As you rightly point out in your question, Canopus star trackers are pretty nifty instruments. I don't think it's really possible to do them justice in a single post, so I'll include a few summarizing points here and will include links to other articles, etc. for more information.
1. How do they work?
NASA Technical Report 32-1559 does an excellent job of explaining the general principle of the star tracker from an engineering perspective.
The Canopus tracker operates in a closed-loop servo control mode when it is tracking a star. This electro-optical control system tends to keep the star centered within a narrow field of view once a star of the correct intensity is acquired.... starlight entering the Canopus tracker travels first through a baffle, then a sun shutter, and then on to an objective lens which focuses the light on an image dissector tube. The purpose of the baffle is to prevent unwanted stray light from reaching the tracker optics. The sun shutter is another protection device that is actuated if the sun gets near the tracker field of view. When the starlight reaches the image dissector tube it is converted to an electrical signal. This signal is amplified by the image dissector tube and by the preamplifier. The demodulator then receives the signal, detects phase, and delivers a DC voltage to the voltage-controlled oscillator. This voltage is variable in both polarity and amplitude and determines if the voltage-controlled oscillator will issue count-up pulses or count-down pulses and what the pulse rate will be.The pulses from the VCO are routed through the count-up/count-down logic and then into the digital integrator. The digital integrator stores the pulse count, transfers position information to the roll error shift register, and controls the input to the 11-bit digital-to-analog converter. The voltage output of the D/A converter is used to control the roll deflection amplifiers. These amplifiers provide high voltage for the image dissector tube deflection plates. The voltage on these plates controls the position of the electron beam within the tube, and in this way the loop is closed. Thus, from star image around the loop to the deflection plates, the system automatically aligns itself to place the star in the center of the scan field of view.
The loop is closed by the VCO enable line. This line allows the VCO pulses to pass on to the digital integrator,thereby closing the loop. The acquisition logic, along with the programmable gates, determines if a star of the right intensity is in the field of view. If so, the loop is closed and automatic star tracking takes place.
The rest of the article goes a little further into the digital logic used to drive the star tracker circuity. Technical Memorandum 33-586 goes into some of the math and physics behind the star tracker, and NASA SP-8026 has an overview of the use of the Canopus tracker on various missions. (It also includes some information on other star trackers as well).
2. Are there any still in service today?
Yes! Voyager 1 and Voyager 2 both had Canopus star trackers, and given that they were launched in the 70s, they star trackers on board would have been of the original design that you're interested in. The US Air and Space Museum has a page on the Voyager star trackers (and has one of the test units in its collection). Given that the star trackers on the Voyagers are used to guide antenna pointing, and that the Deep Space Network is still in contact with them, one would presume they still are in use.
That being said, ye olde fashioned Canopus star trackers are not nearly as popular as they used to be, primarily because the design used in the 70s was susceptible to failure if overexposed to light reflected off the spacecraft or another unexpected source (for instance, see Lunar Orbiter I). Modernized star trackers and other attitude monitoring systems are now preferred, as they have been engineered to address some of the weaknesses of the original trackers and also have greater accuracy.
A good article for additional reading on modern star trackers is:
Zacharov, A. I., et al. "On Increasing the Accuracy of Star Trackers to Subsecond Levels." Solar System Research 52.7 (2018): 636-643.
