Stepper Or Servo ???

[TPMX]

After much time spent on this Series II Boss 6 and getting nowhere
I have decided to do a retrofit.
I would like opinions from you guys that have done this on whether you are happy with conversions using the stock steppers or do you wish that you would have gone to servos.
Are the stock steppers powerful enough to stay on location and not loose steps or are you loosing location to some extent.
I have a 3 in 1 that I built a control for using Gecko's,steppers and Mach 3 and I will have instances where I loose location if I push the machine too much.I have a 3 axis readout on the machine that I zero out when I zero out the machine so I have something to alert me that i have lost location.
Without the readouts I would not know that the steppers were off.
This is what I am worried about if using the stock steppers,There isn,t a "backup" way to watch the machines location.
Maybe this isn't a problem with the steppers on these mills,thats why I am asking.
I am leaning heavy toward servos and geckos but the cost of servos is hard to swallow.
The best I can find for direct bolt ons is $600.00 a piece for 29 in-lb motors.
Does anyone know of a source for servos with a decent price?
I thought about machining mounts for NEMA 34 to NEMA 42 with a 2.1 or 3.1
reduction but would sooner have more powerful motors even if they were NEMA 34's and just machine an adapter plate for them.
I just wanted to know how different setups worked on theses machines.
I found the regulated power supply on my BOSS 6 has an issue with one of the supply voltages and am just concerned that if I can fix this problem there still may be something else wrong.
My steppers lock once i apply power and push the limit reset but thats as far as it will go.
Thanks too all for their feedback!

[Hood]

Have 850oz/in (2.5:1) steppers on my series 1 and never had a problem with missed steps, if I had a bigger travel then maybe would have gone with servos.
Not sure what the 29in-lb is that you are quoting but if its peak then have a look here http://www.kelinginc.net/ServoMotor.html

Hood

[vipers95]

Steppers are great if you want to cut something where the tolerance does not matter. They are old school. bite the bullet and go Servo.

[jdelaney44]

This seems to be the question of the ages. This does not seem to be so black and white which is why it seems to be often debated.

Assumed set up - Mach3 Control, Gecko G320 Drives, Servo Motors

Fact: Under too much acceleration, max ipm, load, tooling going away, etc., you will experience step loss and fault with servos using the Gecko G320 and Mach3 will not know about it unless you either:

a.) design a circuit to tie the Gecko err signal to the e-stop or.....

b.) get a board from Roger's Machine that acts as a DRO and feedback device to stop the software or.....

c.) go to the G-Rex set up from Gecko.

I have NO idea how the G-Rex is going and working with Mach3 at this time. It is more $$$ too.

Gecko is pushing the G-Rex to be a totally closed loop system as understand it. I think it's coming along. I'm not sure it's plug & play yet. But you may already know about that.

The nut to crack here is making the software aware of the step loss condition AND getting it to slow down the step rate as a result. I don't *think* Mach3 is able to do that *yet*. But I am admitedly out of step with current hapennings at Art Soft.

The Rogers board does allow you to set lower thresholds than the 128 steps that the G320 will fault at. 128 steps on a 3 to 1 gear ratio and a .20 pitch screw turns out to be enough to create problems. I think the Roger's board is a great, reasonably priced alternative to full blown feedback.

So I would say this:

Go with servos if you are going to go the extra mile and make the Geckos talk to Mach3 somehow.

Go with steppers if not. Tune the machine best you can. Over time you will find the right acceleration and max ipm.

Having said that......I am half pregnant as it where. I have the servo's w/o the feedback to Mach3. Having at least the fault at the Gecko's has been helpful in determining how fast I can go w/o step loss. The Gecko can see things I can't. So it is a useful tool at the very least.

If you are going to use a control that has the feedback built in then yeah, go servos.

Hope that helps.

-jd

[Hood]

JD
not really looked at the Geckos servo drives but is there no fault output on them? The Tek10 servo drives I have for the lathe conversion I am doing have this. If I input it to Mach through Input 1 then I can set that input to fault Mach out if the drives fault. Obviously with two or three drives you will need an and gate for this.
This still doesnt close the loop but from what I gather talking to Brian Barker the loop has been closed on a huge mill he has, but it is using a Galil.

Hood

[jdelaney44]

Hood, yes there is an fault terminal. It's called ERR/RES, it's pin 5. It goes from +5V to ground when the drive faults at +/- 128 steps out of whack. Then when you feed it +5V for 3 seconds, it resets. 3 seconds is more than enough to trip Mach3.

Interestingly there is also a position error test point that is supposed to drop or increase .04 volts for every step it's out. It would seem that this test point could be hacked to provide a closer error than 128 steps. Just thought of that actually.... there goes my warranty again....;-)

What I was trying to really communicate is the notion of the encoder step counting needing to make it back to the control software for a servo to actually be much better than a stepper. Also that some error triggering is better than nothing.

Question, do the NC Bridgeports have shaft encoders on the screw or something on the stepper rigs?

I think this gets confused sometimes too. Just because you are using a stepper motor does not mean you can't have feedback to some kind of control software / hardware.

Best,
-jd

[Hood]

JD my Bridgeport was/is steppers and no encoders although I did have a loan of Rons board for testing and had glass scales to work it.
The steppers when set up right are extremely accurate and although I have not yet had experience with servos I would say they are every bit as accurate if not inherintantly more so. Servos have the advantage of the drive faulting if a given error is seen (125 for Gecko, Tek10 has a few different) but as you say Rons board can add this function to Mach. The big advantage I see with servos is the velocity and acceleration.
Hood

[TPMX]

Thanks for all the input.
When I was converting my Smithy 3 in 1 over to cnc I was using EMC.
My hope was too come up with a true cloosed loop system.
My machine already had readouts on it with rotarty encodrs so a friend of mine designed a pci board that we hooked the encoders from the readout to but could not get it to work right.
The problem I think came from so much backlash in the acme screws that the software faulted because the axis in movement was too far off the location that the software was targeting.
It was then that I converted the machine over to ballscrews.
I then started using Mach 2 so I never went back to the original setup.
HAs anyone ever converted a dc motor over to a servo and had good luck with them?
I found some new 1/2hp PMDC motors that fall within the gecko guide lines that will bolt on the bridgeport.Just need the encoders bolted on.
Thought maybe I would try this route.