Z-axis motor

[Rudy from Norway]

Hi,
I want to improve the Z-axis on my ZX32 mill. Not a CNC conversion, but it will probably require the same considerations as if I did. I will of cause make it compatible with making it a CNC if I go to the next step.
The reason I want to do a motor driven Z-axix is that I want a more rigid and stable Z-axix feed. The quill on this machine is not the best in the business. And the depth control is somewhat "strange". With a stepper motor on Z, I can tighten up the quill for good. At least that is my plan.

I have read threads here on the subject. Thanks everyone for your contributions. So much to learn from you.
I have figured out that I should probably use a 25mm ball screw wit 5mm pitch. That is called a 2505 screw I believe. Will I want a double nut on this?

The motor some use on the ZX45 mill is NEMA42 size with some 20Nm torqe. I don't know if that is overkill for my ZX32. Maybe a NEMA34 with 12Nm torqe is suficient? I plan to put on some gas struts to lighten up the share weight of the head.

Guess the electronics is easy. DM860N?
I also need a hand controller for this.

Appreciate some advise on this.

Rudy

 

[RM-MN]

I found little information on how to size the stepper motors for any axis as I attempt to CNC my SC 2 mini mill. I finally just bought some NEMA 23 with 650 mNM holding power that require a maximum of 2 amps. Fastening one directly to the X axis I tried milling aluminum and found that it had sufficient torque for my purposes. I do intend to use timing pulleys and belt to get a 2 to one torque amplification for the Z axis. I'll be sacrificing speed for increased torque. Hope this helps.

 

[Rudy from Norway]

Thanks. I guess I'm into a lot more torque. Pulling X and Y on a smaller machine doesn't have to lift heavy loads, just overcome friction and tool pressure. The head on my medium size mill weights maybe 30 kg. And the rails have to be relatively tight.
Thats why I'm brought to believe I need 12-20 Nm. That's the gap between NEMA 34 and 42 I believe.
Rudy

 

[blighty]

For comparison.

I have an RF45. on the Z there is a NEMA 34 1200oz stepper running on 61.5v. ball screw is C5 2205 . ball nut has over sized balls.
when it had dove tails on the Z the head was counterbalanced with about 30Kg of weights, rapids of 3500mm/sec. I now have linear rails
on the Z no counterbalance . Now set to 4500mm/sec.

 

[Rudy from Norway]

blighty, so if your bigger machine handles the Z with a 1200 oz/in (8,5Nm) motor, mine will certainly do with the 12Nm NEMA 34 I suggested to my selves. I will also put on gas struts anyway to neutralize most of the head weight.
Thanks for the info!
Rudy

 

[ddmckee54]

I was watching some of the stuff on YouTube that CLOUGH42 has been doing to his Grizzly CNC conversion, which I think is about the same size as your machine. He has upgraded his CNC conversion from the original NEMA 23 motor on the Z axis to a NEMA 34 and is quite happy with it. Apparently he was never really happy with the Z axis rapid travel moves. I forget what kind of Z axis speed he was getting out of it after the motor upgrade, but it was impressive.

I think your choice of the NEMA 34 motor over the NEMA 23 will probably be a wise move, and one that you'd probably want to do in the future if you went with the NEMA 23 stepper.

Don

 

[Rudy from Norway]

Thanks Don. I watched those videos. He's machine is smaller than mine, but I'm more confident with NEMA 34 size so far. I will also get a closed loop type in case I will go for a full CNC conversio som time later.

So far I have figured out that I need a controller, power and a manual hand controller for this manual Z axis motor system?
I see some buy from fasttobuy2012. I sendt them a mail, but no reply. Not a good sign...

Rudy

 

[RM-MN]

blighty, so if your bigger machine handles the Z with a 1200 oz/in (8,5Nm) motor, mine will certainly do with the 12Nm NEMA 34 I suggested to my selves. I will also put on gas struts anyway to neutralize most of the head weight.
Thanks for the info!
Rudy

Unless you are in a hurry to get the machine set up you might wait to install the gas struts. I read someone's post that said his setup worked better without the gas struts. You could decide to add them later if you wanted.

