DV-59 Hardinge Lathe going CNC

I started under another thread. A member gave a good suggestion to start a new thread, so here is the first posting.

Hi!
I am new to this forum.
A few details first:
Location: Montreal, Canada
Objective: To learn and have the basic insight of CNC
Project: To convert a Hardinge DV 59 manual lathe to CNC lathe
Material to machine: Stainless Steel 316
Size of Components: About 75% of the items will be 1” dia or less with a max. length of 2” and the remaining 25% can go up to 2.5” dia and 4” in length.

I propose to complete this project in two phases

Phase I
I will keep the head and spindle of the DV 59 lathe as it is, driven by 1 HP induction motor.
The bed of this lathe is having a longitudinal slot in the middle of the bed and chamfers through out the length.
The existing Hardinge cross slide can be moved and fixed on the bed of lathe. Due to the size of this slide, I do not like to modify this slide but intend to proceed as follows:

1. To purchase an imported heavy duty mill & drill table (sold on ebay item # 7578353458) The table has11” longitudinal and 7.5” cross travel
2. Remove both the lead screws and replace with rolled ball screws and two nuts on each screw. The screw size will be ¾” or 5/8” dia depending upon the available space to mount nuts.
3. The guide ways of the cross table slides will be scraped and improved upon as required, to a reasonable condition.
4. Stepper motors will be mounted on both axis of the table, directly coupled to the ball screws if space permits, or else through timing belt and pulleys. The torque of motors will be decided depending on the condition of table.
5. The table is to be mounted on the lathe bed and fixed to the bed so that the cross slide - 7.5” travel of the table is Z axis of the lathe and the longitudinal slide -11” travel is X axis.
6. The above arrangement will give sufficient place on the longitudinal slide of the table to mount three or four tools and possibility of mounting a live tool also.
7. Initially the setup will be tried for drilling, boring, turning and facing operations only, with the spindle running with the existing arrangement of induction motor drive.
8. The spindle axis is at 4.5” height from the top face of the bed whereas cross table height is going to be 5” when fixed on bed. So for initial trials the tools will have to be mounted in special tool holders to match the spindle axis. Alternately depending on the dimensions of the longitudinal table, holes can be drilled and reamed in the side of the table so that the tool mounting arrangement is similar to the Hardinge turret tool mounting system.

Phase II
On successful completion and testing of the above, I will proceed to raise the height of the spindle head by providing spacer on the bed and the spindle will be driven by a servo motor of suitable power. This will enable to cut threads and provide full CNC capability.

Thanks,

RRossey

The mill and drill table is expected in two or three days. I will open it up to measure the available space for mounting ball screw and nuts.

A concept drawing is attached to show how table will be mounted on the lathe bed.
The spindle head will be raised by providing a spacer above the bed.

The imported mill and drill table purchased from ebaystore.

Loooooking good rrossey :)

I see what you were referring to on the slot in the bed, that should come in very handy, is it wide / deep enough to accommodate the screw ?

There are three or four of us doing similar conversions with enough variations for plenty of ideas for others who are considering this, so I am very happy to see you start your own thread/conversion log.

Ken

I built a small lathe using a precision x-y table and stepper motors that I use every day. I primarily turn wood with it so I left the leadscrews in to see how well it would work - good enough for what I need. Someday I will replace them with ballscrews. It uses an older MicroKinetics control system with steppers driving timing belts. The tool changer is a manual turret off of an old Logan 9" lathe that I have. The headstock is from an old photo engraving lathe. I use a 1-1/2 hp 3 phase motor and a VFD. A real pile of parts :) works great. All this to say... Go for it!

I tried to include pictures in the last post.
try again.
Not shown is the tailstock or dust control enclosure.
dave

Nice job DJSHOP.
What is the torque of stepper motors?

Ken Shea
The slot width is 7/16" and 22" long, it goes through the full depth of bed between the two supports of the bed. It can not accomodate the screw, but is good for bolting the cross slide table. The position of the table can be shifted/adjusted as per the job length, if required.

I received the cross slide table. I can fix 5/8" ball screw but some machining will be required to accomodate the rectangular ball nuts.

The steppers are 1125 oz-in motors from Microkinetics (www.microkinetics.com) I found the complete retrofit system on eBay which included the Microkinetics controller, computer+software, two stepper motors, some mounting hardware and all of the cables. The z axis is lifting quite a bit of weight with the slide and tool turret so I am glad for the beefy motors.

