[robhrzic]

Hello all!

I have a large 10 x 54 vertical milling machine, Grizzly Model # G9905 that I would like to convert over to CNC. I bought it new last year, and installed a Newall C80 3-axis (X, Y and knee for Z) DRO. My goals:

Convert X and Y axis lead screws to precision rolled ball screws.

Install servos on X, Y, knee & quill. I want to be able to use the knee as the Z axis for contour milling and increased travel capability (vs. quill), and also use the quill for things like peck drilling, threading, etc.

Use a small rotary table as a fourth (A?) axis.

Use a Gecko G320 / breakout board combo ( or Gecko G100 ?????) to interface the PC to the mill.

Use a 19" touch screen lcd, mounted on an arm off of the mill, for control.

Use Mach 3/Quantum to control mill.

Use a VFD to interface Mach 3 with the motor for spindle speed control.

Interface things like flood coolant, mist, air etc. with Mach 3.

Install a tooling indexer, to be able to quickly set the Z depth of mills, drills, taps, etc.

Use a high speed air motor installed in R8 collet to use as a CNC jig grinder.

Use a spindle mounted electric router motor for routing / carving wood.

I am a Tooling Engineer by trade, with a tool & die maker background.
I work for a large metal stamping company, specializing in proggressive and transfer tooling for the automotive industry. I am also an expert Unigraphics and AutoCAD user.

I know that I'll need plenty of help with the electronics portion of this project from some of the CNCzone members. I have a good understanding of basic electronics and computers(see www.ultramame.com) but alot of this (electrical control) stuff is very foreign to me.

If all goes well, I also have a Grizzly 16 x 40 lathe (Model #G0509) I would also like to convert to CNC. Then ultimately I would like to build a 4' x 8' plasma cutter / router cutter table with 5-axis capability.

Have a great day! :banana:

[robhrzic]

My first dilema with the conversion seems to be locating a precision ball screw assembly for the X and Y axis.

I have seen that companies like Hiwin and Rockford Ball Screw have retrofit kits for Bridgeport type Series I and II mills. None of them seem to have a standard ball screw assembly that will fit my 10 x 54 table / travel size.

I have contacted Nook and Thompson, with little hope. They both told me that they do not have a standard assembly that I can use. They could make a custom assembly to my liking, but I fear that this would be very cost prohibative. Neither of them would give me a "ballpark" as to what the cost would be.

Can anyone direct me to a source for the ball screws that would fit this mill. I also need to know if I need to replace or convert anything else, such as ball / thrust bearings. Will I need to replace the yoke, or the
"cross-piece" that the ball nuts will be mounted to? I see that Hiwin has a yoke as part of their (smaller) retrofit kits.

I would like to make this machine as highly accurate and repeatable as possible. Would I need to get a precision ground ball screw assembly (+/- .0005" per foot)?

Could I instead go with the rolled (+/- .003 per foot) or precision rolled (+/-.001 per foot) ball screw assembly, and use the backlash compensation feature in Mach 3 to negate any unwanted tolerance stack up? Would this have a negative impact on climb milling performance / finish or limit table speed / feed drastically?

Thanks for your feedback.

Here are some (before) pics of the mill:

[wcarrothers1]

Back lash mode is not to be used for ANYTHING other then drilling operations in mach far as I've been told. Basicly the mode makes sure when coming to a location that it aproches it from both sides. (ie when told to move to x4 y4)

try to imagin the moves it would be trying to make while aproching a point while cutting a line or series of small moves. The servo motors can express their unhappyness.

So what I'm saying is mach backlash isn't to cure backlash if you are milling out a piece. only for accurate drilling. Least that is the only time I turn it on. In those cases it's usefull.

Far as the ball screws. You can order rolled from mcmaster car how ever you mention accuracy. In any event those are the least expencive ones I know of but you will have to mill your own ends, buy pullies, angled contact bearings ect to go with everything..

