Re: Low cost Servo Controller

David Howland wrote:

> Has anyone seen high performance Servos with Servo controllers which competed in price to the stepper and stepper controler counter parts? I've picked up a few surplus servos to investigate their controller requirements. The simple ones seem to be a brush type DC motor tied to an encoder wheel. These will work out of the box with the new board. The more complex servos, seem to be multi phase brushless motors and more complex encoders. They are certainly in budget at the surplus stores, but the missing controllers and drivers may not be. It seems to make since to create an electronic interface which is low in cost and will allow the use of Servo technology on a reasonable budget. If that makes since, perhaps Servo like performance can be obtained near the same cost as stepper performance.
>
> Will someone make a list of stepper controller costs vs. Servo controller costs plus motor costs, with a little performance data one might get at each cost? Aside from the main stream sources such as Minarik or the likes. What does it look like as you travel up the performance - cost curve for "underground" motion controller components and difficulty with CAM software interfacing? Even more important, What kind of CNC performance might one end up with along the way? Frankly, I hate dissapointment, even when it's my own fault for not spending enough money.

I designed and built servo amps to run the axis drives of my Bridgeport in a CNC conversion. I used some
fairly low-grade servo motors, because the weight of the machine components dwarfs the motor inertia.
I wasn't interested in 400 IPM rapid feeds, and getting the absolute last second out of a 5-minute part
program. I wanted to get about 120 IPM, but not sacrifice any accuracy at the low-speed end.
It took some doing, but I can go down to about .01 IPM before stick-slip friction on the table appears.
Currently, I can go to at least 105 IPM before I hit the voltage limitation on the raw DC power supply.
Using the EMC program to control the machine with a 1000 Hz servo update rate, I get about .005" following
error up to 105 IPM, including the acceleration and deceleration part.

My servo amps use a DC tachometer for velocity feedback. There are ways to eliminate this, but
I knew this approach would work. I'm using 1/8 Hp (continuous) servo motors, rated at 1.54A at 75 Volts.
I allow the armature current to peak at 5 A. The input to the servo amp is from the +/- 10 Volt DACs on
the Servo-to-Go card.

You can see some details of my system at :
http://ascc.artsci.wustl.edu/~jmelson/CNCconv.html

Jon