The first view shows the cross table assy with new end plates fixed to the top table and blue coloured ball screw end projecting out. The ball screw will be 5/8" dia with sq. ball nuts.

In the second view, the top table is removed. The green coloured ball nuts are fixed to the red coloured block that is bolted to the cross slide.

Due to the limitation of space, the ball nuts must be fixed in this position. So the preloading has to be done to get this particular position, probably by providing shim washers, by trial and error.

Any suggestions?

The ball screw is mounted in the two end plates in two angular contact ball bearings # 7202: 15mm x 35mm x 11mm. Again the shims and spacers are to be provided to eliminate back-lash.
Another alternative can be to provide both the angular contact bearings on one side in a thicker housing, that is the motor mounting side, eliminate the back-lash and mount a plain roller ball bearing on the other side.

Which is the better way?

Looks good. shims will work. One suggestion: only "fix" one end of the ball screw with your angular contact bearings. The other end should just be supported by a plain bearing and allowed to "float" a bit along the length of the screw. This is because of possible changes in rates of thermal expansion.
dave

rrossey-
Nice Lathe! I have the exact same model, except mine has a turret tailstock.

Regarding the preloading of the ballnuts, you can use the thread of the screw to your advantage. Solidly mount the nut closest to the motor, just as you have shown. Next, mount a block to the rear ballnut, but allow it to freely spin with the nut up and down the screw. When you turn the second nut and block toward the solidly mounted nut, it will contact the red mounting block. Now, it you make the block that is attached to the free ballnut spring loaded with a torsion spring, it will hold the "free" ballnut against the red block. This is the same effect as having two nuts on a piece of threaded rod. If you tighten them against each other, they each press against a different side of the thread. This is the same effect, just fine tune the torque, and there will be a constant, automatically adjusting preload on it.
http://www.kerkmotion.com/learn/kerktech/anti-backlash-overview.asp

As far as the bearings to support the ballscrew, you should have a preloaded pair at one end, and a radial bearing at the other. If you were to put one angular contact bearing at each end, it would be tough to remove the backlash. Also, as the screw heats up from continious running, it will expand, and possibly bind the bearings at the ends. Using one pair at one end and a radial bearing at the other allows the radial bearing end to expand axially.

What software are you using? Could I get a copy of your model? Perhaps a IGES or Step?

Do you have a collet closer for your lathe?

This is what mine looks like...

Neatman

djshop/NEATman
Thanks for the information and suggestions.

NEATman, my lathe is similar to yours exept the turret. I do not have a collet closer. I am looking for one at a competitive price. The shipping charges to Canada sometimes exceed the component cost.

I am modifying the design to accomodate both angular conact bearings in motor side plate and placing a plain ball bearing on the other side of the screw.

Preloading the ball nut with a compression spring is good idea. I will work on it.

I am using solidworks for the design and am in the learning phase only. I think the model can be saved in IGES. I will confirm.

I am considering stepper motors, 1200 oz-in directly coupled OR 640 oz-in through 1:2 pulley drive. Which one will be better?

rrossey-
Contact me privately about the collet closer. I also have a bunch of spare 5C collets that I would be willing to part with. Where in Canada are you located?

I guess I did not explain myself properly. I was trying to explain that the spring would be a torsion (twisting) spring, not a compression spring. One nut would be solid, and the other would be free, except it would be twisted against the nut mount by a spring. The torqe provided by the spring keeps the free nut seated against the solid nut mount, and that resulting thrust provides the preload. Perhaps there should even be a thrust bearing between the nuts to prevent binding. This is an idea that is based on the Kerk screw patent, but slightly different, as it uses ballscrews, and the thread of the ballscrew itself. I have been meaning to try it on a small manual milling machine that I own, but the 50 year old acme screws are still do the job well enough. We use hundreds of these exact same non-preloaded ballscrews where I work, and we have quite a solidworks model library. I will try to make a model of the dual ballnut assembly, with the torsion spring preload.

I'm lucky enough to have a plentiful dumpster, connections with the service department and scrap pile at work, so I happen to have internally preloaded ballscrews and THK rails that I plan on using on my lathe. I like your idea about using the XY table though. Very slick!

