[Cartierusm]

I have a couple of questions regarding the conversion of my HF 9x20, I've read other posts but have some basic questions. Hope ya'll can help.

First off I'm using some old 1/2-5 2 start precison McMaster Carr Acme Lead Screws and Dumpster CNC Anti backlash nuts for the two axis'. So for the conversion, it's fairly straightforward, right? I mean just make some brackets to hold some angular contact bearings on one end, a floater bearing on the other and the the nut to tie into each axis. Then hook up an encoder to the spindle so mach can keep time and do threading operations? Setup Mach and you're done, right? It just seems so basic compared to converting a mill.

Now a couple of basic lathe operation questions. If I'm parting a piece and the piece is let's say 2" in diameter, I can't have my parting tool which is like 3/32" thick sticking out that much because it will chattter? So when doing a deep parting operation do I program in a stop and readjust the tool length, or does the parting blade stick out as far as I need from the beginning?

If I don't mind changing belts for different spindle speeds then I don't need to do anything else besides add the spindle sensor for doing CNC threading operations?

When setting up a part in a CNC lathe how do you determine where the tool zero is, do you measure the part exactly then use feelers to touch the tool to the non-moving part, zero out mach and then offset by the amount of the feeler? I'm cluless on how to program CNC lathes so far, so my questions are basic.

[acondit]

I would suggest that you want a finer thread pitch unless you are using servo motors with about 1000cpr or better resolution. Unlike a router or a mill, typically you are wanting very precise control of the x-axis, and moving 0.001 is going to take 0.002 off of the diameter. Also, I tried a delrin (acetal) nut for the x-axis and it was too flexible for good accuracy. My z-axis is 5tpi and I am running 1/2 step mode. So that gives me 0.0005 per step with reasonably good repeatability of 0.001. For my x-axis I am running 10tpi with 4:1 gearing and 1/2 step. This is theoretically 0.0000625" per step, but, it is still difficult to maintain much better than 0.0005" (0.001" diameter) repeatability.

So, you really need to think about what you are trying to achieve before deciding what kind of screws to use.

Alan

[Cartierusm]

Thanks. I'm going to be using 495 oz/in steppers with 72VDC at 2000 steps per rev. so a resolution of .0001. Plenty of power and I already have the screws so I'll try them and see how they work and if I get errors then I'll switch to ballscrews.

[acondit]

Thanks. I'm going to be using 495 oz/in steppers with 72VDC at 2000 steps per rev. so a resolution of .0001. Plenty of power and I already have the screws so I'll try them and see how they work and if I get errors then I'll switch to ballscrews.

If you're talking about microsteps as part of your count, you really can't rely on the position of the microsteps when it comes to holding position for a turning cut. You need to base your repeatabliity on full step positions.

On the x-axis I am using 200 oz/in steppers with 4:1 reduction gearing or the equivalent of 800 oz/in to the screw with a 10tpi precision acme screw running a cast-iron nut.

Good luck,

Alan

[Cartierusm]

I understand what you're saying but it seems like everyone uses microsteps in lathes, routers and mills. I find it hard to believe that all these people are wrong. There is even a company selling their machine just like the way I'm going to do it. http://www.d3cnc.com/. Again I'm not trying to argue with you just stating some facts trying to figure out what the correct solution is.

[acondit]

I understand what you're saying but it seems like everyone uses microsteps in lathes, routers and mills. I find it hard to believe that all these people are wrong. There is even a company selling their machine just like the way I'm going to do it. http://www.d3cnc.com/. Again I'm not trying to argue with you just stating some facts trying to figure out what the correct solution is.

1/2x5 - 2 start is effectively 2.5tpi. D3CNC is using .200 pitch ballscrews (or 5tpi). Twice as fine as what you are suggesting.

I stand by my earlier comment, you need to define your goals before proceeding. If the goal is just to have a lathe the "runs under CNC control", you can accomplish that. If your goal is some certain degree of precision, then you need to design for that (remembering of course that our 9x20 machines will never be a Hardinge).

Regarding microstepping, search for information on repeatability, accuracy and microstepping and look for comments by Mariss Freimanis of Gecko fame.

Alan

[Cartierusm]

Thanks I'll check it out.

I was mistaken my screws are 1/2-10 5 TPI so they're the same.

[Cartierusm]

Acondit, what program do I use to make the gcode for cnc lathe? I use Rhino for most of my regular mill cad work and rhinocam, but have no clue what to use to program the lathe? Thanks.

[acondit]

I hand write most of my code, so I can't really help you with that.

All you need to do is draw 1/2 of the profile that you want, and generate the gcode for that profile.

Alan

[jhowelb]

If you are using Mach 3 there is a wizard included for most any basic shape you would want and a threading wizard for inside and outside threads.

There is Lazy Turn in progress but still not complete yet.

http://www.machsupport.com/forum/ind...topic=5767.new

[Cartierusm]

Acondit, what do you mean by draw half of it and then generate Gcode from that? I take it you mean draw half of it either vector or in a CAD program? Then what do I use to generate the Gcode?

Jhowelb, I've started to look at the mach built in wizards, how would I string a whole bunch together to make a complex piece? Would I just do each opertion separately and then cut and paste the posted codes together?

Thanks for the help. I've finally found a program Dolphin Parmaster Lathe and I've got a RFQ in for the hobbier so we'll see how expensive it is.

[acondit]

Acondit, what do you mean by draw half of it and then generate Gcode from that? I take it you mean draw half of it either vector or in a CAD program? Then what do I use to generate the Gcode?

