[probinson]

Hey Nikel,

My opinion on the Bridgeport CNC conversion is quite different. A CNC mill makes a lousy hand operated mill but a hand opperated mill converted to CNC does both quite well. If you have ever tried to do a quick machining job using the jog wheel on a CNC you know what I mean so be sure to have an easy way to disengage the CNC for hand operattion.

Although putting a ball screw on the quill is a real pain in the butt, the rest of the conversion is very simple.

The accuracy of the ball screw in meaningless. Your glass scales are already more accurate then a precision ball screw so just use the scale outputs instead of rotary encoders on the screws.

As long as you don't go with a really small ball screw, backlash is the only issue you need to worry about. You want double ball nuts on each screw so that you can pre-load them for zero backlash.

Pete

[nikel]

Thank you for the input. I was coming from this point of view myself. Although, very good info on the ball screws.

I took the other input to heart and went on a serious search of CNC mills. Depressing, most used equipment in the 10k budget is pretty old and would more than likely require too much work.

This all however brings up another question. I was planing on using AC servos, but I am now wondering if that might be simple over kill?

Hey Nikel,

My opinion on the Bridgeport CNC conversion is quite different. A CNC mill makes a lousy hand operated mill but a hand opperated mill converted to CNC does both quite well. If you have ever tried to do a quick machining job using the jog wheel on a CNC you know what I mean so be sure to have an easy way to disengage the CNC for hand operattion.

Although putting a ball screw on the quill is a real pain in the butt, the rest of the conversion is very simple.

The accuracy of the ball screw in meaningless. Your glass scales are already more accurate then a precision ball screw so just use the scale outputs instead of rotary encoders on the screws.

As long as you don't go with a really small ball screw, backlash is the only issue you need to worry about. You want double ball nuts on each screw so that you can pre-load them for zero backlash.

Pete

[probinson]

The difference between AC and DC servos is just electronic commutation. Some power is lost in brushes, they wear out, and they take up space in the motor which means that an all electronic motor can be a more powerful than the same size DC motor and be virtually maintenance free.

AC servos are cheaper to build but they are newer technology, higher power density, and maintenance free so they can fetch a higher price. Your best bang for the buck, however, will probably be a DC servo system, especially if you are shopping the surplus market.

If you were building a CNC that had to run 24/7, needed maximum power in a small space, or needed really massive motors then AC servos would be the only way to go. For hobby or small business applications, the bushes will never need replacing and you won't see any difference in performance between the two. You are shopping for power per dollar in your servo system, not power per cubic inch or bragging rights.

I think I went over my 2 cents worth. Oh well, inflation..

Pete

[ichudov]

The difference between AC and DC servos is just electronic commutation. Some power is lost in brushes, they wear out, and they take up space in the motor which means that an all electronic motor can be a more powerful than the same size DC motor and be virtually maintenance free.

AC servos are cheaper to build but they are newer technology, higher power density, and maintenance free so they can fetch a higher price. Your best bang for the buck, however, will probably be a DC servo system, especially if you are shopping the surplus market.

If you were building a CNC that had to run 24/7, needed maximum power in a small space, or needed really massive motors then AC servos would be the only way to go. For hobby or small business applications, the bushes will never need replacing and you won't see any difference in performance between the two. You are shopping for power per dollar in your servo system, not power per cubic inch or bragging rights.

I think I went over my 2 cents worth. Oh well, inflation..

Pete

Another advantage of AC servos is that current goes into the stator (outside), not the rotor (inside), and so AC servos are easier to cool.

[probinson]

Another advantage of AC servos is that current goes into the stator (outside), not the rotor (inside), and so AC servos are easier to cool.

Yes, thank you. That is definitely the most important selling point for the AC servo and I should not have left that out.

For an application where the machine does a lot of rapid moves, motor heating is a critical issue. With the Bridgeport conversion, however, we are loading the work piece manually and doing manual tool changes so, unless your motors are seriously undersized, motor heating won't be an issue.

[nikel]

@probinson

This concept of using the dro scales got me doing some research. Thus far I have not found much information, which leads me to think I am not searching for the right subject. Do you (anyone really) know where I could find more information on how this kind of setup would be wired?

If I go the DC route I have some Gecko drivers that might be used: G320's I think. If I go the AC route I will purchase a driver/servo set.

[nikel]

For anyone else interested; I just found a nice tutorial on dual closed loop & improved dual closed loop systems here: http://www.galilmc.com/checkreg.php?...sation-methods

@probinson

This concept of using the dro scales got me doing some research. Thus far I have not found much information, which leads me to think I am not searching for the right subject. Do you (anyone really) know where I could find more information on how this kind of setup would be wired?

If I go the DC route I have some Gecko drivers that might be used: G320's I think. If I go the AC route I will purchase a driver/servo set.

[probinson]

@probinson

This concept of using the dro scales got me doing some research. Thus far I have not found much information, which leads me to think I am not searching for the right subject. Do you (anyone really) know where I could find more information on how this kind of setup would be wired?

If I go the DC route I have some Gecko drivers that might be used: G320's I think. If I go the AC route I will purchase a driver/servo set.

The scales on your DRO generate the same 2 phase signal as the rotary encoders you would normally be using (there are sine/cosine output encoders that require interpolation to generate the digital phase pules but it is unlikely you will be dealing with that type of encoder). If you have ZERO backlash, then you can simply connect the scale outputs to the servo encoder input. If you have any backlash at all, however, this will produce chatter.

Think about how the machine acts when you dial in by hand. The DRO reads smoothly as you dial into your desired location but as soon as you need to reverse direction the handle turns freely until you take up the slack and then it bumps the carriage, you overshoot in the other direction, and usually repeat this several times before that analytical engine between your ears modifies your actions and you finally settle in the the exact location you wanted. The servo driver does not have such adaptable logic and would just bump the carriage back and forth indefinitely. This is where the dual loop servo system comes in. Using either a tachometer or encoder directly on the motor shaft, the servo system can detect when the motor is moving but not driving the carriage and will limit the velocity so the motor doesn't bump the carriage when the backlash is taken up.

Things are a little simpler if you use stepper motors (or servo drives that take step inputs) along with seperate feedback because the DRO scales (feedback input) don't affect the stability of the loop. In this case, the controller always moves the motor to a calculated position. The control system only uses the feedback from the scales to correct the error between calculated position and the actual position.

Pete