My ELS starts at 0 for the X axis value and a thread is set up with a Total depth, depth of first pass and depth per pass plus a number of spring passes that don't change the depth.

?

CNC requires 'knowing' where the centerline of the lathe is.? This can be set by knowing the diameter after a pass and dividing that by 2 and setting the X axis to that point.? From now on you could be turning by setting the X to a specific position for each pass.? Kind of complicated from my perspective but trivial with G-Code the

?

So for my ELS, by deciding that 0.000 is where the thread or turning starts one can do a clean pass and then measure the diameter.? Say it's 0.566" and you want 0.50".? So the diameter has to be reduced by 0.066" or a cutting depth of half that at 0.033".? That's your thread depth.

?

The thread pitch is any value you'd use for normal turning to get a nice finish with the tool radius you have ground on your tool bit.? Let's say 0.003" makes a nice smooth almost polished finish so that's the threading pitch.

?

So what I would do is set the first pass depth to 0.010" and each normal pass at 0.010".? The ELS calculates the final depth for threading at 0.003" which, for HSS, ?makes a nice cleanup pass with a smooth finish.? Maybe do a couple more spring passes to deal with the work deflection and any lathe instabilities.

?

Press START and cut the 0.003" pitch thread which is narrower than the 0.010 radius on the tool bit so it's not really a thread.? The total depth of cut was 0.033" so at the end of it all the diameter is reduced by 0.066" and final diameter should be 0.500".?

?

Now pretend you want to reduce a 1" piece to 0.5".? Starting again at zero the only difference is thread depth which is now 0.25"? Number of passes at 0.010" would jump way up to 25.? Might be better to do 0.024" per pass for 10 passes and the final pass will be about 0.010 and then the spring passes will clean up the last half thou or so.

?

Point is.? Like zeroing the dial on the cross slide to zero.? Measuring the work to determine how much to remove.? Then setting up thread depth to that value and depth per pass.? Done.

?

Hopefully it's just as easy with the RELS.

?

John Dammeyer

?

?

?

?

?

BTW, one of the nice things about castings and patterns is that if you screw up you can melt it down and do it again.? Or if you decide you want a different motor cast another bracket.? Brushed DC Servo on the right, AC Servo on the left.? Same speed/torque rating.

?

Now it's true that a chunk of 1/2" plate and 4 spacers do the same thing.? But plate is expensive.? This was scrap melted down aluminium window frames I think. ?Maybe a dollar or two worth of gas and consumables.

?

John

?

?

Bill,

Thanks

You've got a couple of years yet to catch up.

?

I just went through one of those treadmill stress tests.? They keep upping the speed and the incline while the 10 probes stuck all over the chest record heart and breathing I guess.? Also not allowed to hold tightly to the hand rail during this but take long steps. ?And every minute they take blood pressure.? ?All very intensive.? Just about fell off it at the end.

?

Then I'm given a prescription for a drug to lower my heart rate and a CT chest scan is going to be scheduled.? All because I complained about being out of breath climbing stairs and being off the scale on the low end of red blood cells etc.? Apparently there is something wrong.?? They aren't saying what.

?

Explains also why, during our Silver Level Standard Ballroom Dance test I had trouble remembering the steps.

?

Anyway, about the CAD stuff.? Fusion 360 I'm told is pretty good.? So start watching the video tutorials, pick a project like the Gingery Lathe and use the CAD to translate Dave Gingery's 2D drawings into 3D.? Spend about 2 hours per night doing it with no rush and no completion date nor any need to use them when you are done.?

?

As an example, I drew this assembly back in 2006 using, the at the time very inexpensive, AlibreCAD.? I'd already cast and was using the part.? The drawing from 2D out of the book to 3D was purely an exercise.? To learn how to do CAD.? If the attachment comes through you can open the pdf and then click on the drawing to rotate and look at it from different angles.? Alibre allows publishing drawings to 3D PDF files.

?

?

The reason I suggest something like the Gingery Lathe instead of your own project is that this makes it a true training exercise like a class assignment.? The only reason for doing it is to learn.? Not as a way point to an end project where the "This is good enough.? I want to make chips now!"? So then the drawings are never completed.

?

And it's amazing how much you will learn just duplicating something as seemingly simple as a compound slide.

John

?

?

?

I think because I made the Gingery lathe and a lot of other bits and pieces the practice has resulted in repeatable results.? Like a musical instrument, practice, practice, practice.? And I know nothing.

?

The X axis motor mounts are an example of dealing with limitations.? In this case not a large enough crucible for the entire mount plate so I made it in two pieces and bolted them together.? The exploded diagram shows what I wanted to do including the angular contact bearings.

?

The castings came out without any issues.? The one that holds the bearings is a two part since there's the relief on the bottom side.?

?

?

