Hello everyone,
I want to see if I can make a metalworking manual/cnc lathe. For many years I have wanted a metalworking lathe, but the cost of a quality machine has stopped me every time. I have looked at all the cheap imports that I can afford, and I cannot bring myself to buy one. I want a beautiful precision instrument that is within my abilities to maintain and adapt to my needs as they develop. And I feel the skill and abilities I gain in building my own machines will be worth it.

This website and the projects of various members have greatly inspired me. So, here is where I am so far in my past attempts and in my current thinking on this project.

I have read the Dave Gingery series, and I began my version of his lathe project in fits and starts. I made the lathe bed and ways using a square 2 x 2 steel tube filled with cement that had a 1/4" plate screwed and jb welded to it. I learned how to cast aluminum and made a headstock and feet for the bed. The project stopped when I could not get the headstock moving on the ways smoothly.

Next, I got a 52" Atlas lathe bed on ebay for 20 bucks. The shipping cost more than the bed. This I sand blasted and repainted and then I had the ways reground. The company that ground the ways did not have the ability to grind the underside of the ways, so I began studying scraping. I purchased "Machine Tool Reconditioning" by Connelly, I got a 24 x 18 surface plate and I made my own carbide tipped scrapers.

I was going to then piece together the lathe with parts from other Atlases, but while trying Atlas machines, I realized they did not have the weight and tolerances I was looking for. They were all really worn out.

So, I started looking at old SB 9 and 10 inchers again recently, but the decent looking ones are out of my price range and I do not trust buying one without the thought that I would have it reground and reconditioned. This is an expensive proposition.

So I figure, if I am going to go to all the trouble of rebuilding old machines or rebuilding new, low quality imports, why not start from scratch. In the process, I can also make my lathe more modular, so that upgrading any component is not very hard. DIY makes sense also, because in my various projects I do not need thread cutting ability until I have cnc working. So, I will be happy without all of the split nuts and power feeds and quick change gears. Turning two cairrage handles if fine until I have the steppers hooked up.

Today, I have gotten a lot of ideas studying the various threads on this site and I would like to run them buy anyone who is interested in helping me.

1. I was thinking I could still use the Atlas 52" lathe bed I have. The tops of the ways are ground flat (I do not know how flat, however). Being old, I assume all of the twist has settled out of the casting. I could buy Thompson-type guide rails and attach them to the ways by drilling and tapping holes in the ways. The carraige and tailstock would attach to plates that attach to the Thompson bearings.

2. I have noticed that a lot of builders are starting with the mini-mill and mini-lathe headstocks that Little Machine Shop sells. I like that idea. I am wondering if the mini-mill and the mini-lathe headstocks have the same dimensions for the bearings, because I would like the get the mini-mill headstock and the base casting it attaches to. These two castings I could then bolt to a plate that bolts to the Atlas bed where the original headstock is screwed down. The v groove that is ground in the mini-lathe headstock in also not necessary. Could I get the mini-mill and swap in the mini-lathe spindle? Does anybody know?

3. I am thinking that I could push the mini-lathe/mill headstock, spindle, and bearings to handle a 9 or 10 inch swing as long as I do not take huge cuts. The lathe bed is designed for the 12" Atlas, I think. Any thoughts on this?

4. I could clamp the tailstock along the ways by pushing a gib tight against the outside surface of the original Atlas bed ways. A type of apron could hang down and one or two set screws could be tightened by hand. I don't think this would torque the Thompson guides and bearings too much.

5. I could power the spindle with a timing belt and pulley. I will have to learn which one to get. I am not sure what motor I will use to power it. Any suggestions would be great. It looks like a lot of people are finding tread mill motors and speed controllers.

As, I get further along I will show some drawings, I am also learning to use Rhino 3D.

Thanks

Hi,

I am also in the same position. After all the reading and data collection from CNCZONE and other sites i have planned to by a small manual lathe. I plan to start my conversion as follows.

1. Convert the Two axis into Servo driven system.

2. Convert the Spindle to a DC drive with VFD

3. Design a Automatic tool Changer

Yes its good to start from a old machine because lot a time will be saved on the mechanical part. Its better to check the if you are on the right path with a 100% error free accurate machine thats what is called a CNC.....?

My first plan was to start right from scratch using a 30 or 45 Deg slant bed. But when i started to plan the time could well be over 3 months for the mechanical part as i have to do all by myself. As i was in a dilema i came across many threads in the forum regarding converting a old lathe/Mill in CNC machine. Then an Idea Struck me the cost of my brand new lathe with 6" chuck will be arould $1000 Dollars without tooling. Thats pretty cheap that building one. Donot try to start with a Stepper as they are not accurate and a waste of money. Go for a Servo no second thought. I plan Gecko drives or Uli servo driver With Mach3

I estimate you can finish the project in approx. 45 days if you go with the old machine else it could take 4-6 months of your precious time if you dont have a helping hand.

Well i will be begining from scratch with the purchase of my lathe next Week.

Please share you ideas.

Regards,

Vishnu

I am planning to use my Atlas 52" lathe bed, an import mini-lathe headstock, a mini-mill cast riser block, and a mini-lathe tailstock. I will then use lots of adapter plates to mount it all to the lathe bed and the THK or Thompson linear guides and bushings. I think this approach will get me off to a good start without needing to weld and bolt up everything from scratch.

I am interested in having a modular approach, where I am not stuck because I do not have a South Bend part that no one is making any more, for example. It is cool to see in this forum how people are continually upgrading and expanding their machines to fit new needs. I want to first succeed in getting a manual lathe working. Then, I can use it to make the cnc parts and make a cnc router mill.

I think the flat ground Atlas ways are a good stable flatness reference point upon which to add the linear guides. I know that drilling and tapping the ways will destroy the lathe bed for any other purpose. But, it is not a big loss at $20 if it doen't work out.

My main concern so far is to finding a way to clamp down the tailstock, without torquing the linear guides and bearings too much.

Vishnu, I will consider your idea of only using servo motors.

I will share some drawings and pictures soon.

How about mounting your linear rails toward the outside edges of the ways and having the tailstock ride on the inside edges of the ways not on the linear rails? Then use a standard type tailstock clamp to lock it in place.

Alan

dfro, My concern is such a long bed and the mini lathe headstock. It is much to small to handle 50" long parts.
I think I would consider shortening the lathe bed some and it will be cheaper to build with shorter rails. But yeah, the flat ways of the Atlas would be a good surface to start from.
A mini mill dc motor is a pretty good match for the mini lathe head stock. Several dc drives you could choose from. Or a 1/2 hp 3ph with vfd might be a nice choice.
I just milled off the dovetail on my mini lathe headstock base as well as on the mini mill riser base. Messy cast iron but it machines real easy and nice.
I totally disagree with above opinion on steppers.

Steve

acondit, your suggestion is worth pondering. So far, my thinking is to have some apron type pieces drape down off of the plate that connects the tailstock to the guide bearings. These would drape down over the outside surfaces of the bed ways. I could then have a couple of set screws on both sides which would tighten some brass gibs into the ways and secure the tailstock/plate/bearings structure. I figure that if I am tightening it from both sides I will balance out some of the torque on both guide rails. I will draw a sketch in Rhino soon.

S_J_H, I hear you about the bed being too long. I would have to be careful about what machining I did. But, I could use the whole bed for turning wood and plastic, probably. I could mount a dremel and a carbide bur bit to the carriage and turn some pretty sweet braids and posts. For the metal turning, I would stay within 18", maybe. (?) Any other suggestions on what I could make the headstock out of? If you have looked at an Atlas 12" lathe, I am sure you noticed how little metal is in the headstock. These Chinese 7x headstocks look pretty beefy. I have tried to follow how people are beefing up the bearings in other forum threads, but I have to study the subject of bearings quite a bit more. Again a modular approach, where I am not tied to any one companies parts is appealing to me.

Thanks

It seems a nice addition to the lathe would be to have the headstock set up so that it could operate two ways. It could be a normal free running, fast spinning lathe headstock for normal turning and threading operations, or it could be set up like a 4th axis. Then I could do more complex machining like spirals, braids, and flutes. It could also be used as an indexing head and cut gears. Hell, it could turn a replica of the 'Venus de Milo' for that matter.

I would just have to figure out a way to change belts to engage or disengage the different motors.

A brushless dc motor would be very cool to try for the free running headstock motor. I have done a little PIC and AVR programming, and I have looked into this a little for some other projects. Maybe a lot of people are doing this already on their mills and lathes.

dfro, yes I think the mini lathe headstock could probably handle long parts made of woods and plastics, especially with a live tool. I am to "metal" minded at times.:D

Well the mini lathe head stock is pretty inexpensive, it's cast iron and simple to fit a belt drive to. For a small hobby machine it's not bad at all. You can also fit tapered rollers into the mini lathe headstock.