Think about looking down from the top of a finished piece. If you were to cut longitudinally along the center of the z-axis, you would have 1/2 of the finished piece. Then rotate the piece 90 degrees and lay it on the flat surface, and take an infinitesimally small slice the same direction as the first cut (but on piece rotated 90 degrees). You are left with a theoretical 2d slice that has a straight edge at the centerline of the z-axis and the profile of the cuts made by the cutting tool. Program to cut out that profile with X and Z moves, remembering that the part is hidden in your starting stock so by repeated cuts you have to uncover the hidden part.

You said that you currently use Rhino, how do you generate gcode from that? I started off drawing the parts, exporting a DXF and then using NCPlot do create gcode from the DXF. Eventually I had created enough sub routines that it became faster to write the gcode by calling my own subroutines.

Without showing all of the subroutines that I have written, this is what the code looks like to cut this part:

o230 call [1] [500] ( Change to tool 1 Facing & turning )
o500 call [0.650] [0.010] [-0.0625] [0.000] ( Face off end )
o130 call [0.630] [0.00] [0.375] [0.750] ( Turn end to 3/8 for spider )
o130 call [0.630] [0.75] [0.500] [1.500] ( Turn bearing surface to 1/2 )
o130 call [0.650] [2.25] [0.610] [0.500] ( Turn body to just under 5/8 )
o230 call [6] [800] ( Change to tool 6 threading )
o120 call [0.500] [0.700] [0.550] [0.4987] [0.050] ( Thread for bearing preload nut )
o230 call [9] [400] ( Change to tool 9 parting )
o240 call [0.650] [2.75] [0] ( Part of ballscrew end )

However, I use EMC2 rather than Mach.

Alan

[Cartierusm]

I use rhinocam to generate the gcode. I don't know enough about gcode to write by hand, but thanks for the help. I'll try messing around with ncplot and a dxf.

[jhowelb]

You can use a text editor to join one operation to another if you want. I have always wanted more direct control, I don't like crashes brought on by unforeseen complications.

[acondit]

I use rhinocam to generate the gcode. I don't know enough about gcode to write by hand, but thanks for the help. I'll try messing around with ncplot and a dxf.

I didn't when I started either. However, you really have to learn it to debug problems (even those created by CAM programs). I wrote software for a number of years, so writing subroutines to handle common tasks is second nature.

Alan

[Cartierusm]

Alan, I've watched the mach turn video and it helped wonders. It helped me understand cnc lathe setup and such. Plus mach lazycam and lazy turn are useable I just have the mess with them. I learn little bits of gcode here and there eventually I'll know enough but it won't be anytime soon.

I just checked out your webpage, pretty cool stuff, I've always wanted to do home forging. I can build anything so this shouldn't be too hard. Can you answer some questions? If you go to this link you'll see part of what I'm trying to make.
http://hbsboard.com/index.php/topic,2219.0.html
I tried contacting the poster but he is no longer on and his email is dead.
If you can help I'd appreciate it. When casting aluminum does it have to be anything special, I mean a certain grade? Also in the link is he just using regular steel to make his molds out of? Do you need a release agent? If I make a mold similar on my Bridgeport CNC will the casting come out accurate or will I have to remachine for the bearing journals? I think the bolt coming out of the part is steel that he imbedded into the mold, is that possible and how hard to keep it aligned so it's in the center? Thanks.

[acondit]

I tried the link but it wouldn't let me in. Maybe you could post a picture of what you're trying to do.

I started out melting aluminum pistons but I eventually found out that I could buy ingots of 356 Aluminum for less than I had to pay for scrap. If you cast with scrap you have less control over your final product. I worked in a pattern shop for a while and we had lots of castings made. The material was often spec'd by the engineer who designed the part. There are all kinds of issues depending on what you are casting, issues like shrink rate, machineability, porosity, and strength. For most of the one off's that people do, they are just worried about getting a part that works for them and they don't have to satisfy an engineer.

If you are familiar with Lindsay Books you can find a lot of books on home casting. It is relatively easy to get started but it is also potentially dangerous. Do some reading before you start, so that you can do it relatively safely.

Alan

[Cartierusm]

Thanks, there is a book at my library that will get me some much needed info. I read all last night. I've always wanted to do it and now I acutally have a use for it. I have an old stainless steel container originally used for Liquid O2 in a lab, it's 14" high and 15 wide, should make a perfect furnace shell.

I do need to have some control over quality, where do you buy your ingots and how much per pound?

Here are some pics. I also signe up for hobbycast on yahoo.

Here's a link to the book Metal casting : appropriate technology in the small foundry

[acondit]

Thanks, there is a book at my library that will get me some much needed info. I read all last night. I've always wanted to do it and now I acutally have a use for it. I have an old stainless steel container originally used for Liquid O2 in a lab, it's 14" high and 15 wide, should make a perfect furnace shell.

I do need to have some control over quality, where do you buy your ingots and how much per pound?

Here are some pics. I also signe up for hobbycast on yahoo.

Here's a link to the book Metal casting : appropriate technology in the small foundry

Just remember that you need to plan on at least 2 to 3 inches of refractory cement to form the walls of the furnace. So 15" minus 4" leaves you 11" and 15" - 6" = 9" of open diameter with that shell. I use a #16 crucible in my furnace (16# molten aluminum) and it is about 8" in diameter. My furnace is about 18-20" in outer diameter. You need room for the flame and for the handle to pickup the crucible.

A diecast mold needs to have appropriate draft on the "vertical" surfaces and very smooth "vertical" surfaces. It looks like that mold is designed to cast around the metal rod. You want to be able to get the part out of the mold after you cast it. I have never done diecasting so that is about the extent of my ability to help.

I haven't done any casting for awhile. The last ingot that I bought I paid $0.83 / pound, but scrap aluminum was about $1.00 /pound then.
I live in Oregon and up in Portland there is a foundry supply house, Lagrand Industrial.

Alan

[Cartierusm]

Thanks for the help. What material did he make the die out of, is it just steel?