Tangential Saw shaving 4th axis

[David]

Hi Art

Some of the issue must be virtual machines in my case both VMware and Virtualbox both on Macs.
I have zipped files up for you to see.
The gth file name ending in macVM was produced on my Mac, the tap and screen shot  on the win7 PC I use with Mach3
The files ending in win produced solely on the win7 Mach3 PC
You can see the A90,s in the macVM files, but those gth file created using the same gear parameters on the win7 PC are fine.
Never seen difference between VM,s and PC before. I use 2 different Macs each running a different VM both produce the A90 problem.
As I store files on a NAS I was not really aware were I had created them until I found no difference reducing saw width and you finding no problem when you used the same settings. 
I think I have unwittingly caused you a lot of work, sorry. Perhaps you could ask if anyone else has tried running Gearotic on a VM and had problems?

David

[ArtF]

David:

  Not at all. I have just uploaded a fix to correct it. I dont know why the gear tooth form created by mac's were slightly different
than windows, but they just crossed under the zero line by about .00001 units and that was enough to create a bug.
Try the new version, the toolpath should look much better...

Art

[David]

Hi Art

You certainly work quickly
Tried latest version in both Mac virtual machines, looks good (see attached screen shot) and no A 90 degrees.
Its gone midnight here so will try a test gear over the weekend.
I am sure you have had enough of me, but do you know if anyone has tried to use your program to produce say a four lobed relieved involute gear cutter?
I am thinking using a standard size endmill to give the cutter profile and cnc motion to give relief. For the larger gears it would be easy enough to find the correct endmill size.
I wondered if a blank was gashed quite widely producing 4 lobes then when correctly aligned it should not be too difficult for the cutter to be correctly relieved.
Please feel free to ignore those thoughts you must have plenty to do!
David

[ArtF]

David:

  Ahh, that looks much better.

  Im not aware of anyone doing a cutter that way.. I have been asked and have promised to make the program
put out code to make cutters, so its something Ill have to look into eventually..

Art

[JustinO]

I can't seem to get the guy who wrote SolidWorks to pay any attention to me. Just as well; their spur gear tooth faces are made from one circular arc.

You're doing well, Art.

Thanks,

--Justin

[David]

Hi Art

Had a couple of hours, so started by finding a newish 0.016" saw, a better mandrel and some supporting discs (leaving about 0.2" teeth exposed)
I am still using the over long and unsupported mandrel for the blank, the blanks are the cheap hardware washers.
The teeth (72) are 0.6 module, at the base circle tooth width is 0.0325" and the tooth space is 0.037" which given the difficulty measuring is very close to the theoretical 0.0708".
Is the tooth space a little wide? The quality is just superb (IMHO)  ;)
Many thanks for your great software and support
David

[ArtF]

David:

  Nice work!..

Probably fine for clock wheels, but the space is a bit large, Im thinking it would require a bit thinner saw to make it thin
enough. The program is very good at accuracy, but a bit forgiving as to the diam being too large, it will cut anyway and
make the space too wide. Im thinking this is why the spaces look a bit wide.. but I'll have to check the math.. It may just be
the rooting is too wide.. seems I meant to check somethign there..

Art


Art

[David]

Hi Art

Moved on to trial pinions for clock, 8 tooth 0.6 mod. I noticed what may be an error in that for a 72 tooth wheel the OD given on the tools and project screen is the same at 1.762" the material size is calculated to be 1.777" an increase of 0.85%. For an 8 tooth 0.6 mod pinion the tools screen gives OD 0.238" the project screen 0.25" and the material size 0.2649" an increase of 11.25%. Not sure if this is supposed to happen or is an error. In both cases the OD on the tools screen is correct according to tables.
I have encountered a practical difficulty in that a 0.238" blank needs to be turned and then a length turned to say 0.165" for about 1" to provide clearance for the saw especially during A axis rotation . This is a bit flimsy and although I have done a couple of teeth in brass, I doubt the finish would be good in the preferred silver steel. Is it possible to have the cut applying Y axis retract before A axis rotation for small tooth numbers? This would permit cutting on a blank of the correct size without the length turned down  for clearance, but perhaps more productively also cutting of a length of material that can then be parted into a number of pinions saving material and ultimately the planet :D
Best wishes
David

[ArtF]

Hi David:

  That sounds reasonable. I will look into adding this for saw selections. Ill let you know how I make out, might take a few days...

Art

[David]

Hi Art

Had a few days break from gears, drew up some clock plates and machined them on the CNC mill, at least I can see them without an eye glass or microscope.
Back to gears, and tried a 0.55mod 12 tooth 0.125" bore gear in brass still, until you are able to look at the saw retract issue.
Very pleased how it turned out (pics attached) pitch circle at 1.171mm ( 0.066"0 very close to the standard 0.070") given the linear measurement rather than on the arc.
The space is 0.65mm (0.0256") and the tooth 1.041mm (0.041")
Question do you think this is now about right, both in visual appearance and measurement?
Best wishes
David

[ArtF]

David:
 
  The rooting looks pretty good, though the involutes look somewhat pointy.. but when you get
real small things get so critical that those really dont look too bad. I guess the test is if they run..

Art