You could also try a 9x20 headstock. Very similar to the mini lathe but larger with tapered rollers and probably more expensive. Drawback is the spindle has a screw on chuck.They both have MT3 spindle tapers.


Make your own of course, I think that is only realistic if you have larger machines or access to them.
Anyhow, go for it! Sounds like a great project.
Steve

S_J_H, where could I get a 9x20 headstock and tailstock? That sounds like a good option - more metal would be good. Could you tell me which tapered bearings you are talking about. I would not mind improving the bearings as long as they do not cost $200 per pair - like I have heard some people mention on other threads.

I could get a SB 9" headstock and tailstock on ebay, but I am worried about the spindle, bearings and tailstock ram being worn out.

I could cast or fabricate a headstock from scratch, but I don't want to do that again, for now.

I cast a pretty nice little Gingery-style headstock, but I ran out of steam on that project. I will show some pics of it soon. I might make it into a cnc indexing head or something later.

Two questions: what kind of forces can linear bearings withstand? Are they used on production cnc lathe's?

Grizzly sells parts for the 9x20 at: http://www.grizzly.com/products/G4000

The total cost of the headstock parts I need and a complete tailstock is ~ $450. Ouch!

I think starting with the 7x headstock and tailstock is the best option. I can learn and make my mistakes on a smaller, less expensive setup and then maybe later upgrade to something bigger.

dfro,

What about buying a complete HF 8x12? They sell for $539 and you can now have them shipped to your local store. That way you'll have the spindle, motor, controller and chuck as well as the tailstock.

Alan

Alan,

I have thought about your advice and I looked at your website, also. Nice website.

I am considering this senario.

I will buy a Grizzly 9 x 19 lathe. Having a manual lathe and my Jet benchtop mill always 'on line' keeps me moving on all of my projects.

With these machines, I will build the cnc lathe I have been talking about and I will build a gantry mill. I am also interested in replacing the Jet mill, which has a round column and a lot of slop in the table, with a homemade cnc mill that uses a square column. Or i will buying a square column cnc mill.

I have heard the saying 'the perfect is the enemy of the good'. I fall into perfectionism easily. With these manual machines, I can build better machines and keep my other projects 'on line'.

What do you think? Also, how do you like you Grizzly 9x19? I like your idea of adding more rigidity to the compound rest with a four bolt clamp.

dfro,
I have a grizzly 9x19. I have done many mods to it. Some mild some wild. Have a look here-
http://s109.photobucket.com/albums/n48/S_J_H/9x20%20mods/?start=all

It's a nice lathe to use in it's current state.

I easily could have cnc'd it but I wanted to have both a manual and a cnc lathe.
I did consider a 8x14 or 8x12 for CNC. But after cnc'ing 2 small mills with dovetail ways I really wanted to test the waters with a scratch built machine and linear rails. So far I am very happy with my decision as the build is going well.
Steve

Steve,

Your photos are really inspiring. So, you have the Grizzly G4000, correct? Did you completely take apart your 9x19 machine and paint it grey? I am very interested in what you did to the tailstock ram, too. Did you machine a part that screws into the back of the existing ram?

Your diy cnc lathe build is the main inspiration for my cnc lathe design. I just will be starting with an Atlas lathe bed instead of making the bed out of steel plate. Could you tell me what type of linear rails and bearings you are using and where you got them? I will check your thread again to see if you have already shared this info.

What are your plans for a tailstock? You might want to consider securing the tailstock in the way I am thinking - tighten some gibs to the outside surface of the steel bars that the rails are attached to. Just and idea. I will get some cad pics up soon.

While saving up for the G4000 (and clearing up a spot for it in my small shop), I will get busy drawing up the cnc lathe. I will also start planning my gantry mill.

Any other advice you may have on the 9x19 table top lathes will be greatly appreciated. Thanks, Alan and Steve for helping me with my game plan.

Dave

Steve,

Is that a Grizzly G0619 mill I see in the background of some of your cnc lathe pics?

How do you like it?

I would like to replace my Jet round column benchtop mill with a square column mill.

Dfro,
Two questions: what kind of forces can linear bearings withstand? Are they used on production cnc lathe's?
Oh these things are trick! They can handle forces in any direction and quite a bit! Take a look at the manufactures specs for exact numbers.
Yes they are used on production machines. Here is a nice little article about them-
http://www.cncmagazine.com/vol6thru8/v6i21/v6i21f-LinrGuide.htm
I am using Hiwin brand rails/blocks and I bought them from www.automation4less.com.

As for the 9x19, yes it is a Grizzly G4000. I just painted it grey.
The tailstock ram was made by welding an attachment to it. But you could probably do it simpler with a screw on attachment.
It's a very very good mod IMO. Saves so much time having a lever feed. I also am into mini engine building. The tailstock can now be used effectively to hone cyls with a nice cross hatch.

I have the standard X3 (Grizzly G0463)
It's cnc'd now. Great little mill, Here's my cnc thread on it-
http://www.cnczone.com/forums/showthread.php?t=24983

My plans for the tailstock on my cnc lathe is to mount a mini lathe tailstock on a plate and use 2 rail blocks for it to ride the long axis rails. A pillar with a 7/16" thru bolt will drop down to the bed surface. The bolt will firmly lock it into place to any of the tapped holes in the bed. The pillar will be adjustable in length( only need a few thousands of an inch adjustabilty) to allow for flatness variance on the bed surface and precisely allow the tail stock to be bolted in position on the rails without brinneling the rails or bearings.

I am interetested in seeing your castings. I would like to learn metal casting in the future.
Steve

Steve,

I tried to download some images of my diy lathe #1, but the image size is too great. I will have to figure out how to reduce them.

I also did some sketches in Rhino of my tailstock plate idea. IMHNO (in my humble newbie opinion), this might be a better approach than what you are considering. The gibs tighten down with four threaded knobs. It is very easy and convenient to loosen, move, and retighten the plate/tailstock assembly. I think you will get very weary of unscrewing, realigning, and rescrewing down your tailstock assembly to the treaded holes you have on your design. Also, I think the locking leverage would be better with my design. There might be some flex in the aprons, though. Any, comments are welcome. This is just how I am thinking of doing the tailstock, so far.

http://www.cnczone.com/gallery/data/500/medium/tailstock_plate_1.png

http://www.cnczone.com/gallery/data/500/medium/tailstock_plate_3.png

Here is another idea: A vertical steel plate could run down the middle
The clamping knobs could be in front and behind the tailstock base that is clamped to the tailstock plate. These would then push the gib into the lathe bed to lock the tailstock assembly in place. The only problem is this eats up a lot of distance on the ways, reducing it's distance between centers. There could also be problems with the plate flexing up, also.

http://www.cnczone.com/gallery/data/500/medium/tailstock_plate_4.png

Please, let me know what you think.

I am also considering buying the slightly bigger Grizzly G0602 10x22 lathe.
I like that it has a 1" bore instead of the 9 x 19's 3/4". I am not sure though about the knobs that are used to change the gears, though. I will have to study the schematic to see how it works. Maybe the 9 x 19 would be better, since it is sold by so many different vendors. Parts would be easier to find if anything needed to be replaced.

Your mill conversion thread is very inspiring, and your resonance damper device is pure genius!! Could the holes be loosely filled with bb's or buckshot?

Dave

Photos of diy lathe #1:

Patterns:

http://www.cnczone.com/gallery/data/500/medium/foot_headstock_patterns.jpeg

Headstock, countershaft, lathe bed:

http://www.cnczone.com/gallery/data/500/medium/diy_lathe_1.jpeg

http://www.cnczone.com/gallery/data/500/medium/diy_lathe_2.jpeg

The bottom of the headstock:

http://www.cnczone.com/gallery/data/500/medium/headstock_bottom.jpeg

I decided to put some old-school curves in the feet:

http://www.cnczone.com/gallery/data/500/medium/feet1.jpeg

http://www.cnczone.com/gallery/data/500/medium/feet2.jpeg

I filled the lathe bed's square steel tube with Commercial Grade Quikrete Fast-set Non-shrink Grout with a piece of rebar going down the middle.

I also used Bondo where the castings shrunk from the sand a little. I was going to paint it all green like the countershaft.

As I said before, I am going to move on to diy lathe attempt #2. What I plan to try with this lathe bed is to make it into a scraping reference plate. I will attach handles where I screwed in the feet. I have heard the advice that steel is not a good substance for making scraping references, because it twists as you take material off. Granite makes a great reference. So, I figure maybe by adding the non-shrinking grout, some of solid properties of stone will be gained. I figure it is worth a try. For almost no money, and some elbow grease, I could have a 36" flat reference to use in scraping mill and lathe ways. It has already been ground, so it is close to flat already.

Comments very welcome,

Dave

Dave,
Your method for the tailstock lock looks good. I would use the side clamp version. I think my idea will work very well though so I'll stick with it since the bed is already tapped. I'm using a 7/16" bolt which can create some serious clamping pressure!
Realistically the little tailstock will be the weakest link, not either of our clamping methods. I don't plan on doing a lot of tailstock supported work so unscrewing a bolt to position it won't be any bother to me.