 

[Rudy from Norway]

RM-NM, I'm in no hurry. I assumed the struts could do nothing but good, making the head lighter. However, thanks for the tip. I can easily do afterwards it if needed.
Rudy

 

[Nibby2226]

Hi,
I used a NEMA34 motor with 2505 ballscrew on my BF30. The head is counter balanced to retain 20 - 30kg of downward weight. I think the motor is 12Nm running at 42V with a Gecko driver. The Z speed is limited to 20in/min to give chance catching mistakes before they cost a tool or a part The ballscrew nut is loaded oversized balls to limit backlash.
https://www.cnczone.com/forums/benchtop-machines/107594-cnc-cad.html

 

[ddmckee54]

You don't NEED closed loop motors. If your control system can handle the encoder feedback they will allow for greater positioning accuracy, i.e. compensating for missed steps. It sounds like you aren't going to a CNC control software yet, and are just motorizing your Z axis using a stepper motor telling it to run at a fixed speed until you tell it to stop.

If it were me, I wouldn't commit to a closed loop stepper immediately. I'd buy a cheap-ish NEMA 34 with about the same torque/holding power as the closed loop version you've got your eye on. That way if the NEMA 34 doesn't work out it won't hurt quite so bad.

Don

 

[Rudy from Norway]

Don, I'm trying to resist the urge going full CNC. I might breake down some day. And then, I would probably regret. Looks like the price difference isn't all that big.
I sit by the computer all day with electronic "CAD" at work and I really enjoy the old school manual machines on my spare time. Going home an do even more CAD, Well.. I don't know yet. However, as a tech nerd I would really want to build that CNC machine..
Motorize the Z axis is in first hand to get a more stable machine by locking the quill witch is the weakest link on mine.
Rudy

 

[Muzzer]

You don't NEED closed loop motors. If your control system can handle the encoder feedback they will allow for greater positioning accuracy, i.e. compensating for missed steps.
Don

Encoder feedback isn't designed to compensate for missing steps, so much as achieving fast and accurate positioning. If you are missing steps you have a serious problem with your system. How many controllers actually have inputs for linear encoders to be used for positional feedback - can you suggest a few examples? I'm not aware of (m)any in this market sector. They are also not very happy trying to compensate for backlash which is very nonlinear.

Closed loop steppers are not expensive and have many advantages over dumb, open loop setups. There is quite a choice out there these days - don't be put off.

 

[ahughes3214]

Rudy,

I’ve left some info before on my setup; sorry if some is repetitive. I switched to full motor control for the z-axis and do not use the quill at all for positioning on my PM30 mill. I did not setup for, nor do I intend to go to CNC. Manual milling is my interest.

I am currently running a Compumotor 83-135 Nema 34 stepper rated at 400 in-oz (2.83Nm). I do have a gas strut (McMaster-Carr 4138T630 / 100lbs) for the head. The gas strut not only offsets the head weight, it also provides some dampening effect. The head with carriage weighs 127lbs. Power to the stepper is 80VDC.

This combination moves the head easily up and down. For manual jogs (joystick) I am currently driving it at 35-40 inches per minute. For my automatic memory moves, I drive it a little faster. I use the memory points to raise the head quickly to change a tool, etc. and then lower back to a rough position (+/- 0.010) prior to final positioning for the cut. If I find some time, I can make a video and try to post.

The ballscrew is imperative for this performance. I use the RM1605 ballscrew which is more than sufficient in this application as it is rated for about 1500 lbs. dynamic load. Direct drive to the leadscrew is preferred to keep it simple and limit potential slop in the mechanism. I did not change the standard ballscrew or use a double nut. As a result I have about 0.007 backlash which is not a problem for me and manual milling. I rely on my DRO for all accurate and final positioning.

Depending upon your method of motor control, if you plan for program/memory type movements, I recommend a closed loop system. I have been using an open loop controller, and if steps are missed (for example gib accidentally left tight when a move is initiated), the motor will stall and position is lost without the stepper controller knowing. As a result, the next move will not be accurate and a potential crash hazard.

After experiencing the above issue, I decided to go to closed loop controller so my control system will know when a step is missed and stop the motor drive signal, or perform some other logic. I just ordered a closed loop stepper and controller combo (HSS86 hybrid servo driver & nema34 Stepper Motor Closed loop 4.5N.m 82mm 6A) and will be experimenting with this setup over the next month or so.

I’d be happy to provide an update once I get it setup and running.