Frankly I can't see it being done for cheep in any event. Bearings get expencive and I don't think you can count on re-use of any of the thrust bearings used on the acme screws currently on the mill but I could be wrong.

After picking up my s1 bridgeport CNC and seeing it works so well I have a hard time with cost justification of conversions now seeing I only paid 1000.00 and the machine came with all those parts you are looking to make this CNC..

that mill looks nice though. even have a dro on the table height.. nice..

[robhrzic]

Thanks for the feedback, wcarrothers1.

I thought that you could minimize any backlash in a ball screw assembly using software, because I have seen a couple of companies with a guide to programming backlash compensation based on length of the ball screw and it's manufacturing tolerances.

I do realize that it won't be cheap to replace the lead screws on this mill.
I would like to do it with as little cost and pain involved though. I have most of the components, or know where to get them, for the rest of the conversion.

I just can't seem to find someone / somplace that can get me a ball screw assembly that I can use. I can't believe that I am the only person around that has ever attempted to retrofit a 10 x 54 mill with a ball screw.

I also do not know exactly how to go about the disassembly / reassembly process. What else will I need to replace / modify? What do I remove first? What do I stay away from? How does it come apart? What do I need to tweak or check after reassembly to verify that all is stright / square, etc....

Unfortunately the owner's manual that I received with the mill is a copy, of a copy, of a copy, of a copy, if you know what I mean. It has a few pages with exploded views of the main assemblies, but you really can't make anything out.

I wish I could find some sort of repair or service manual that would be able to guide me through the process. Does anyone know if there are any out there?

[klxrcr]

I think you are confusing backlash comp. with the screw mapping capabilities of mach 3 (and others).
The screw mapping compensates for the lead deviation found in rolled ball screws. With a ball screw no backlash compensation is required, unless of course your screw is completely shot.

For your conversion you may need to purchase screw stock, that is a ballscrew with no machining on the ends, and turn it to match your bearings, pulleys and possibly thread it for your bearing preload. But ballscrews are pretty hard and a 54" one may be difficult to fit in your lathe.

But!!!! don't be put off, many hundreds/thousands? have done this conversion before, the results will be worth it.

[Dennis Storzek]

The problem with backlash comp is that it doesn't solve the main problem caused by backlash. All backlash comp really does is tell the stepper (or servo) to add some extra steps when it reverses direction, on the theory that the table isn't going to start moving until the screw has turned that far, which may, or may not, be the case. It works reasonably well if all you use the mill for is to drive around between drilling locations. The problem comes in when you try to mill something. Much CNC milling is done climb cutting, as that yields a better finish, provided there is no backlash. If there is backlash, there is nothing to keep the table (and work) from continually jerking forward into the tool, which is bad for the machine, tool, and finish on the work. In manual machining on a mill with worn lead screws, one works around this by planning the work to never climb cut, but this isn't an option when doing CNC.

[MarcL]

Dwayne Elrod should be able to supply ballscrews for you. See http://www.elrodmachine.com/CNC%20XY...l%20Screws.htm.

He also has yokes, motor mounts, and other hardware you way want.

[robhrzic]

Thanks for the input klxrcr and Dennis Storzek.

Hey klxrcr, yea it's looking more and more like I'm going to have to fabricate my own ball screw assembly. I just need to figure out what type of screw I need to get (ie rolled, precision rolled, or precision ground).
I also don't know whether I need to be looking at a single preloaded ballnut, or a double preloaded ballnut. Any ideas or pointers?

Dennis Storzek, yea I know all about the perils of trying, either intentionally or accidentally, to climb mill on a manual mill without ball screws. That's why I want to make this mill as "tight" as possible, but without wasting time / money where it's not needed. I want it to be able to climb mill and contour mill and be able to get a nice finish with no chatter marks or blemishes.

That the main reason that I am having a hard time deciding on which way to go with the ball screws, in respect to a rolled vs. precision rolled vs. precision ground assembly.