I would probably go with the 10x22 lathe. It looks better made and heavier than the 9x19.
Thanks for the kind words on my mill and damper.
The dampers work ridiculously good. I can't say if bb's or lead shot would work as well. One thing is for certain. Resonance is gone completely.

I like your pics of DIY lathe#1. I see a nice machine as a possibilty there..
But I guess a nice long flat reference surface would be very nice to have as well.

Steve

Steve,

Thanks for the comments and advice.

In the future I would like to try the lost foam casting method. It really simplifies a lot of things. I like this primer on diy lost foam casting:

http://www.buildyouridea.com/foundry/lost_foam_howto/lost_foam_howto.html

I would like to build a cnc foam cutter. It could be a good first cnc project, since the mechanical parts do not have to be very rigid and the steppers can be small. It could be a good place to work out all of the electronics and software for the first time, too.

It has been way too long since my last post!

Since then, I have been following the thread on epoxy-granite and doing some of my own experiments:

http://www.cnczone.com/forums/showthread.php?t=30155

I think casting parts with e/g could be a very viable technique for anyone who wants to make small to medium sized machines that are massive, rigid, and accurate. People should be able to find most of the materials locally. It allows one to cold cast parts for machine tools as a replacement for cast iron. And, it is a room temperature process - no need to build a foundry and pour molten metal. These features are what made me seriously consider using e/g.

E/G is a moldable material made out of different sized aggregates, measured in certain ratios so that they pack together very tightly. The epoxy binds it all together. E/G is massive and it has 10 times the damping properties of iron. That means better surface finish on your parts and longer tool life. E/G can also approach the strength of aluminum. Adding steel reenforcement, like rebar, greatly increases its tensil strength.

I have built some small propane and electric foundries, and I have had a lot of fun casting aluminum parts. But, there is no way that I could cast in metal the kind of massive machine tool castings that I want to make. So, I am giving e/g a try.

***

I am now using epoxy-granite along with an old Atlas 6" lathe bed and a Chinese import headstock to build my cnc lathe. Not all of my experiments have been successful, as I will show everyone. Maybe you can learn from my mistakes, maybe you can give me some suggestions.

The first step was to build a vibrating table. The e/g mixture is thick like dough and will not flow if you don't vibrate it. Here are some pics of the table I built.

http://www.cnczone.com/gallery/data/500/medium/vibrating_table_2.JPG

http://www.cnczone.com/gallery/data/500/medium/vibrating_table_3.JPG

The next thing I needed was a table vibrator to cause the e/g mixture to settle in the mold. I decided to try and make a variable speed vibrator using a wood router, a router speed control, and an eccentric weight spinning in a wood box. I was able to make a few test samples of the e/g and make a few parts with this vibrator, but the bearings got extremely hot and started causing the wood to smoke. In the pictures bellow you can see I used brass bearings. BAD IDEA! I then replaced them with ball bearings, and I still had them getting too hot. I think this might work, if the whole thing was made out of aluminum and some special bearings were used. Here are two pics of the design. DO NOT COPY THIS DESIGN!!

http://www.cnczone.com/gallery/data/500/medium/table_vibrator_2_BAD_DESIGN_.JPG

http://www.cnczone.com/gallery/data/500/medium/table_vibrator_1_BAD_DESIGN_.JPG

My next try was to make a vibrator that runs on compressed air. I chose to make it out of steel tube, aluminum plate, e/g, and a steel ball. It is a simple design and a couple of these could be made very easily to shake the table. You do need a sufficient supply of compressed air or else they won't run for very long. Here is a link to a post on the e/g thread where I showed some pics of a cast aluminum vibrator I made several years ago:

http://www.cnczone.com/forums/showpost.php?p=341777&postcount=2079

Here are some pics of the casting of my current compressed air vibrator:

http://www.cnczone.com/gallery/data/500/medium/compressed_air_vibrator_1.JPG

After casting with e/g:

http://www.cnczone.com/gallery/data/500/medium/compressed_air_vibrator_4.JPG

The compressed air spins a 1.625" steel ball around:

http://www.cnczone.com/gallery/data/500/medium/compressed_air_vibrator_5.JPG

Flattening its top surface with a coat of table top epoxy:

http://www.cnczone.com/gallery/data/500/medium/compressed_air_vibrator_6.JPG

I think one of these works pretty well at shaking my table. Two would work even better. With all the constructive and destructive interference of the two vibrators, you could really get some good chatter going. I think I would have to get a bigger air compressor, though.

My goal with the vibrator is to get the e/g to settle. While a few bubbles do rise to the surface of a casting and pop, there is no way of getting rid of most of the bubbles. It is way to thick.

If I make another of these in the future, I would make it round. I would have the inner steel tube, the hex bolts, and then an outer steel tube. A small amount of e/g would fill in between the tubes. That way, it would weigh a little less.

More pics later,

Dave

I have read that playing a blow torch quickly over the surface bubbles pops them.

While the mathematical and materials science wizards on the e/g thread are working out the best e/g formula, I decided to settle on the basic formula that Cameron (ckelloug) created and posted. I think if I make the castings thick enough, and put some rebar or steel rod in them for tensile strength, I will be fine. Time and experience will tell. Here is Cameron's formula:

#6 Agsco Brown Aluminum Oxide = .18744
#4 Agsco Quartz = .38420
#2 Agsco Quartz = .12950
#2/0 Agsco Quartz = .11514
3M G800 Zeeospheres = .11352
3M G200 Zeeospheres = .07020

All the ratios by volume add up to 1.

I sourced all of the quartz aggregates of different sizes and found Zeeospheres and began making test samples. The epoxy is from US Composites - 635 low viscosity epoxy with the 2:1 ratio slow hardener. Here is how I calculate the mixture of the various aggregates:

http://www.cnczone.com/forums/showpost.php?p=578984&postcount=3534

I created a spreadsheet to help me calculate how much aggregate and epoxy to measure. I start with the volume of e/g that I want and the spreadsheet calculates how much of each material I need to measure out in grams. This is a very handy tool.

http://www.cnczone.com/gallery/data/500/medium/epoxy-granite-spreadsheet.JPG

Here are some pictures of me mixing the epoxy in by hand. I did vacuum degas the epoxy before adding it to the aggregate mixture. For my purposes and the relatively small volumes I am dealing with, mixing and squishing everything by hand is the best method. It works bubbles into the mix, but I am not that concerned about it. I am more concerned with making sure all the mix is wetted with the epoxy.

http://www.cnczone.com/gallery/data/500/medium/measuring-epoxy-granite.JPG

http://www.cnczone.com/gallery/data/500/medium/mixing_epoxy-granite_1.JPG

http://www.cnczone.com/gallery/data/500/medium/mixing_epoxy-granite_2.JPG

I settled on mixing in 20-22% epoxy by volume - I had trouble getting it to flow at a smaller ratio. I am going to try West Systems 105 resin with the slow hardener in future batches. I noticed that its viscosity was thinner than the US Composites epoxy - that means I can use a smaller percentage of the epoxy in the e/g mix.

After making some small square samples, I decided I was ready to infill a small 6" Atlas lathe bed with e/g. The ides was to take a cheap, thin walled lathe bed and turn it into a solid mass of iron and rock.

More pics later,
Dave

I have read that playing a blow torch quickly over the surface bubbles pops them.

Mark,

That works with straight epoxy with nothing added to it, like table top epoxy. You pour a 1/16" to 1/8" film and the blow torch causes the bubbles to expand and pop. But, with the e/g, it is so viscous and the castings are so thick that there is no way for the bubbles to get out. The stuff is like fresh asphalt when it is mixed and ready to be molded.

Here are some pictures of me infilling the Atlas lathe bed with e/g:

http://www.cnczone.com/gallery/data/500/medium/bed_eg_casting_1.JPG

The lathe bed is clamped to the vibrating table and ready for the e/g. The springs under the top table are excellent at isolating the vibration from the workbench. I used my first vibrator to do this casting. I built a box that screws to the table. The vibrator is tipped vertically and screwed into the box, so that the vibration is along the horizontal plane.

http://www.cnczone.com/gallery/data/500/medium/bed_eg_casting_2.JPG

I used copper tubing to keep the e/g away from the bolt holes in the bed that are for attaching the headstock.

http://www.cnczone.com/gallery/data/500/medium/bed_eg_casting_3.JPG

A long shot of the casting. As the e/g is shook, it settles and gets a slight shiny sheen on top.

http://www.cnczone.com/gallery/data/500/medium/bed_eg_casting_5.JPG

Later, I decided I wanted to permanently cast the feet into the lathe bed, so I had to extend the copper tubes.

http://www.cnczone.com/gallery/data/500/medium/bed_eg_casting_6.JPG

Casting with the new and improved compressed air vibrator. It is very quiet compared to the wood router vibrator. Also, there are no shaft bearings and nothing gets hot - except maybe the air compressor motor. Air swirls around pushing the steel ball and eventually escapes out of the small hole in the center of the top plate.

http://www.cnczone.com/gallery/data/500/medium/bed_eg_casting_7.JPG

The finished casting with some cosmetic filling of small gaps and some shaping. That bodifile with its holder is a great tool.