I am using solidworks as well. If you have already taken apart your headstock, and have an accurate model of the riser block, I am very interested in that model. But a nice model of the entire bed would save me some time too. If you like, contact me privately so I can give you my e-mail without getting spammed.

I am also planning on using steppers. I am going to go with a direct coupling, and my steppers are about 1000 oz-in. Either of your options should work well, direct drive or 2:1 reduction. It all depends on your budget, but higher torque steppers and direct coupling vs. smaller motors, belts, pulleys and mounting may cost almost the same.

I believe you also mentioned that phase II would include a servo motor to drive the spindle. I happen to have a few large brushless DC servo motors with internal encoders that could do the trick. I haven't had the time to get them on ebay yet, so if one of these would work for you, let me know. I was planning on keeping one for a later phase two on my lathe as well. I was planning on removing the dual V belts, and using a toothed belt to ensure that there is no slipping between the motor and spindle. I have been looking at straight toothed belts, but that may transmit too much vibration, so something like this goodyear belt may be a good alternative.
http://www.goodyearindustrialproducts.com/polyurethanebelts/index.html

I really love seeing new uses for old machines. You probbably can't get a much better spindle than a hardinge!

Neatman

For pre loading the ball nut, I got this solution: I am shifting the position of the block on which the ball nuts are screwed. As shown in the attached drawing one ball nut will have space to get positioned where ever the preload is reached by tightening against the block.

I have not opened the head. I will be doing that only on completion and trial of the cross slide table.

I will go for 1200 oz-in steppers directly coupled to the ball screws. I hope this will take care of increased friction due to machined guides as compared to the guide rails and ball slides.

1000+ oz in torque step motors? Those seem incredibly large for a lathe this size. I'm doing a 9x30 lathe and most of the conversions I've seen are sporting motors in the 200 oz in range and seem perfectly happy with their performance. I have a 400 oz in for the leadscrew and a 210 oz in for the cross slide on my conversion.

With ballscrews you will eliminate an awful lot of the friction in your machines. I read one account where a guy's rapids rates went from 30 ipm to 110 ipm when he converted to ballscrews. One thing to be very cognizant of is the best torque rpm range for your motors versus the pitch on your screws. Somewhere I read a Mariss Freimanis quote that said max torque for steppers is 900-1200 rpm. Check on your drive ratios and screw pitches to see what that would give you for rapids.

No harm done in providing some "headroom" with more powerful motors, I'm just curious what advantages you perceive for such large motors.

Best,

BW

Thanks for this thought provoking question.

I hope some members can comment on the torque requirement based on their experience, having successfully completed their projects with stepper motors.

I am going to use stepper motors for the first time. I talked to a few suppliers and the following is my understanding:

1) The torque mentioned for a particular motor is the max stalling torque. This should be 30% to 40% higher than the actual requirement.
2) When motor is in micro stepping mode the torque further reduces in inverse proportion to the step rate.

I considered this torque after going through the threads regarding conversion of mill/Drill to CNC; many have used 1200 oz-in motors.
I will be mainly machining stainless steel 316.

These may be over size motors; perhaps it is just playing safe. I am open to suggestions and still have time to finalize the design.

The reason I have gone for a cross table on the hardinge lathe is that I can avoid tool changer or the tail stock or the turret as this increases one more axis to be controlled.

Multiple tools can be mounted on the top slide of the cross table, somewhat similar to the tooling of screw machines. there will be enough space to mount live tool, as I am thinking of grinding a cam profile on hardened material.



I do not propose to raise the lathe head initially. I will mount offset head tools adjusted to the spindle axis height for trials.



Latter on raising the head may provide space to mount the spindle drive motor in the spacer itself.

BobWarfield - I Worked on the torque requirement and finally decided on 640 oz inch motors.

The following items are ordered;
640 oz inch stepper motors
Gecko 202 drivers
PMDX 131 Breakout Board
5/8" ball screws with square ball nuts

The machining of cross slide for fixing ball nuts is in progress.

The lathe after assembly will take shape as shown below.

Hey rrossey, nice work with the models.
Just for comparison, my CNC Router (gantry style) is running 640-oz steppers, and that’s with a 60” gantry supporting a 10-hp spindle and performs very well.
One observation on your lathe design. Is it necessary to mount the cross slide stepper on the end were you will be working/loading parts from? It looks like more of a reach and one more opportunity for a “hip-check” while running it.

yeah
I'd move it to the back

This is a good idea. I will shift the motor on the back side.

