[duducati]

Dear forum members,

This is my first post and I hope I am doing it right.

I came across a Yamaha cartesian robot. (My company was going to scrap it)
It has very nice linear rails and brushless servo drive which is connected directly to the ball screw.
Now, my question is if you could help me use it for my DIY CNC router project I have in mind?

The problem is that I don't know if there is any easy way to control the AC servo motor with the build in encoder using a modern hobbyist grade CNC controller.
I will upload pictures of the cartaisian robot later. I found the following information on the labels:

- Tamagawa BRX Smartsyn, TS2620N654E122
- TBL-i N 112, E200
- The robot controller is Yamaha DRCX: https://www.yrginc.com/Catalog/PDF/C...CX_E_V6.08.pdf
- Yamaha DRCX

My plan for the DIY router is:
- work area around 40"x30"x8"
- cheap (~1000$) but somewhat rigid enough to work with Aluminum
- I'm planning on using 4x8cm aluminum extrusions that I have laying around but everything else still needs to be obtained including the controller.

I thought I could use one axis of the robot for the Z-Axis.

I also thought about just using the rails and the ball screw but the screw is also the drive shaft of the servo motor at the same and the thrust bearing for eliminating backlash is mounted directly on the encoder.
The motor mount is also not compatible to the Nema motors because the holes are about 50mm apart.
So it is probably difficult to use it with conventional stepper motors or am I wrong?

I would really appreciate it if someone could tell me an easy and cheap way to use it.

Thank you!

[duducati]

Here are the pictures:

[Jim Dawson]

It would be pretty easy to mount a standard NEMA frame motor with an adapter plate or just redrill the holes.

[duducati]

Thanks Jim. You are right but do you know any easy way to remove the axial play of the ball screw if I follow your suggestion?
The opposite end of the ball screw (not the motor side) is not threaded and only fixed radially by some bearings.
The thrust bearing is mounted on the encoder as you can see in the picture.
Thank you.

[spumco]

Main problem with the existing motor: you don't know the specs (voltage, amps, etc). And I couldn't find anything on the TBL-i N series in 10 minutes or less.

You could get an agnostic servo drive and use the existing motor. The "encoder" appears to be a Tamagawa resolver, not a standard quad a/b encoder, so you'll need a resolver-input drive. Ebay has plenty of drives like that.

Example: there's a Parker VIX500IE for $95 on ebay right now that takes resolver inputs and has 5A continuous. Uses an 80VDC bus so you'd need a power supply, but you could use the same PS for steppers or low-voltage servos.

Once you have a drive for the existing motor, any 'hobby grade' control system should be able to provide step & direction signals to the drive.

No machining option: perhaps a sleeve on the shaft to take up the distance between the tension nut and the AC bearing in that housing. Fit a timing pulley on the OD of the sleeve, and mount an off-set stepper or servo and belt-drive it. Motor mount could use existing 50mm hole pattern on the bearing block.

[Jim Dawson]

It looks like the thrust bearing is in the large aluminum housing that the motor bolts to, looks like the thrust bearing is preloaded with the nut on the end of the shaft. In order to make it all work with a new motor, some machining would be required, but simple if you have access to machine tools.

What spumco said makes sense also.

[duducati]

Spumco, Jim, I appreciate your inputs.
If I won't be able to find an easy way to use the motor, I would certainly try to adopt a NEMA stepper motor.
So far, I am seriously considering the Parker solution.

Today I tried to get some information through the customer service of Tamagawa Germany but without any success.
I found some information about the BRX resolver: http://images.100y.com.tw/pdf_file/79-TAMAGAWA.pdf
I can't judge if the resolver is something that is supported by many drives.
How could you tell if it's compatible with the resolver?
The motor seems to run at 200V AC (based on "E200" designation). That's not surprising because 230V is the standard voltage in Europe.
I could not find out if the mentioned Parker controller can run such motor. Does anyone know it?
I am also still wondering if I can use the Yamaha DRCX controller.
If someone knows how to use it with step and direction signals, please let me know.