Thanks,
Dave

Thanks for posting all those photos.

I really like your idea for the air driven vibrator. By varying the air flow can you change the frequency until the mix flows? There was a cool video on the main EG thread showing how a stiff mix went liquid when the right frequency was hit. Then the air should rise for the bubbles to be flamed off.

Mark,

Yes, the change in air pressure causes a change in frequency. I like the steel ball, because it vibrates slower than the router vibrator. I think the slower frequency is better at causing the e/g to settle. However, the vibration seems to work over a fairly wide frequency.

I took a test square (approx. 4x3x1.5") of some e/g that I cast to the shop of some friends. We put it in a 20 ton hydraulic press to see how much pressure it would take to break it. We had no measuring instruments to measure the pressure when it broke, but it went off like a gun shot!

That (unscientifically) confirmed to me that e/g is strong enough for my purposes, especially if I am going to cast some rebar in it to increase the strength.

Where the test sample cracked, there were small bubbles dispersed throughout it. E/G is just too thick to get all the bubbles out - it is 80% rock and dust. The only way I see to cast e/g with no bubbles is to mix and cast it in a vacuum.

The torch works great on table top epoxy. I used US Composites Kleer Kote Table Top Epoxy and a torch to cast the top surface of the e/g air vibrator.

Dave

Next is the headstock. I bought a generic import headstock and spindle for a 7x12 lathe from Little Machine Shop. I also got tapered bearings for it. I wanted to make it taller, but instead of buying a cast iron riser block or making my own, I decided to try e/g again. The first thing I did was to cut a steel tube and glue it inside the headstock. This keeps the e/g away from where the spindle needs to go. I also had to drill and tap a fourth hole on the bottom for attaching bolts. Are these headstocks attached on these 7x12's with only three bolts? If so, I do not see how anybody can get any kind of accuracy out of machines made with these headstocks.

The goal here is again to turn a cheap, thin walled casting into a solid mass of iron and rock.

http://www.cnczone.com/gallery/data/500/medium/headstock_casting_1.JPG

Headstock with bolts to key into the e/g.

http://www.cnczone.com/gallery/data/500/medium/headstock_mold.JPG

Mold with coupling nuts and hex bolts. These will be the attachment points for the headstock to be bolted to the lathe bed.

http://www.cnczone.com/gallery/data/500/medium/headstock_casting_2.JPG

Ready for casting. Steel rings and transmits vibration, so the bolts are not touching.

http://www.cnczone.com/gallery/data/500/medium/headstock_casting_3.JPG

Casting the e/g.

http://www.cnczone.com/gallery/data/500/medium/headstock_demold.JPG

Demolding. I did not use a very good release agent - car wax. I planned to destroy the mold anyway.

http://www.cnczone.com/gallery/data/500/medium/headstock_casting_4.JPG

I leveled the front of the headstock and cast a coat of table top epoxy to smooth the front face. I made walls around the perimeter with tape. Do not use electric tape if you are going to pass a torch over the epoxy, use masking tape. The electric tape started to shrink and I had to use molding clay to try to stop it from leaking!

http://www.cnczone.com/gallery/data/500/medium/headstock_casting_5.JPG

The finished headstock casting with some cosmetic fill and sanding. Notice the steel tube I glued inside it.

http://www.cnczone.com/gallery/data/500/medium/headstock_and_spindle_2.JPG

Headstock with spindle - rear view

http://www.cnczone.com/gallery/data/500/medium/headstock_and_spindle.JPG

With spindle - front view. This thing is really solid and heavy.

Thanks,
Dave

Head stock looks awesome. I could have done something similar and saved myself some money and machining time. how do you plan to align the axis of the spindle to the ways? shims?

Nice work Dave,
If you are thinking of making a thinner mix next time, why not try the table top epoxy?

The product is thinner, unless there is a cost issue, it just might be the solution for a thinner mix.

Best regards,

Bruno

Dave,

Congratulations for actually getting something done. Fabulous work I might add.

You mentioned release agents so I'll provide a quick comment:

Have a look at Mann Easy Release 200. At $11.00 a can, it cant be much more expensive than car wax. I've been using it for test specimens. It leaves a satin finish but in a polished steel specimen mold, I got an epoxy impression so accurate as to be able to see .0005 inch or less imperfections in the surface grinder finish after the thing was lapped to 3000 grit diamond.



--Cameron

Nice work on that head stock rise! I have a 7x12 which I CNC'd and the 3 three bolt interface to the bed is the least of these machines worries.

I was really surprised that mine has a very accurate headstock alignment (could be a fluke). The tail stock is the real turd!

http://www.wrathall.com/Interests/machining/Alignment_check1.htm

Thanks guys for the kind words.

will,
I have recently noticed your 'Granite surface plate lathe' thread. Your progress looks great! That is going to be a monster lathe! I also just checked out your www.gothamchopshop.com website - very cool work. I have been fascinated with metal forming work. I got several of Kent White's videos at http://www.tinmantech.com. I learned a lot from them and got a lot of ideas. He is a real artist. I play saxophone, and, at some point, I would like to fabricate my own custom tenor saxophone neck out of brass sheet. That project is way on the back burner for now, though.

As far as aligning the headstock, I am not completely happy with how I have done it. After casting the headstock and the lathe bed, I made a series of time consuming mistakes that I will share with everyone, later. I hope others can learn from it and that I can get some advice from people. The alignment of the headstock is included in those mistakes.

What I plan to do is a fix for something I could have easily included in the e/g headstock casting. I am going to make two 1/8" or 1/4" thick steel strips that are the width of the bed ways and sized to be glued to the underside of the headstock with a toughened epoxy like 'jb weld'. They will each have two holes in them, so that the clamping bolts can reach the four coupling nuts that are cast into the headstock. I will roughen the glue side of the steel strips, and scrape the bottom surfaces that will contact the ways. I will also plug the headstock's coupling nuts with modeling clay to keep the glue from getting in the threads. I will then glue the strips to the bottom of the headstock, clear away any squeeze out, turn it rightside up, and place it on one of my surface plates (the glue is still wet at this point). Beforehand, I will wipe some blue spotting paint on the surface plate to protect the plate from any stray epoxy.

Next, I will knock and tap the headstock to get the front surface of the spindle to line up perfectly square with a machinist square that is placed next to the headstock on the surface plate. This is all by eye, but I think I will get close. Then, once the 'jb weld' has cured, I will mark the lathe bed with 'hi-spot' blue paint and scrape the headstock into the ways. Maybe I will have to shim, eventually - we'll see.

I think, in this case, the 'jb weld' is kind of like a poor man's Moglice. Here is a quote from the Moglice site:

"Molding to fit precision: The molding process gives a full surface contact even given irregular shapes and angles. This eliminates the need for costly match machining of parts."

Come to think of it, this method might be easier than casting the steel strips into the headstock. I am liking this idea.

Bruno,
It is my recollection that the US Composites - 635 low viscosity epoxy with the 2:1 ratio slow hardener is thinner than the US Composites table top epoxy. I could be wrong, though. I definitely know that the West Systems 150 with the slow hardener is thinner than the Kleer Kote table top epoxy. I will explain how I know that, later. It's not pretty.

Cameron,
Thanks for the advice on the release agent. I will get some. I know you guys worked that out somewhere in the e/g thread, but at the time I didn't feel like digging for it.

And, thank you for your efforts in producing the e/g formula that I am now using. I wouldn't be trying this without you and the other contributers to the e/g thread sharing your work!

Mark,
I enjoyed browsing your site. There is a lot to learn there.
You might want to get a spare tailstock at Little Machine Shop and do an alignment trick with toughened epoxy, like what I am thinking of doing with my headstock. Find some way of having a film of epoxy between two parts of the tailstock (grind away some iron between the two castings). You could chuck a precision ground rod in the headstock chuck and also the tailstock. Then, clamp and knock and adjust the tailstock until a dial test indicator shows that the rod is perfectly aligned along the length of the lathe bed - top and side. When the 'jb weld' cures unclamp everything and clean it up with a file. You could even drill and tap some screw holes in the bottom to hold the two pieces together, if you are afraid of the glue breaking. I don't think it will, though.

I am thinking of doing something along those lines for the tailstock on my lathe.
It might be worth trying.

Thanks,
Dave

Sounds like you have a good plan for headstock alignment. I'm a bit worried about it on my lathe. Scraping is my last resort. I'm hoping that it magically falls into alignment (I used stacked ground plate for my spacer).