How important it is to have hand wheels on the slides for manual movement?
Once in a while, the hand wheels can help to have a good feel while adjusting the gibs.

For rolled groove ball screw, purchased from McMaster, screw lead accuracy mentioned is .004" per ft.

Is it correct that preloading will eliminate only backlash in ball screw and ball nut?
Even after preload the maximum lead error will be .004" per ft.

Hello! CNC Guru's

Say something: yes, no or partialy true.

The power supply and drive components placement in an old computer enclosure is shown in the attached photo.

Two Gecko 202 drives are mounted on a 1/8” thick Aluminum plate and then PMDX -131 Breakout board is mounted on top. In the enclosure the board will be vertical as shown.

Stepper motors are 640 oz-inch from HomeShopCNC.

I will appreciate if somebody can provide the following information:

1) How to calculate the motor current set resistor value for Gecko Drive?

2) In the power supply diagram provided by member Cyclone, a capacitor of 470 micro farad is shown across power supply to each gecko driver. As you see the PMDX board is mounted directly on Gecko driver and a capacitor is mounted horizontally above each drive, is this capacitor serving the same function?

3) Similarly, are the fuses shown in front of the capacitors are for each Gecko drive?

4) How can I provide power switch and power indicator lamp for each drive?

5) In another post Mr. Jerry suggested to connect a bleeder resistor of 5,000 ohm, 5 watt across the capacitor. Is this one resistor required in parallel to the main power supply capacitor?

Any suggestions please?

RRossey,

1. CURRENT SET RESISTOR:
The following formula comes from the G202 Manual. Use the following equation to calculate the value, (in kilo-ohms) of the current set resistor:
R (in kilo-ohms) = 47 * I / (7 – I) or for the low current range, R = 47 * I / (2 – I)
2. Yes, but not required on the G202.

3. Yes, may be replaced by lower amperage but not more than 5 amp fuses.

4. Can't answer this one for you.

5. Yes

I found the answer to 1 in the G202 manual you can download it from Gecko.com. I found the answers to 2, 3, and 5 in the PMDX-131 manual. You can download it from PMDX.com.

Alan

Alan,
Thanks for the information.

I just recently put a ballscrew cross-slide on my 7x10, similar to what you're doing and for the same reasons - I needed more travel for gang tooling.

I'm using the 250 oz nema 23's for steppers, 1:1 on a 10 TPI ballscrew, geckos for drives and THK RSR20 for the slides. This gives me 200 ipm for the cross slide, which is quite snappy. I think it could go faster still, but 50 kHz is as much as I can get out of TCNC with the computer I'm using.

I use one nut and preload, which currently uses a cable arrangment with a counterweight, eventually plan to use a pneumatic cylinder and a regulator. There is lead error in the ballscrew per spec, but I don't see it in the parts, whereas the old ACME leadscrew had lead error galore.

dkowalcz, I am pleased to note your experience.
It confirms that as per specs, max lead error is specified but one can be lucky to have a small length of screw with a negligible lead error.

As there are no comments regarding the preload issue, I conclude that preloading can eliminate only the backlash due to play between ball screw and ball nut.

the lead is fixed; pre-load will not change the lead; if it's 0.004/ft that will stay
Now if you run Mach3 you can screw map; hence remove the errors if there are parts of the screw significantly off pitch
If the lead is basically the same error all the way along; adjust the steps to compensate; you should be able to get close to Zero error; however the backlash will remain
I just loaded a 48" ball screw 0.500 pitch 2 start 0.500 diameter with 0.126" diameter balls; fixed the slop right up; feels great the whole length (I use a cordless to spin the screw)

Stevie,
Thanks for the information. It is good to know that Mach 3 can take care of the significant lead errors.

Another point regarding the handwheels; You have built very impressive machines and I observed that you are not installing hand wheels on them.

Is it that when a conversion is done from a manual machine to CNC; because the hand wheels are available, we mount these on the machine, or, we are used to manually operating the machine so we like to retain the feature?

might be
I really have no need to manually adjust anything; the jog in Mach3 is just fine

I switch off the drives to set limit switches etc while building; but after that I use jog

I am installing a collet closer on the lathe as shown.

How the collet closing pressure is applied in the attachment? Is there any spring is to be provided?

I will appreciate if somebody can provide a x-sectional drawing of the attachment.