John D Thanks for a concise answer. ?
John L
It’s been 13 years.? Let’s see if I can remember.? Each turn of the stepper motor is 10 degrees or 160 steps per degree.? For 13 divisions in a circle we need 27.69230769230769 degrees or 4430.769230769231 steps.? But that’s a hard number to figure out. ? Distances on the ELS are set up as Motor steps per rev and lead screw pitch and finally distance per jog button. ? So 1600 steps per rev is fixed by the micro-stepping driver.? One revolution means 10 degrees on my rotary table so I’ll call that a Pitch of 10 TPI or 0.1” per rev or 0.01” per degree since one turn is 10 degrees. ? I have to move 360/13 for every jog so that’s 27.69230769230769 ?degrees. ??Multiply that by our 0.01” per degree and I enter into the jog distance 0.2769” which is the maximum resolution I’m allowed for entering numbers; not need to round up.? Once the enter key is tapped it shows to 3 digits or 0.277 but internally is still 0.2769” per jog press. ? With the table set at 0 then pressing the jog button 13 times brings me back around to 3.599 on the display.? ?The marks on the rotary table look like they are at exactly 0 again.? I think from a cumulative perspective over many turns the error would add up.? On one turn I doubt we’d see a difference. ? A test might be to use a gear cutter on a disk to cut 13 slots.? Then couple that to an encoder with 1000 lines or 4000 edges per rev.? Then measure the gap and tooth width in encoder counts. ? Then do the same thing with the arm and disks etc. the old fashioned way on a rotary table.? My guess is my rotary table (made in India) likely has imperfections in the worm and drive gear and the errors wouldn’t be any different. ? John Dammeyer ? ? ? ? ? ? Understand, but if you needed 13 divisions?? -------- Original message -------- Date: 5/16/24 07:09 (GMT+01:00) Subject: Re: [digitalhobbyist] #CNC gear cutting I took the easy route 13 years ago.? My ELS has the stepper motor drivers inside.? I connected to the rotary axis motor and configured the Lathe Z Axis parameters so each single jog step moved the number of degrees required for what I was doing.? When I was done I set the parameters back for the lathe and put it back on my South Bend. ?
? John Dammeyer ? ? This is similar to what I built, I attached a belt and pulley and not direct drive. The Arduino sketch composed by Gary Liming is the same. ? Cool thanks . I built one of the Arduino rotary tables from one of the forums but never got it to work , pretty sure it had something to do with my stepper driver boards , same thing with the L'il MaxNC that I tried to convert over to Arduino using a CNC shield . I wanted to get setup to make graduated dials , but last I looked into it Grbl didn't support a 4th axis . Mach3 just looks like it's too complicated for me at this stage . I do have the 4 nema 24 steppers that came with the Max, so I know that they are up to the task . thanks animal On 5/14/24 1:23 PM, John Lindo wrote: I have a mini Weiss 16AV mill that I fitted with Nema 23 steppers to the X Y Z and a homemade indexing?head that also has a Nema 23 that I use for the fourth A axis. The mill can be used for simple power feed from a control box connected to push button feeds and rapid mounted on the mill,? also a different separate set of cables to the steppers by the way of 4 x quick release plugs and these feed into a control?box (drivers etc) that receives info from a CNC DDCSV ver 2.1 box. Photo 1 CNC gear cutting. Photo 2 The Red box, Power feed unit with DM 542 steppers and power supply boxes, the box underneath? is the steppers and power supply for the CNC. Box boxes initially were originally tea storage boxes from a Chinese shop, the crystal glass was taken out and I replaced them with clear perspex.? No room in the workshop bench areas, so I have to use the walls and ceiling for any available space.LOL. Photos of power feed X Y Z operators controls. Photo of a homemade Arduino step indexing control box, It has several setting commands, set Ratio, (currently I use 1:1) IE 45T stepper 45T indexer, but I can select? 90:1 45:10 all depends on the project,? step Jog, Continual run, Angular moves per any full degree selected and a Division module. A very useful piece of kit in the armoury. I have the .ino code also a complete "how to" build info file. Hope all of you are interested. ? Nice Job ! This is why we do things like this , to see if we can . I forget what is yer CNC machine ? thanks animal On 5/14/24 10:51 AM, John Lindo wrote: Hi Mike and members.
Finally got there with the gear cutting of the 9 T chuck pinions, I think LOL.
Only cut one so far, but entering the pinion into the alloy chuck scroll ring/gear, the pinion rotates without?
any jamming etc. The acid test will be to cut the other 2 pinion blanks and see if they all marry up.
To remove the arrow head created by cutting down a cone, indexer set at 20 degrees, the gear cutter running down the front?
portion of cone IE parallel to the X axis, cut 9 divisions, 40 degrees rotation, thus creating the arrow head effect between the teeth.
See the 12 line handwritten g code, basically I? tell the machine to do a function, in this case repeat 9 times.
This is a very useful code for simple gear cutting.
Then with the MPG, CNC, manual pulse generator, hand wheel, rotated the A axis +4 degrees, set A to home 0 degrees and then ran the programme again,?
then with the MPG set? to -8 degrees , again reset? A to 0 and ran the programme again.?
I could of consolidated basically 3 programmes into 1 but decided not to risk scrapping the phosphor bronze.
I agree that there will be so many armchair engineers who will disagree with this approach to making basically an old mangle type set of bevel gears,
and yes the 2 cutter approach as seen in your video you attached is the way to go if it was a crown wheel and pinion set for a Formula one drive train.
but this works for me and I am confident that the next 2 pinions to be cut if I follow what I just did on the first one then I should be home and dry.?
see attachments.
--
John
--
John
--
John
|
John Lindo