Keep it up! I'm really impressed by your work in this thread.

I was thinking of doing something like what you suggest with the tailstock, but haven't yet, cause it would seem like investing time polishing a turd :)

The foot is so badly machined and scraped, and the joint between the foot and the other casting so lacking in perpendicular surfaces, that I really don't see it being a great starting point. On top of that the alignment changes measurable between the ram locked and ram unlocked.

A CNC lathe doesn't need a set over tailstock, so I would rather start with a block of cast iron, and machine and scrape the base properly, and then set it up to slide on the bed, driven by the saddle. Then with a boring bar between centers on the existing tailstock, the new tailstock could be bored true to the bed.

Actually what is more likely is that I finally move out of the apartment into a house, and buy a 12x36!

I was thinking of doing something like what you suggest with the tailstock, but haven't yet, cause it would seem like investing time polishing a turd :)

I hear you. I am doing the DIY route, so that I can control the quality of my machine. I want modular components that can be replaced from several sources. And, with e/g, I can make the castings I need for it.

Maybe you could make an e/g tailstock for your lathe to practice casting. You could bore a better housing for your tailstock quill out of steel rod, make a steel base to travel on the ways, then find some way to mold it all together with e/g. Might be fun.

Thanks,
Dave

It is tempting...

Another good idea, share by Larry (lgalla) on the e/g thread, is to use epoxy to make an accurate surface plate. Thank you, Larry!

http://www.cnczone.com/forums/showpost.php?p=241687&postcount=17

Epoxy has the wonderful property of conforming to the curvature of the earth like water. A (skillful) pour of a large 4' x 8' epoxy surface plate will give you a table that is flat to within .005". I thinks that number includes the curvature of the earth and the inevitable imperfections that will happen, like dust and small bubbles that were missed by the torch during the pour.

This idea was a real epiphany to me. I had studied and practiced hand scraping a little, looking for ways that I could make my machines more precise. Hand scraping is a long, exhausting, thankless task that becomes exponentially harder as the parts gets bigger. And, I could never figure out a way to flatten a large table or a long steel tube for supporting linear rails. This epoxy surface plate idea seemed like the answer.

So, I thought I would experiment on my small DIY lathe. Make small mistakes, if they are going to be made. The plan was to:

1. Roughen the top of the lathe bed
2. Plug the four bolt holes in the lathe bed and the headstock with modeling clay
3. Level the bed as best I could using a precision level and shims
4. Cast an epoxy surface plate on top of the lathe bed
5. After the lathe bed has cured, set the headstock upside down on the, now flat, bed
6. Shim and adjust the headstock so that the front surface of the spindle was perfectly perpendicular to the bed
7. Cast an epoxy surface plate on the bottom of the headstock
8. Open up the bolt holes again with a drill bit and a 45 deg. countersink
9. Drill and tap the holes for the linear rails
10. Clamp the bed, headstock, and rails together and have an accurate machine with no scraping.

So, after infilling the lathe bed, I roughened the top of the ways with an angle grinder. If you are going to ruin a lathe bed, go all out, use an angle grinder!

http://www.cnczone.com/gallery/data/500/medium/bed_eg_casting_4.JPG

The next step was to set up my casting area. I wanted a very stable workbench to work on, since I would be trying to hold a very accurate leveling of the bed for a long time. I chose to hold up my maple workbench slab with concrete blocks. I put sand between the blocks and seated them together so that there would be no wobbling. I also set up a cover with plastic to stop dust from settling on the surface as it was curing.

http://www.cnczone.com/gallery/data/500/medium/epoxy_surface_plate_pour_2.JPG

Larry also suggested, when pouring the epoxy, let the extra seep over the edges in order to avoid a meniscus. That sounded like a good idea to me, so I used masking tape and scrap paper to collect the over-pour.

http://www.cnczone.com/gallery/data/500/medium/epoxy_surface_plate_pour_3.JPG

The beginning of the pour

http://www.cnczone.com/gallery/data/500/medium/epoxy_surface_plate_pour_4.JPG

After passing the propane torch over it several times, it smoothed out like glass. There is a fine line between getting all of the bubbles popped and burning the epoxy, so that it starts to curdle and warp. I did go over it several times with the torch as new bubbles rose to the top. I waited at least 2 minutes between passes to give the epoxy some time to cool of a little. I also painted the epoxy over the edge with a brush, all the way around, so that the epoxy would seep over evenly.

The first major problem I had with this process was - FRUIT FLIES. Don't ask me why they wanted to be in a metal and wood working shop, but they were a real problem. They were attracted to the smell of the epoxy and kept landing in it! They ruined a couple of pours. So, I went to the hardware store and got those sticky hanging strips that smell like over-ripe bananas, and I commenced the slaughter of innocent fruit flies.

What I had to do was dig the flies out of the epoxy while it was still rubbery and do another pour over the last one. Eventually, the fly strips got all of them.

Once the lathe bed had cured over a couple days, I began setting up the headstock for a pour. I used the new lathe bed epoxy surface plate as the stable reference upon which to pour a surface plate on the bottom of the headstock. If I set everything up correctly, my thought was that the lathe bed and the headstock surface plates would both be made parallel to each other by gravity.

http://www.cnczone.com/gallery/data/500/medium/epoxy_surface_plate_pour_5.JPG

The machinist square has no light showing through at the top and bottom of the spindle's front surface.

http://www.cnczone.com/gallery/data/500/medium/epoxy_surface_plate_pour_6.JPG

A wide shot of the headstock epoxy pour

http://www.cnczone.com/gallery/data/500/medium/epoxy_surface_plate_pour_7.JPG

A close up of the pour. Smooth like glass. The blue modeling clay is keeping the epoxy from getting into the threaded holes.

http://www.cnczone.com/gallery/data/500/medium/epoxy_surface_plate_pour_8.JPG

Here is the headstock placed on top of the lathe bed after both pours and sufficient cure time. I also rubbed some oil on both surfaces to keep them from sticking.

The second major problem with the epoxy surface plates arose when I tried to drill out the bolt holes and drill and tap the holes for the linear rails. I was having problems with the table top epoxy adhering to the metal and e/g surfaces.

More later,
Dave

When I drilled and countersunk the bolt holes in the headstock and bed, I noticed a halo of white forming outside of each hole. The epoxy was delaminating from the surface. It happened on all of the holes, no matter how slowly and carefully I drilled. So, being like a scab that I had to pick, I went after one of the holes with a chisel, to see how bad it was.

A big piece came off very easily. The epoxy stuck better to the e/g than the metal, but it still was not very hard to remove. In comparison, toughened epoxy like 'jb weld', adheres very well to e/g and metal.

http://www.cnczone.com/gallery/data/500/medium/adhesion_problems_1.JPG

http://www.cnczone.com/gallery/data/500/medium/adhesion_problems_2.JPG

Here I am destroying the epoxy surface plate on the bed.

http://www.cnczone.com/gallery/data/500/medium/adhesion_problems_3.JPG

Big chunks came off of easily.

I then decided to try West Systems 150 Epoxy with the slow Hardener. I did the same setup - leveling the bed, dust cover, etc. I learned that not all epoxies are the same for pouring an epoxy surface plate.

Firstly, the West Systems cured with lots of small ridges and waves in the surface finish. You can see that in the picture below. Maybe this is because West System epoxy is more sensitive to the propane torch.

Secondly, its viscosity was lower and therefore much more of it was seeping over the edges. I did not notice how severe this was until I tried pouring the bottom surface of the headstock. I would do a pour and come back and notice that one corner was still not flat. While the epoxy was still rubbery, I would do another pour, and the same bump would show up. I finally discovered that most of the epoxy was flowing off of the headstock, down the masking tape and paper, and all over the back of my workbench and onto the floor! That was a mess! Luckily it came up fairly easily with a body file. Everyone should have one of these. The handles are showing in the picture.

From these experiments, my impression is that the makers of table top or surface plate epoxies design it to have a low enough viscosity so that it will self-level, but high enough so that it will stay on the surface without too much seeping over the edges. And I am sure they design it to be more resistant to the heat from a torch. I can't see anybody making this technique work without using a torch or a heat gun.

http://www.cnczone.com/gallery/data/500/medium/adhesion_problems_4.JPG

There were still problems with adhesion. The West Systems epoxy stuck much better to the e/g, but it still did not stick very well to the metal surface. Notice the chip missing on the corner in the above picture.

So, for the headstock, I decided to machine the West Systems epoxy surface flat with a fly cutter on my benchtop mill.

http://www.cnczone.com/gallery/data/500/medium/milling_headstock_bottom.JPG

For, the lathe bed, I decided I wanted the ways reground. I asked the owner of a local grinding shop, if he could grind down ways that had been mangled by an angle grinder, and how much. He said he could do it and it would probably cost $60.