Drew

 

[Rudy from Norway]

Drew, Thaks for the helpful information. I'm starting to get a picture in my mind of how to do this.
I would appreciate to follow your updates on yous setup.
What are you using for joging and memory?
Rudy

 

[ahughes3214]

Rudy,

I have changed out my original open loop Pacific Scientific stepper controller for a JSS HSS86 controller. I did this so I could get an alarm (ALM) signal if the head stalled (missed steps). This new controller is a hybrid drive (closed loop). I really needed to know if steps were missed so the movement command shuts down and more importantly subsequent memory moves do not crash the head. I learned this the hard way as I initiated a memory move with the gib locked, lost steps, and then made a subsequent move which ruined an end mill and marked my vise. Now if the stepper motor misses steps, the stepper controller sends an alarm (ALM) signal to my controller and I reset my memory settings to null so the head crash is not possible. It is just a matter of rehoming and then I am back up to working.

If you go with a closed loop stepper controller like the HSS86 or similar, I do not recommend the JSS that I purchased. It works well enough but lacks the ability to program the parameters. I made a bad assumption when I saw the 232 port; on the JSS, the port is just for initial programming. There is a very similar unit which I think would be better; the JMC HSS86 which looks to be programmable based upon the documents I have seen. With regards to the programming, I would have liked to make the alarm output (ALM) normally closed. It is normally open from the factory. I can work with that but is not my preference. The only downside to the normally open alarm is that if power is lost to the stepper controller and movement stops, you only get the blip on the ALM output when power is reapplied. It works, but I would prefer the ALM error as soon as the power is lost, not upon reapplication of power. I setup my controller so that a power loss to the stepper controller also resets the memory positions to null and forces a rehoming. this is necessary because steps could be missed in this scenario as well.

As part of the swap, I also needed to replace the stepper motor to a closed loop (encoder output) motor. I bought this motor with the stepper controller for cheap (about $166 for both w/shipping). The motor is 4.5Nm which I have found is more than sufficient for my application. I can only stall it by hanging my 210lbs on the head PRIOR to movement. Once it is moving, there is no stopping it with my weight, it will lift me.

For jogging and memory, I have a custom unit that has a teensy 3.5 microcontroller at its core. It is running code that I programmed using the Arduino IDE. I can move the head with a joystick or rotary encoder for fine movements, two memory positions, and a boring function.

Thanks,
Drew

 

[Rudy from Norway]

Thanks Drew, for the update. I have a good chance to select the right parts now. Surprised that the 4.5Nm motor is so powerful. I have never owned such a motor, but from reading around it looks like people are using a relatively big motor on the Z. My mill is a bitt larger than yours though. It weights about 180kg/400lb.
I'm not into programming so I will have to get some kind of a controller.
Thankful for your help.
Rudy

 

[ahughes3214]

Rudy,

A couple clarifications;

My mill is about 500lbs. Just the head and carriage is 126 lbs .

The motor size requirement is very dependent upon the application. As I am using it for just jogging up and down, and do not need to accelerate aggressively, this motor is plenty sufficient.

When you get into the applications for CNC, etc. where very quick high acceleration and precise movements are needed, a larger stepper motor may certainly be warranted. If you want CNC in your future, you may need to consider a larger motor or replace later.

When you get to setting up your controller, if you have some questions, post them and maybe I can help.


Thanks,
Drew

 

[user 40653]

I never plan to go full CNC but power Z is something I plan on. Either a PWM controller with a worm drive or a stepper via an Arduino to allow for some fancy stuff like 0.1mm jog, fast and slow traverse with feed rates for boring etc. With the aim of keeping things compact and low power I have been looking at nema17 planetary gearboxes. I don't think the backlash in them makes cnc viable but it might be an idea that helps.

 

[Rudy from Norway]

Drew, I will certainly keep you in mind when I start this. Today the start of a 4 week holly day is the main focus! It feels good to have something to look forward to when the holly days are over.
I might go the whole way to CNC one day. That's why I wanted to do this Z axis motor compatible for the future without having to go back, once again!.
You see, I did buy a geared speed controlled AC motor for this. Total failure and wasted money. It wasn't nearly as powerful as needed to lift the head. I connected it to the crank. Think I put up something about it here somewhere. Micatroid, I suspect a PWM motor will have problems. However, a 45Nm battery drill is doing the job for now. And it is not over powered. Probably a good solution to just having faster and easier Z operation, but it doesen't give us a fine feed when the mill is running.
As I mentioned above, I want to lock the quill for better stability and do the fine feed with the head.
Rudy