My experience with variable speed drives has transcended through several generations of "new technology". The progress in the efficiency and control characteristics of "modern" VS drives has moved along as fast as the "new technology" of solid state electronics and PC control. Compare a desk-top computer of 2006 to one of 1996 and you will get a idea of the progress and changes that have occured. I'm behind now as it's been several years since my last experience.
My history with VS drives started in the 70's with the Original "Ward Leonard" drive.
This drive first became available around 1900 and found wide application. The systems that I was familiar with were still in operation into the 70's. I'm sure that some of the systems are still being used.
This (Ward Leonard) system used a constant speed (prime mover) AC motor to drive a (constant speed) DC generator. Output voltage (DC) from the generator was regulated by increasing or de-creasing the (DC) field voltage supplied to the generator. This generator output was then used to supply power to a DC motor (or several in parallel) that produced the work.
The constant (excitation) voltage for the generator (and work motor) was usually supplied by another (small) DC generator known as the "exciter". It was also driven by the prime mover and usually mounted on the (tail end) of the larger generator.
This was replaced in later years by a "static exciter" using a early model "rectifier bridge". Either way, two seperate sources of DC power were required in order to power a third variable speed DC (work) motor.
Primary speed control (variable voltage) for the work motor(s) was provided (and varied) by using a rheostat to increase or decrease the field voltage supplied to the generator.
If several DC (work) motors were powered (in parallel) by the same generator output, tandem speed control was possible, The generator supplied the same variable DC voltage to the (work) motor armatures, causing each motor to run at basically the same speed. Unison in speed between each individual motor was regulated by individual (manual) rheostats used to increased or decreased (+-10%) the field voltage to each motor.
Believe it or not, after the motors speeds were "synchronized", the system was pretty reliable.
These systems were used a lot in the textile industry on systems that might have up to 12 seperate DC motors. Each motor would power a different process on a continuous output machine, and variable speed control was necessary, along with the need for all (work) motors to run in unison.
It would be considered a "dinosaur" by todays standard, but it was a great advance in technology in those days. The system was really dependable if not efficient.
That progressed to a common (solid state) power rectifier (DC drive) to supply the DC voltage to the tandem (work) motors. Control was basically the same, variable armature voltage with constant (+- 10% controllable) field voltage for each motor. This eliminated the prime mover, generator and exciter.
Then came Individual individual DC drives for each motor and
using "tachometer" (voltage) feedback from a tach (mounted on each motor) to supply a reference "speed" (voltage signal) to a central PC control. This control automatically regulated the speed of the motors. This was a great improvement in regulating speed of tandem motors.
This (IMO) is the best system for a "constant horsepower drive. You need to size the motor HP (and drive) to the maximum torque required at the minimum speed.
Next came the AC "Variable frequency" drives. These control the frequency (Hz) of the supply voltage to a AC motor and thus allow variable speed operation. This is the best system (IMO) for a "constant torque" drive application. Most industrial application are "constant torque".
The motor HP (and drive) needs to be sized to the maximum torque at the maximum motor speed.
The advantage to variable frequency drives is the lower motor costs and higher operating effeciency.
When they first "came out" regular AC motors were all that were available, and they weren't designed for variable speed operation. The torque characteristics of the motors and the generated heat was a problem and led to premature motor failure. This has been vastly improved by the introduction of AC motors specifically designed for variable speed operation.
This is just my opinion based on my experience, others experience may vary.
Feel free to critique.
Another long post, sorry but it's a interesting subject (to me anyway).
steve