I will also try to contact other Tamagawa offices to get more info on the motor. How much torque do I need for the Z-axis in order to be able to do light cuts on Aluminum? The pitch of the ball screw is 5mm.
Would the belt drive solution significantly affect the CNC's capability by introducing increased axial play?

Thank you

[Jim Dawson]

I don't know about the resolver. Normally it is best to pair the motor with the drive that it is designed to run with. It is possible to mix drives and motors in some instances, but it is normally not an easy task to get mixed solutions to work correctly.

A very quick review of the manual you linked to indicates that the controller would not accept step and direction input. In fact, that controller seems to be designed for PLC control of a robot. Not a good fit for a CNC cutting machine.

With a 5mm lead screw, very little torque is required to hold the position. Almost any NEMA 23 motor or greater would hold the Z position when cutting. There is not a lot of axial load on the ballscrew when cutting.

A belt drive should not introduce any axial play into the system. Many industrial CNC machines have a belt drive on the Z axis.

[spumco]

The Parker VIX I mentioned is for an 80VDC bus, not for that motor. It would run it as the VIX can take resolver inputs, but the maximum motor speed would probably be reduced quite a bit.

You need a 200V-class (240VAC 1-phase input) servo drive capable of resolver inputs. There are a few of those out there - Parker Gemini series, Copley Controls Xenus, Allen-Bradley Ultra 300. Baldor/ABB also make them as most of the servos are resolver-types, but they are not as 'universal.' Unfortunately (perhaps), resolver-capable servo drives are not as widely available on the secondary market.

If you're determined to use the existing motor you could remove the resolver and fit an encoder to it. Measure the shaft diameter and look for a hollow-shaft encoder with standard quadrature signals plus commutation (Halls). Then you can use any of the encoder-input agnostic drives.

A resolver-capable servo drive will cost you way more than $200 on ebay, not to mention the cost of cables. And an encoder-capable drive might be $150, but you'll probably spend that again (or more) on an encoder.

Or... about $250 would get you a power supply and a sweet Nema23 3-phase closed-loop stepper & drive combo from Automation Technologies, plus some pulleys and belts. I've used those steppers; they're fantastic and stupid easy to set up. And you probably won't need a brake for the Z-axis if you use a stepper.

Shaft sleeve, two timing pulleys, bracket, stepper & drive, power supply - done.

How far down the servo-to-drive compatibility rabbit hole do you want to go? You've got to match:
Feedback type (resolver, quad encoder, BISS, serial, proprietary)
Drive/motor voltage (DC in, AC in, 1/3phase, etc.)
Drive/motor amps
Command signal type (I still have a Parker GV6K drive I thought was a screaming deal until I found out it only takes software commands - no PWM or step & direction. Perfect for a factory floor machine, not so much for CNC.)
Connectors for I/O & feedback (Manufacturers all seem to pick different connectors when they build drives. Have fun building this-to-that cables with obsolete connectors. The bargain drive doesn't look so good when you shell out another $100 for cable ends - just the ends.)
Drive programming software (don't laugh, it's an issue. Parker's Gemini series are great drives, but you need WinXP or Win7 32-bit to run the ancient software).

Jim's advice, as usual, is spot on. It's way easier to pair a motor/drive combo from the same manufacturer AND series/family.

P.S. Maybe you could gut the motor and use the rotor (with the magnets on it) as your shaft sleeve to preload the bearings. Break the magnets off and attach the pulley to the outside - no custom shaft sleeve required. It'd look like hell, but who cares?

[spumco]

...but it is normally not an easy task to get mixed solutions to work correctly.

Understatement of the day.

[duducati]

Thank you so much for all the inputs. I was afraid it won't be as easy as I hoped.

Great to know that using belt drive is not a problem.
I have a manual lathe so fitting a sleeve is going to be easy.
However, I wish I already had a CNC to make the brackets to mount the nema motor.????

I will probably go with the cheap nema 23 stepper motors for all axes (open loop). I guess the 3Nm ones are still good enough for my first CNC build?

If you have any recommendation or tips, please let me know.

Cheers