It sounds like, if I am understanding correctly, that I do not need to waste my money on a more expensive ball screw assembly. I can get the same accuracy with a rolled ball screw vs. ground ball scew assembly, as long as I map the screw correctly in the software.

Though I am sure that the ground ball screw would be more efficient, relatively speaking, and allow for higher ipm speeds. Is this correect, or am I totally missing something?

[robhrzic]

MarcL, Thank You!!!!!!!!!!!!!!!!!!!!!

I sent an RFQ to Elrodmachine. Hopefully they can accomodate this mill.

[klxrcr]

Thanks for the input klxrcr and Dennis Storzek.

Hey klxrcr, yea it's looking more and more like I'm going to have to fabricate my own ball screw assembly. I just need to figure out what type of screw I need to get (ie rolled, precision rolled, or precision ground).
I also don't know whether I need to be looking at a single preloaded ballnut, or a double preloaded ballnut. Any ideas or pointers?

Though I am sure that the ground ball screw would be more efficient, relatively speaking, and allow for higher ipm speeds. Is this correct, or am I totally missing something?

A rolled ball screw should be plenty accurate for your machine, a rolled/precision rolled will not have as accurate of a lead as a ground, but can be mapped to nearly equal precision. A single or double preload nut will work, a single preloaded nut has 2 seperate "races" in it that preload against each other, where as a double would be 2 unpreloaded nuts back to back that preload against each other.

Ground screws are slightly more efficient but a rolled screw still has a 90% efficient force transmission which is a ton better than an acme. A rolled screw will not limit the feed rate of your mill, unless you plan on 1000ipm rapids. Which I hope your not. also a rolled screw will have a bit more noise than a ground screw, just the sound of the balls spinning around inside the nut.

One last bit I priced a set of screws from elrod, precision ground for a 9x42 bridgport, $1750...... I'm looking for a cheaper set now (rolled) And I hear that Rockford has sets starting around $700.

[robhrzic]

Well I finally got a hold of somebody at Rockford Ball Screw that could help me out, I hope. I talked to Ryan T., and he told me that they could make me some custom rolled ball screws.

I would need to send him my original lead screws so that they can duplicate them, or send him a drawing. I think that I need disassemble my table to see if I can take some measurements. Then I'll e-mail it to Ryan to see what they can do.

He gave me a ball park figure of about $800 for a complete assembly including the yoke.

This is a lot cheaper than Elrod Machine's $1525 for a X&Y axis kit, though I do realize that Elrod's kit has .0005 per foot precision ball screws. I am not sure if Elrod's kit includes a yoke though. I e-mailed them some questions, but never got a reply.

It seems that alot of these companies want nothing to do with hobbyists, or "one-offers". Too bad, I bet that they lose a lot of potential revenue that way.

Anywho, I priced some raw (unmachined) rolled ball screw stock and single pre-loaded ball nuts and flanges through Nook Inustries, and recieved a quote for $520. That does not include a yoke, and would require me to machine the ends.

Now all I need to do is to figure out how to take this thing apart! Anybody got any suggestions?

[robhrzic]

On another, completely unrelated note, I just finished installing my new 8-inch 6-jaw "set tru" chuck on my lathe.

I wanted to upgrade the plain 3-jaw chuck that came on the lathe standard, especially since it did not have reversible jaws.

I thought that this would come in handy if I had to machine the ends of the ball screws. If not, I'm sure it'll be put to good use on other projects!

Here are some before and after pics:

[robhrzic]

I received my 19" Touch Screen LCD on Thursday. It is an industrial open frame type with a 3M touch screen and serial interface. I pruchased it on eBay, item # 230156459157. This thing is awesome!

I unpacked it, connected it to my PC (future machine controller), downloaded and installed the software driver from the 3M site, and it all worked as planned. It took all of about 5 minutes.

Now all I need to do is fabricate a mount for the LCD, and an arm to hang it off of the side of the mill.

