperature control of the laser diode (LD) in a Crystal Laser CL532. The LD is mounted on top of a small Peltier-effect TEC module (with a maximum current rating of 1A). An ILX LDT-5910C temperature controller (which has a full PID control loop) drives the TEC and is provided with feedback via the factory-installed thermistor on the LD mount. The LD can be powered as well - it serves as a heat source for the LD mount.

In addition, a separate thermistor was added to the LD mount allowing the temperature to be monitored using a DMM or an oscilloscope. The scope is particularly useful as temperature variances can be graphed on a very long timebase (e.g. 5 seconds/division) to show oscillations and response to temperature settings. A circuit linearizes the output of the thermistor producing a voltage as a function of temperature as follows:
Voutput = 5 * (0.0092 * Temperature_in_C + 0.2675).

PID Experiment Setup
The LDT-5910C temperature controller (left bottom) controls the temperature of the LD mount. Above the controller is a Newport 505 laser diode driver - with current flowing through the diode, heat is injected into the system. The oscilloscope on the right allows actual temperature of the LD mount to be plotted over time (see the equation above).

Temperature Plot
Temperature of the LD mount plotted against time on a digital oscilloscope. In this example the vertical axis is 100mV/division and can easily be calibrated in absolute units of degrees-C while the horizontal axis is in units of 5 seconds/division. The TEC was set for a higher temperature at about one second into the plot and an oscillation can be clearly seen here. Use the cursors (Y) to determine the actual voltages (and hence temperatures) corresponding to each grid on the display.

The Experiment

Begin by setting the P, I and D loop gains on the LDT-5910C temperature controller to zero and the limits to +1.00A and -1.00A. Enable the control and turn the output on. Select a temperature of 25.0C.

FIRST, begin by investigating the operation of the P loop. Set the P gain very low (1.00) and observe the effect. Next, set the P gain very high (25.00) and observe the effect again. Graph the variations of temperature on the scope and capture the output (using a cell-phone camera is easiest). Note the minimum and maximum temperatures of these oscillations. With the same gain settings, switch the Newport laser diode driver ON at 500mA so that the diode dissipates about 1 Watt (assuming a voltage of 2V across the diode). Again note the minimum and maximum temperatures of these oscillations - does the presence of heat improve the situation or make it worse?

When recording the scope screen, use the cursors (Y) to determine the actual voltages corresponding to each grid on the display. This will allow the screen shot to be calibrated in actual degrees-C.

Turn the LD driver OFF. Next, optimize this gain using the "Simple Method": set the P gain to one-half of the value at which steady oscillation is observed. Graph the output when the temperature is changed from, say, 25C to 30C to observe the reponse.

SECOND: investigate the operation of the I loop. Set the I gain very high (10.00) and observed the effect on output. Decrease the I gain to zero then optimize: Increase the I gain until the offset is decreased and response time to a temperature change is reasonable however the system is stable.

THIRD: investigate the operation of the D loop. Set the D gain very high (10.00) and observed the effect on output, especially when the temperature changes. Decrease the D gain to zero then optimize: Increase the D gain until response to temperature change is improved. Increase the gain further to improve response however at the expense of increasing overshoot. Graph both.

Finally, having set the controller yourself, allow the controller to Auto-Tune. Graph the response to temperature change using these values.

Lab Report

For this experiment, an abbreviated lab report is required (word processed, never hand-written) with the same format as PHTN1300. Answer each question in the form "4a., 4b., 4c. ..." with each new question (#4, #5, etc) beginning on a new page. Do NOT answer an entire question (e.g. question 4) as a single paragraph without identification of sub-parts ('a', 'b', etc). Submit the lab report in a bound folder NOT simply a pile of loose, stapled papers nor a thick binder!

Each student must submit a unique lab report - no portion other than the results must be shared between lab group members.

  1. Summarize the tuning method used for the PID controller. Include graphs as required (calibrated in degrees-C) to describe each step and demonstrate effects such as oscillations, overshoot, and response time to temperature changes. Use actual numbers from the lab and report the final gains used.
  2. Show the Auto-tune values and a graph of the response to temperature change using these calues.