So, before bringing it over to his shop, I needed to lower the surface of the e/g with respect to the top of the ways. I decided to drill multiple holes into the e/g with a hammer drill and chip out the remaining e/g with a concrete chisel. This was a very difficult job, especially since I had a cheap hammer drill that would not keep the drill bit tight in its chuck. Chiseling out the extra was exhausting. E/G is hard stuff.

http://www.cnczone.com/gallery/data/500/medium/epoxy_surface_plate_removed.JPG

The lathe bed with all of the epoxy surface plate chipped off.

http://www.cnczone.com/gallery/data/500/medium/e-g_removal_from_lathe_bed_1.JPG

A close up of my drill/chisel work.

http://www.cnczone.com/gallery/data/500/medium/e-g_removal_from_lathe_bed_2.JPG

A wide shot. I am glad I made this mistake on a small machine!

Next I re-leveled the lathe, and filled in the rough, pitted surface with some aggregate and table top epoxy.

http://www.cnczone.com/gallery/data/500/medium/lathe_bed_epoxy_fill_1.JPG

http://www.cnczone.com/gallery/data/500/medium/lathe_bed_epoxy_fill_3.JPG

Below is a picture of the lathe after being precision ground. The machinist charged me $60 to grind the top and both outside surfaces of the ways to within .0005"/ft accuracy.

http://www.cnczone.com/gallery/data/500/medium/ground_ways_2.JPG

I still think that an epoxy surface plate is a great idea. I think it really shines at leveling large surfaces.

http://www.precisionepoxy.com./TestBed.htm

Maybe some of you reading this have ideas about how to get the epoxy to adhere better. I would also like to try the surface plate epoxy that www.precisionepoxy.com makes, to see if it adheres better to e/g and metal.

I do, however, like the idea of embeded steel rails protruding from the castings that can be ground flat, and I think I will incorporate that in most of my e/g designs. Eventually, I would like to make a large cnc router mill, in which I extensively use the epoxy surface plate technique.

Thanks,
Dave

You are making a very fine job of that lathe :)

I'm very impressed by your uncompromising experimental approach The finished machine will be a real credit to you.

Thanks for sharing :)

Bill

dfro,
Let me know your block's dimension so I can imagine the lathe's size.

Re the problems adhering, I did not see any mention of degreasing the metal surfaces. Maybe I just missed it, but my experience filling my mill base with E/G is that any kind of grease or oil (and probably a lot of other contaminants) are problematic.

I'd look for some solvent-type degreaser that will evaporate completely. Epoxy, after all, can be used to join metal and should not delaminate so easily.

Cheers,

BW

Bill,
Thanks. I purposely chose a small project, so that my big mistakes are not that big - start slow and small, they say.
I have an Atlas 12" lathe bed that is 52" long. It is also waiting to be turned into an e/g-iron lathe. I am very glad I did not have to drill out the e/g and chip off the epoxy surface plate on that one!

Asuratman,
My workbench is 72" (6') long and 16" wide. The lathe bed is from a standard Atlas 6", which I think is 36" long. The headstock is a standard cheap import 7x12 headstock. I have increased its LxWxH with e/g.

Bob,
I did forget to mention that I went over both parts repeatedly with naphtha solvent. On the lathe bed, I ground away all of the surface that touched the mold with the angle grinder, anyway. I also went over the bottom of the headstock with some sandpaper. I will edit my earlier post to add that.

Thanks,
Dave

It is a shame you didnt succed to make an epoxy surface plate since it looked like it leveled realy flat. I belive epoxy to metal bond failure could be easily overcomed. At least i had positive experiences bonding steel especialy with West System but a couple of things should be kept in mind.
First is Amine blush - a thin layer of amine formed on the surface after the epoxy enters the gel - faze(especialy developed in a cold or humid environment).New layers of epoxy cannot stick to old ones covered with Amine layer. Since epoxy makes no chemical bond (except with its self and that before gel-faze) with its substrate it relies entirely on mechanical bond. So a good surface preparation is a must. First one should rub a hardened epoxy with scotch bright and water (plenty of it), sand for good adhesion, and then clean with water again. Also degreasing ( clean industrial aceton or similar) should take place especialy if steel surface was soked with oil previously (used lathe ways..).
There are couple of tips from MOGLICE Handbook:


Sandblasting has been proven to generate the best adhesion surface, althoughit is not practical for largercomponents.
Devitt Machinery also has a deoiling powder used in rebuilding, primarily to act
like oil dry; it draws excess oil out of old oil soaked cast iron. The deoiling powder
will change color when it absorbs oil and should be reapplied several times until
it stops changing color. Gentle heat can also be used to sweat oil out of a casting.




As for mechanical bond - even a very corse sandpaper (80 or 60 grit) rotating on a high RPMs will not produce deep enoug cuts to ensure good holding key. One should make a series of angle cuts on the surface to ensure good adhesion. Also When aplying you can rub a first layer of clear epoxy with a steel wire on the surface to ensure it entered all the small cuts and holes.



Miro

Miro,

That is very good advice. I think the surface plate idea can definitely work.

The owner of the grinding shop talked to me about how porous cast iron is - that old cast iron machine tools have soaked up oil like a sponge.

Thanks,
Dave

Thanks for posting all this detail of what is working and what not. The E/G thread is great, but highly theoretical. Your practical examples are a big help to understand E/G for us hobbyists.

That lathe bed looks outstanding cosmetically. Nice work.

What was the plan for the epoxy levels lathe ways, if the adhesion had worked? Were you then going to mount a ground steel way on top for a saddle to run on, just run the saddle on the epoxy, or are you mounting linear guides?

If a most machine shops would will willing to do a grinding job like this for only $60, then this method of machine manufacture is easily in reach of hobbyists.

My local hardware store sells granite blocks for steps, for about $20. Two blocks epoxied together with steel plates glued on, and then ground perpendicular and flat would make a killer small mill frame.

My next step was to attach and adjust the linear rails. That took a couple of hours. I tested the adjustment of the rails against the ground outside surface of the ways with a dial test indicator. I think I got within a couple ten thousandths.

http://www.cnczone.com/gallery/data/500/medium/testing_linear_rails_1.JPG

Linear rail - headstock end.

http://www.cnczone.com/gallery/data/500/medium/testing_linear_rails_2.JPG

Linear rail - tailstock end.

http://www.cnczone.com/gallery/data/500/medium/headstock_bed_and_linear_rails_1.JPG

Top view of the bed with the attached linear rails and headstock

http://www.cnczone.com/gallery/data/500/medium/headstock_bed_and_linear_rails_2.JPG

Here is the machine with everything bolted together. It is one solid mass of e/g and iron (except for where the spindle passes through the headstock). I will weigh it to see how heavy it is at this point.

I am going to paint it green.

***

The next step is the machine base. I wanted it to be massive and cheap. So, I thought I would experiment with cinder block. I have made two columns of cinder block - one for each foot of the lathe bed. The two columns are also tied together with two 1/2" threaded rods that span the gap. The two rods could support a shelf, if I want.

Each column has three stacked cinder blocks and a cap block. The two hollow cells in the blocks have rebar and poured concrete in them. This bonds them together into one solid mass.

Each block cost $1 and I used about $7.50 in poured concrete. This is a cheap way to create a massive, vibration damping base to hold up your machine tools. I have almost finished making them. I will post pictures, soon.

Thanks,
Dave

Mark,

Thanks, for the kind words.

Thanks for posting all this detail of what is working and what not. The E/G thread is great, but highly theoretical. Your practical examples are a big help to understand E/G for us hobbyists.

I want to share all of the small details of what I am trying, because I think the details are what are going to snag people. You really want everything figured out before you cast your parts, otherwise you might end up with a very heavy paper weight! I am also giving a lot of details, because I want advice from others on how I can do better.

My advice is to start slow and small - make big mistakes on small parts.

What was the plan for the epoxy levels lathe ways, if the adhesion had worked? Were you then going to mount a ground steel way on top for a saddle to run on, just run the saddle on the epoxy, or are you mounting linear guides?

My intention was to mount linear rails to the epoxy surface plate on the lathe bed. I do not think table top epoxy would have the durability to withstand direct contact as a bearing surface. Also, the bottom sides of the ways have no precise reference to the top surface plate. So, there would not be a way of using the bottom surface to accurately clamp a carriage.

If a most machine shops would will willing to do a grinding job like this for only $60, then this method of machine manufacture is easily in reach of hobbyists.

I think E/G is wonderful stuff that will allow anyone, if they are willing to learn and practice, make highly accurate machines at a relatively low cost. I also like the idea of locally sourced materials.

My local hardware store sells granite blocks for steps, for about $20. Two blocks epoxied together with steel plates glued on, and then ground perpendicular and flat would make a killer small mill frame.

I think you have a good idea. If you are uncertain about the flatness of the granite surface, you could rough it up with a grinder, level it, and cast an epoxy surface plate. Before doing that you could mark the attachment points for linear rails, drill over-sized holes, and glue in threaded inserts with toughened epoxy. Plug up the holes with some modeling clay before casting the surface plate. You could do some great experiments for very little money with these granite blocks.