The only have two issues with the LCD.

Issue # 1, I can't seem to get the Mach 3 screen to "scale up" to a 1280 x 1024 resolution (see attached pic). Is this a limitation of the trial software (haven't purchased a license yet), or do I need to get / create a custom screen layout for Mach 3? I works fine at 1024 x 768, but the LCD looks much better and "crisper" at 1280 x 1024 since that is it's native resolution.

Issue # 2, I would like to be able to use the LCD in a portrait orientation. This way I can use the upper (approximately) 2/3 of the screen for Mach 3, and the lower 1/3 for an integral keyboard.

Can anyone tell me if there is a "generic" software driver out there that will allow me to rotate the sceen on any LCD?. I have "Pivot Pro" software that I got with my Samsung 23.1 LCD to perform this function, but it seems to be LCD model specific.

I would also need to make a custom 1024 x 1280 screen layout for use in the portrait orientation. It would have the standard Mach 3 control functions in the top of the layout and an integral keyboard layout in the bottom of the screen. Is it possible to create a custom screen layout for Mach 3 to accomodate all this?

I think that this would be the ultimate interface for a conversion like this, and it would really give it that "just like the real thing" flavor.

I have also received my Combo Breakout board from Bob Campbell Designs(see attached pic):

http://www.campbelldesigns.com/Combo-board.php

This is a big board! I like the fact that it is seems to have anything that you would ever need integrated already.

Finally, I got my contact probe from IMService (see attached pic):

http://www.cadcamcadcam.com/index.as...PROD&ProdID=10

This will work great for digitizing surfaces and generating point clouds for surface models, as well as making measurements like a CMM, or as a quick tool indexer when changing from one to another.

[robhrzic]

It's been a busy week. Got a lot more pieces and parts delivered. I think the UPS guy's been here every day!

I got four G320's from Gecko Drive:

http://www.geckodrive.com/product.cfm?pid=13

I got three Keling KL34-180-90 1125 oz-in. peak servo motors, along with UD Digital 500 CPR encoders (see attached pic):

http://www.kelinginc.net/KL34-180-90.pdf

Could anyone tell me how to figure out the proper limit, damping, and gain settings on the G320's when using these servo motors?

I got my 80 VDC, 20 amp power prep module from PMDX (see attached pic):

http://www.pmdx.com/PMDX-135/index.html

I purchased my transformer on eBay, item # 270152210432. It is a 56+56V 1KVA Toroidal type (see attached pic). Thanks again Al_The_Man for the assistance. Wow, this is a big sucker.

Now I can start laying things out to see how big of a steel enclosure that I'll need to fit all this stuff into.

I also received one of the two air cylinders that I plan on using for the Z (knee) axis air assist, or counterbalance system. This has got to be one of the biggest air cylinders that I have ever seen (see attached pic)! It has a 80mm (3.149" bore and 500mm (19.685") stroke. I purchased them on eBay, item # 330155200698.

They (both of them combined) should be able to give me about 1240 lb. of lift or "air assist" with 80 psi of air. I plan on cutting them down somewhat, as they are longer than I need them to be. I am going to beusing them in a "closed" type of system with an accumulator or surge tank. I need to find an air tank with about 4000 cubic inches of volume to minimize pressure rise as the cylinders are stroked. Hopefully I can find a 30 gallon air compressor tank somewhere, as this should fit the bill nicely.

[robhrzic]

Today I made my first baby step in the conversion. I made a mount for the four Gecko G320's, with four integral aluminum heat sinks and fans.

The heat sinks and fans were some old P4 CPU coolers that I had laying around, so I decide to put them to good use. They were a perfect size for the G320's, 2.875 x 2.960 x 1.375 tall, almost as if they were made for them.

I took the CPU coolers apart and used a 3-1/2" inserted face mill to flatten the back of the heat sink and get rid of a .075" boss that was sticking up, being careful not to fold the heatsink in half in the vise jaw.