The close-up with the bearing cap and the angular contact bearings shows the surface finish isn't great but isn't too bad either. My green sand is actually quite coarse. ?If I was I using Petrobond or surrounding the pattern with a really fine grit green sand I'd see much better finish.

?

厂迟颈濒濒…

?

The Shumatech DRO made repeatable positioning for the holes a no brainer. ??I use the ELS, turning my rotary chuck, to position the holes in the hub.

?

It all fit together exactly as designed.

?

What I like about my Gingery furnace is that the small ATMEL micro-processor that runs the ignition coil and PWM speed control for the fan is completely turnkey.? I walk up to it, plug in power, turn on natural gas, lift and pivot the lid and press the start button.? The Fan comes on at a very slow speed, then spark plug starts making buzzing noises, and then the solenoid that allows the gas through opens and there's a bit of a whump and a very rich flame yellow is coming out the top of the furnace.? Then the fan accelerates up to burn speed and the flame pulls down into the furnace, becomes blue and the noise level increases slightly.? Between 12 and 15 minutes later I have about 5 pounds of melted aluminium.

?

Switching it off is much the same.? Turn the switch to off (center position) and the solenoid goes out and the fan blows for a few more seconds to evacuate any fumes or gas. ?Then it shuts off.? I can then lift up the body and pull out the crucible.? By now I've also done any degassing etc.? Once the metal is poured the crucible goes back into the furnace, and the master switch is turned to the other ON position which runs the fan at a very slow speed to slowly cool down the interior.? In the winter it warms the shop up nicely.

?

One day I may add the flame detector.? The software is there for it but that requires some reconstruction and I've just not had the time.

?

John

?

?

Lots of draft?? The green sand was just the nice consistency.? Rammed hard enough.? Then tapping in all directions so that I could pull it straight up.? It was done on the 18FEB2018 so almost two years ago.? And I still don't have my big crucible and tongs done.? Some projects seem to take forever.?

?

But things should improve as the mounting plates for the Mill Spindle will also need this larger crucible so I'm more motivated.? Spindle via CNC control is far more important than the 4^th axis which doesn't even have a face plate for mounting a chuck yet.?

?

John

?

?

?

I'm working on that concept.? I still think that it will be either a Raspberry Pi or a BeagleBone Black since running LinuxCNC (MachineKit) under the covers is the nicest solution to creating an ELS that with the flip of a switch so to speak is also full CNC.

?

However, I don't know, and until I have LinuxCNC on a lathe I won't know, if you can manually turn the spindle by hand and have the lead screw track.? If that's a critical point for end users they are better off to stay with the hi res encoder generating step pulses to the lead screw.

?

Where that feature is really useful is when you want to mount something threaded back in the lathe and line up the cutting tool with the existing thread.? I've done that with my ELS by using the compound to tweak where the tool appears to be as it makes a pass.? Done without the tool engaging in the thread it's pretty easy to get it close.? But precise probably not.

?

And there's so much backlash in my half nut that rocking the spindle back and forth will snap the edge of a carbide insert in a heartbeat as the spindle moves but the lead screw doesn’t? Even with backlash compensation.

?

John

?

?

The back plates were stamped and bent aluminium with some pressed in threaded inserts.? They weren't overly expensive but even at about $12 with setup costs I was still holding onto over $1000 worth of parts.

?

I'm not ready to try lost PLA yet.? Can't stand the smell of lost foam so I don't use that.? The problem with lost PLA is twofold:

1.????? Often takes a long time to print something and if the casting fails that's a lot of hours lost what with foundry setup and pouring time too.

2.????? 3D prints are generally pretty ugly.

?

Having said that though, here's my A axis mount for the mill.? First printed in 4 parts and glued tougher to give me an idea of what a finished machined cast part would look like.

?

Then I expanded the 3D drawing and added draft etc. so it would pull from the sand.? ?I printed that, added body filler, primer and paint and did a trial fit in greensand.? Now just to finish my larger crucible and tongs so I can melt the 14 lbs I need for this.

?

?

http://www.autoartisans.com/mill/HarmonicDrive/TrialSand.jpg

?

There's no way I'd want to melt out this PLA and have the casting fail.? Took way too many hours to make a pretty pattern.?

?

John

?

?

?

Good reference on the issues of micro-stepping.? I believe that is one of the reasons why Gecko switches over to full step above the resonance point.? But I've been trying to tell people for years that you can micro-step a motor as much as 4 micro-steps ?(10 steps/step) before the motor will even move since it has to overcome static friction. ??So I totally agree that a belt reduction drive is a better way to get the accuracy.

?

But then there is the problem with steppers is they just can't turn as fast as you need them to when you add in 1:2 to 1:4 reduction.? Hence that's why I chose to use DC Servos on my mill.? So quiet and smooth too.

?

John

?

?

I have a few of each of those.? Now if I only had the time to play with them.

John

?