I think I am going to cast my benchtop mill base in e/g, so that I can give it a more complex shape - make recesses for the leadscrews, for example. I also like the idea of casting the threaded inserts into the part rather that drilling holes in granite. Drilling rock is not fun! And I want the protruding steel bars (the foundation for linear rails) to be keyed into the e/g base with bolts - like how I cast the headstock.

There are a lot of great possibilities.

Thanks,
Dave

Hi Dave,

Great work keep up the updates.

You where concerned about clamping the carriage and tailstock, did you ever check this thing out? http://www.automation4less.com/linearbrake.htm

Its what you need, fits linear rails and it wont happen the deforming of the carriage itself when you clamp everything, plus no need to worry about the unprescise reference of the bottom within the top since it holds by the linear rail...

Hope this is a solution

Brenck,

Thank you for the information! That is a great idea. I have been scratching my head on this for a long time, but I never even considered clamping onto the linear rails. I really like the feature that the clamp puts no stress on the bearing blocks.

The clamps are $90+ dollars a piece, which seems a bit expensive to me. I am thinking that it would not be very hard to make a pair of these out of a soft metal, like brass. I would want to be careful about where on the rail they are clamping, because I am concerned about wear.

Thanks again,
Dave

Here are some pictures of me building the base for the lathe. It is made of concrete blocks, concrete caps, rebar, steel tubing, and 1/2" threaded rod. The lathe feet will bolt down into coupling nuts, which are glued into the concrete caps with toughened epoxy - 'jb weld'. I cast the columns upside down, and turned them over when the concrete cured. They are very heavy; but, I was able to turn them over without any help from a hoist (or a chiropractor).

http://www.cnczone.com/gallery/data/500/medium/concrete_base_1.JPG

Dry stacking everything, and testing that the rebar is the correct length and that the steel tubes fit through the holes I drilled in the block.

http://www.cnczone.com/gallery/data/500/medium/concrete_base_2.JPG

The rebar keys into the two holes in the cap.

http://www.cnczone.com/gallery/data/500/medium/concrete_base_3.JPG

Mid-way through casting one of the columns.

http://www.cnczone.com/gallery/data/500/medium/concrete_base_4.JPG

The two finished columns.

The threaded rod will pass through the steel tubes in each column, and will be secured with nuts and washers. This will tie the two columns together. I am going to do some cosmetic plastering of the outside, paint them, and pour an epoxy surface plate on top of each column.

Thanks,
Dave

I might have an answer to why the epoxy did not stick, it could be the naphtha. The naphtha may be leaving an oily residue behind. Next time you try this, you may want to use something different, or maybe pass your blow torch over the area once the naphtha dries. Did you know that you can use naphtha in your zippo lighter, much cheaper than buying lighter fluid.

One of the next steps I am puzzling over is how I am going to power the lathe spindle. Little Machine Shop has a 250 W motor and controller for around $150. For $300 I could have a 350 W motor and controller.

But, I think I would like more power - maybe 500 W (2/3 hp). I would also like to be able to control the speed from the computer. Here is what I might purchase:

I am thinking of driving the headstock with a 24VDC electric scooter motor. It is capable of handling 500 Watts:
http://www.monsterscooterparts.com/24vo500wamow.html

I could power it with a 24V@20A unregulated power supply:
http://www.mpja.com/prodinfo.asp?number=6645+PS

And, I am thinking of using this controller:
http://www.robotpower.com/products/simple-h_info.html

The computer or a microcontroller will feed the PWM signal to the motor controller board.

I am also wondering what size timing belt to use to drive the spindle - T5, 15mm wide? I will buy from here: http://www.sdp-si.com.

I am also scratching my head on what my top spindle speed should be? I see metalworking lathes usually advertised in the range 1,800 to 2,000 for the top speed. The motor does 2600 RPM, so I am wondering if I should gear down the spindle speed.

Any advice or ideas would be great.

Thanks,
Dave

Hey Dave,
Many people have used this combo:

Controller:
https://www.surpluscenter.com/item.asp?UID=3524012922354518&item=11-2269

There are a lot of different treadmill motors on ebay, might want to shop around, but here are a couple:
http://cgi.ebay.com/Nordic-Track-Treadmill-Motor-95V-DC-or-Wind-generator_W0QQitemZ140378393926QQcmdZViewItemQQptZExercise_Fitness_Cardio?hash=item20af344d46
http://cgi.ebay.com/18A-TREADMILL-PERMANENT-MAGNET-DC-MOTOR-WIND-GENERATOR_W0QQitemZ270521533253QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item3efc570345

-

Andy

My CNC'd 7x12 uses a 550W 3Ph 2 pole motor driven by a VFD.

The motor drives the spindle through about a 1:3 9mm HTD pulley set.

I have the VFD set up for about 10Hz through 100Hz, which gives a range of spindle speeds from about 250 to 2500.

I am pretty happy with this in general, although there is not much torque at lower speeds (this is relative - it still has way more grunt than the original 250 chinese Watt motor in low gear).

As an example of the limits, turning an M40x1.5mm thread in steel was not successful, even when taking dozens of passes. This was due to the combination of needing a deep enough cut to not rub, slow enough spindle speed for my slowish Z axis to keep up, and limited torque at low speed.

On my set up, 2500 rpm top sped is enough, as my Z is not fast enough to practically use higher speeds. If it had a faster Z, and more spindle speed, then aluminium could be turned down to swarf even faster.

VFD/3Ph motor combinations are really nice to use, as they have large speed ranges, are quiet and smooth, the motors are sealed, designed for industrial duty, and easy to control from a CNC with an interface like Homans.

The ideal for me, would be a choice of two ratios between motor and spindle, 1:9 will give you a speed ratio of about 80-800 with oodles of grunt, while a near 1:1 will give you 600-6000. A 3/4 HP set up like mine would be plenty.

I have considered making a lay shaft setup to do this, but have limited space, and no driving need at present.

Andy and Mark,
Thanks for the feedback.

Mark,
Thanks for the rpm numbers on you lathe. That gives me more of a feel of what I should be shooting for. My shop has 3 phase wiring at the breaker box, but I do not think I want to deal with it. I would like to be able to just plug the lathe into a 110 VAC outlet. I also would like to hang the motor off of the headstock, and those 3 phase motors look way too heavy for this. I may change my mind after studying it some more, though. I also have an Atlas 12", 52" long lathe bed that is waiting to be turned into an e/g-cnc lathe. Is there a way to control the speed of your machine from the computer?

Andy,
The treadmill motor/controller idea looks like a possibility. There is such a wide variety of voltages and speeds. I would like the ability to buy new motors easily, becuase if I fry it, I would like to find the same motor with the same rmp's. But, it is frustrating to try and get specs on new motors - the sellers on the web do not tell you the voltage rating, amps, rmp's. All they want to say is, "motor for the Nordic Track X1000". On ebay, the two main voltage ranges seem to be 90VDC and 130VDC. I am wondering if I buy a 130VDC motor and run it with the Minarik controller at 90VCD, whether I will get a big drop in torque.

Also, I wonder if the motors are sensitive to being driven by a higher voltage. What would it do to a 130VDC motor if I ran the controller at 180VDC?
I see that the Minarik is controlled with a 10K pot. Do you know what the voltage is on that? It seems that to control from the computer, I am going to need one of those PWM to 10V conversion boards.

I am still considering using the lower voltage scooter motors - either 24VDC or 48VDC. The motors cost from $70-180, and they go up to 1000 Watts of power. On Jameco's website I found a lot of switching power supplies that can give me up to 1000W of power at 24VDC and 48VDC. The scooter motor controllers are relatively inexpensive, too. I would just need a PWM to 5V conversion board to run the motor from the computer.

I like that the parts for the scooter motors are easy to find and the specs are straightforward.

Thanks,
Dave

I ended up getting a 90VDC motor and the Minarik controller.

Thanks,
Dave

I did a quick search of Minarik on the board here, and there seems to be plenty of people using it, so support here should be good :)
If I'm not mistaken it looks like LeeWay (http://www.cnczone.com/forums/member.php?u=4929) used it on both his Mill (http://www.cnczone.com/forums/showthread.php?t=32609) and Lathe (http://www.cnczone.com/forums/showthread.php?t=59421).

Looking forward to seeing yours in action

Best Regards,

Andy

My Lathe VFD runs off single phase. I provide three phase variable frequency to the motor. That photo I posted distorts the size relationship, as it was shot with a wide angle lens pretty close up. The motor is not that big or heavy.

Mark,
Could you tell me what VFD unit you are using? And, is there a way to control it from the computer? I would like to learn more and possibly use a VFD on the next cnc lathe.

Thanks for the help guys,
Dave

When using a treadmill motor do you loose torque as you decrease rpm like on a AC motor?