I then drilled and tapped four #6-32 threads on the finished face in a 1.750 x 2.375 rectangular array to mount the G320's. I also drilled and tapped two #8-32 threads on one side of the heat sinks that I'll use to mount them to a riser bar.

I then removed the terminal blocks from the G320's, fastening them using BHCS's. I then re-assembled the fan and replaced the terminal strips, finally installing the four assemblies onto the riser bar. The riser bar will be installed into the enclosure using four aluminum spacers or stand offs using 1/4-20 SHCS's.

Step one done, with about one thousand more to go.

[robhrzic]

Went to go pick up my 2000 lb. capacity elevating lift table from Yellow Trucking on Monday (see attached pic). Purchased table from Northern Tool. This will give me the adjustable height and weight capacity that I will need to remove my mill's table when changing the lead screw.

Got my ball screws from Elrod Machine today (see attached pics). Boy they sure look nice. They are Hiwin precision ground ball screws. Funny, but when I spoke to someone at Hiwin directly, they told me that they did not think that they had a ball screw that would fit my mill. I just hope that these do fit.

These sure are some BIG screws! The X axis is 64" long, the Y axis is 30.5" long.

I bought them from Elrod because Dwayne seemed to be the only person that I've talked to in the last couple of months that seemed to know what he was talking about, and what I wanted. They were VERY expen$ive, but I think in the long run they'll be worth the investment.

If all goes well and according to plan, I'll try to do the actual retrofit this (long) weekend.

[robhrzic]

Started the ball screw conversion yesterday. Everything with the disassembly went smoothly. The problems started when I got the lead screws and yoke out and compared them to the ones that I purchased from Elrod Mechine.

The ball screws are smaller in diameter than my originals, and the machined ends are slightly different. I'm pretty sure that I'll be able to use these with some slight alterations to my support plates. I think I'll also have to make a spacer on the right side of the X axis, as it seems that the bearing journal's shoulders do not line up exactly.

The distance between the X & Y axis bore is about 3/8"-1/2" greater on my original yoke, the bores in my original are .090 larger in diameter, and the mounting holes' center to center distance is about .300 greater on my original. My original also has two 6mm dowel holes.

Looks like I'll need to return Elrod's yoke. I will either modify my original by chroming or sleeving the bores, then drilling and tapping for the ball nut flange screws, or I'll have to make a new one using pre-toughened 4140.
I'm leaning towards making a new one, as I think that will be much easier, and make for a stronger unit in the end.

My goal today is to take readings and precisely measure the original, so that I can create a detail print for a new yoke.

[txcowdog]

Wow... you sure are using the cream of the crop for parts. This will be a rockin' mill when you get it finished. Thanks for the idea using the P4 fans to cool the Geckos. I found them for $.99 at surpluscomputers.com so this was a good day for me. I am looking forward to your progress and love all the pictures.

Did you get answer on scaling Mach 3 up to 1280 x 1024? I am looking at that monitor also if you got that problem solved. Keep it going, it is looking great.

[robhrzic]

Thanks for the kind words txcowdog!

Haven't found anything more on using Mach 3 at 1280 x 1024, but if I do, you can be sure that I'll post my findings here. I will be focusing on that later on in the conversion.

Right now, I need to figure out how to make a new yoke that'll fit my mill.

[robhrzic]

After about 4 hours of making measurements and readings with a digital height gage, I think that I finally came up with a yoke design that'll work in my mill, using the Hiwin precision ball screws.

I used the top of the milling machine's table as a surface plate for making the measurements. Of course, just to make it more exciting, there wasn't a single square corner on the casting that I could work off of with any confidence.

I wound up using Elrod's design concept, while making alteration as needed for use in my machine.

I modeled and detailed the part in Unigraphics NX4. I plan on making the new yoke out of pre-toughened 4140, as opposed to the original's cast iron. Now I just need to send out the drawings to get manufactured, and hopefully it'll be done within a couple of weeks.