Some of them are CCW rotation, is there a way to change it?

When using a treadmill motor do you loose torque as you decrease rpm like on a AC motor?

Brenck,
It is my understanding that Pulse Width Modulation is a good way to control the speed of DC motors, because the pulse goes from full ON with the full voltage and full torque, to full OFF. So, the torque stays relatively high throughout the whole speed range. I know almost nothing about how AC motor speed control works.

I chose to buy an industrial 90VDC motor made by 'Pacific Scientific'. The little research I did showed that they have many motors of different power ratings, all with the same mounting flange and the same rpm's (1,750). I did not want to mess with all of the different sizes, shapes, and rpm's of the treadmill motors. And the total lack of specs on the motors frustrated me. The only people giving sufficient specs were sellers on ebay.com. But, if a motor goes out, I do not want to depend on luck that someone is selling a motor of the same shape, power, and rpm's on ebay.com.

Some of them are CCW rotation, is there a way to change it?

I did read some cnczone posts where people did not seem to have trouble wiring treadmill motors to spin the other way. But, there was a concern that the brushes could possibly chip when being run in the other direction. The angle that the brushes meet the commutator may not be right in the other direction.

Thanks,
Dave

I guess im a day late with this one, but I just came across another great option.

Over in another thread (http://www.cnczone.com/forums/showthread.php?t=59663&page=43) mmprestine suggested these:

http://www.automationdirect.com/adc/Shopping/Catalog/Motors/IronHorse_(TM)_General_Purpose_Motors/Three_Phase_Rolled_Steel_56C_Motors_(0.33_-_2HP)/MTR-001-3BD36
http://www.automationdirect.com/adc/Shopping/Catalog/Drives/GS1_(120_-z-_230_VAC_V-z-Hz_Control)/GS1_Drive_Units_(120_-z-_230_VAC)/GS1-21P0

A bit more but probably well worth it.

Best Regards,

Andy

Andy,

That looks like a great setup. I will definitely consider using it on my next build. I want to make a 12"+ capacity lathe out of my other Atlas lathe bed and e/g. But, I am not sure a 1 hp motor would be powerful enough for that size capacity lathe. Looking at the 'J&L' catalog, the 13" lathes have motors that range from 1 1/2 hp to 3 hp.

I could put that setup on a mill, though. I like that the VFD can be controlled from the computer with no problem.

Thanks for the info!

Dave

No problem,
I also noticed they have a 2HP motor there too that is not very expensive.
The drive to power the 2HP motor isn't too bad either if you have 220/3 phase available.
If you only have single phase available the 1/3phase input drive is very expensive unfortunately.

-

Andy

If you only have single phase available the 1/3phase input drive is very expensive unfortunately.

Are you talking about this one?
http://www.automationdirect.com/adc/Shopping/Catalog/Drives/GS2_%28115_-z-_230_-z-_460_-z-_575_VAC_V-z-Hz_Control%29/GS2_Drive_Units_%28115_-z-_230_-z-_460_-z-_575_VAC%29/GS2-22P0

I would pay $241 for what it does. I think I will do this setup with cnc lathe no. 2. I should probably get the AC line reactor, also:

http://www.automationdirect.com/adc/Shopping/Catalog/Drives/GS_Drive_Spare_Parts_-a-_Accessories/GS_AC_Line_Reactors/AC_Line_Reactors_%28All_GS_Drives%29/GS-22P0-LR-1PH

Thanks,
Dave

The control interface of nearly all VFD (invertors), is the same. I have both Siemens Micormaster on the mill, and a Lenze on the lathe, and they are both controlled through a 0-10VDC input. The Invertor will provide the reference 10V, so you can control them manually with a simple Pot. Hommann designs make a nice cheap widget which replaced the manual pot with direct control from the CNC software of you choice (like Mach3, TurboCNC or EMC).

I believe most PWM DC drives can also be controlled the same way, and Peter Homann does off a wide range of different flavours of interface.

http://homanndesigns.com/store/index.php?main_page=index&cPath=1&zenid=dc90d6ec78a0bac09db346e86c199826

The Fwd/reverse commands to a VFD (Invertor)a simple discretes.

3Ph AC motors are cool as they run sort of constant voltage and current, with just the frequency changing to control speed. Unlike PWM DC motors, whose torque drops off as they slow down, due to the pulse width of "windings energised" being less and less, the torque of these motors is constant up to rated RPM (and I believe the power in constant above this).

Andy and Mark,

Well, you convinced me! I went ahead and bought the VFD and the 3 phase 1 hp motor. It looks like the 90V motor and the controller that I bought will be a jump start on my e/g mill.

My first attempt at a DIY lathe was based on the book, 'The Metal Lathe', by David J. Gingery. I learned how to cast aluminum in order to build a beefed up version of his lathe. In the book Dave came up with a very nice motor mount design that allows you to change the belt on the spindle pulleys easily. I copied that design and welded up a motor mount, which has been sitting on a shelf for years.

Tonight, I took the dimensions of the 3 phase motor over to the shop and measured my motor mount. It looks like the motor will fit perfectly. I think it is going to look very cool. I might also be able to get two speed ranges like Mark described.

Thanks for the great information,
Dave

Dave sorry of your mis adventure.The problems were well explained by the members posts.I will agree with what was posted.
TT/epoxy is formulated to flow out smooth over the edges.Never use laminating resin as it will blush and is oily.
Napatha is oily.Use acetone to clean metal.Your cast iron probably had lots of oil in the pors.
DO NOT POUR FRESH EPOXY OVER GELLED EPOXY.
Glad to see your build.I am still in the design and collecting parts stage.I am planning to use concrete as a base and E/SP to make accurate surfaces as you are.
Good work Dave

Dave sorry I have not had time to read all the posts.I recall Bob at CNC cookbook used E/G as a fill in a mill colum with good sucess at damping.Anyone have a link?
I think I will steal your idea of the cinder blocks.It eliminates the need for a welded steel frame and cheap.You are basically rasing the floor to the machine.
Larry

My point of view is that an used lathe is a verry good start, get some work done with it.
Then use some steppermotors todrive the X and Z axis get some work done with it.
Expend the lathe with live tooling or a tool changer or better place tooling in a row on the X axis.
Build up some experiance and have fun learn a lot and make nice things.
Dream about your lathe and learn to write your own Gcode and be save.
One important aspect you must not forget, use sharp tools.

My point of view is that an used lathe is a verry good start, get some work done with it.
Then use some steppermotors todrive the X and Z axis get some work done with it.
Expend the lathe with live tooling or a tool changer or better place tooling in a row on the X axis.
Build up some experiance and have fun learn a lot and make nice things.
Dream about your lathe and learn to write your own Gcode and be save.
One important aspect you must not forget, use sharp tools.

I recall Bob at CNC cookbook used E/G as a fill in a mill colum with good sucess at damping.Anyone have a link?

http://www.cnccookbook.com/CCMillEpoxyFill.htm

Hi Dave,
That's a really nice looking lathe. As I'm working very slowly on building my own, I'm really interested in what you're doing. I bought an old SB heavy 10 headstock off ebay to use on mine. Other parts accumulated so far are all the THK rails and blocks and two Fanuc AC servo motors.

Anyway, your epoxy/granite and epoxy surface plate adventures are great! Thanks so much for posting all the details. I'm intrigued about using the epoxy surface plate method for my router rebuild (which is my first project) and later for my lathe. My main issue is how to get flat surfaces for mounting my THK rails, and this seems like it would be a much easier solution than transporting my machine to a large grinder or scraping in rail surfaces by hand. Do you know if anyone has actually used the epoxy surface plate method to mount linear rails directly to? I'm wondering if it's hard enough to take the pressures applied to it by the rails and the weight of components resting on top. I'm planning on a gantry in excess of 400 pounds, so I wouldn't want the epoxy to deform over time...

Dave,
Nice job on your lathe project. Very cleaver building methods! I’m really excited to see how it turns out.

I especially like your idea for using epoxy to make a flat surface to put your rails on. It’s a shame that it didn’t work out. I had the same idea myself, but just haven’t gotten as far as trying it out. I want to build a lathe by filling a large structural steel tube with e/g and using that as the base. Hot rolled tube isn’t so flat though, so your method may be a good way to get it there. My other thought was to make a large E/G surface plate and use it to lap the lathe bed. I would either use course lapping compound or just use some spray on post-it glue to temporarily hold a bunch of sheets of sand paper on the plate. We’ll see, it sure it hard to find the time.

I'm surprised that the epoxy did not adhere to your lathe bed better. I was thinking about something else you might try next time. How about etching the metal surface with acid? I know you need to etch concrete floors before you use epoxy paint on them. Also, painting cars often involves using an etching primer. I know you roughed the surface up with a grinder, but chemicals might make the surface even more porous – as if cast iron wasn’t already porous. Acid would also eat up any rust. Just a thought.

Good luck on your lathe.