Identifying Helical Spur Gears

[ozzie34231]

Hi Art, John S,
I want to make my own Sjogren type, 5c collet chuck. As I'm sure you know the working part of this is a set of planetary gears. I bought a helical set originally used in a Chrysler transmission. The sizes look very good for the chuck but the sun gear must be threaded internally and it is hard, very hard. I made a futile attempt at softening it, no go. It wears away a carbide boring bar.
So, no problem, I'll use Gearotic and cut a new sun; soft enough to thread. So Art, your post up a couple, tells how to size it if its a module gear, but how do I know if it is module or whatever the alternative is called?
I'm thinking that there might be a few folks that want/need a short tutorial on identifying and then copying a given gear.
Thanks,
Jerry "Ozzie" Pryor
Old Curmudgeon

[JustinO]

Jerry,

My copy of "Machinery's Handbook, 13th edition" (1948), has over two hundred pages on gearing. There is a lot on helical gears, "checking gear sizes", and explicitly discusses measuring helical spur gears with gauge wires.

Justin.

[ArtF]

Jerry:

>>but how do I know if it is module or whatever the alternati ve is called

  All gears are actually Module gears. ( And DP gears ). The DP is simply the  (25.4/Module) and the
Module of a DP is (25.4/DP); I do everything in Metric, then convert to standard DP when I need to use
non metric numbers for anyone.

    If you use the above letter to figure out the module, and angle of a helical, you should do fine.
Just use a DP of 25.4/Module if you wish to do it in Imperial measurements. Helical angles are
pretty easy, just use a protractor to follow the angle from front of gear to back, ( usually 45 degrees on mating helicals..its the most efficient angle for a helical) , Pressure angles are almost always 20 degrees.
  Helical's may be profile shifted though, or have other modifacations, but if the standard calc's seem to agree , Id assuem its a standard gear.

Art

[ozzie34231]

Many thanks,
Ozzie

[ozzie34231]

Hi Art,
Well I thought I understood it perfectly.
First it is difficult with this size gear to measure the contacting radius or diameter. Since there are 4 planets I did my best to measure the diameter. I got 37--38 mm. 1 1/2" would be 38.1.
There are 29 teeth, counted them 3 times. The helix angle is 30 degrees.
I thought I would get some near integer in either metric or imperial; no such luck; (at least with the math that you gave, which is all I have).
Playing with those numbers I get something like module 1.13. Using that I generate a gear that seems a tad small when I measure the real gear against the screen specs. Overall diameter is small, total depth is a bit small.
If I use a module 1.2 the numbers look a lot closer, now a shade large, but not much.
So I ask, should the numbers come out in a nice integer or are my calcs probably correct?
I was not able to accomplish anything by trying to switch the program to inch mode. I suspect some problem there but not sure.
Another problem I foresee is that the rotary axis is calibrated in steps per degree, (500). Doesn't that mean that any combined feed rate is going to move ponderously slow?
Should I set up a new profile in Mach 3 using Metric?
Ozzie

[JustinO]

Ozzie,

Are you measuring center to center, sun to planet, and using the proportionality of the teeth numbers to determine your pitch radius? If they are not shifted or in some other way optimized, you should get pretty close. If you can measure the CC distance with the gears in their frame, then you'll be including the backlash. If you just push them together without their frame, you won't be including the backlash.

Converting back and forth from regular spur to helical spur will give you funny numbers somewhere. Trigonometry tends to do that.

Are those 500 microsteps? My machine (Sherline) does 320 microsteps per degree and has no problem.

--Justin

[ozzie34231]

Hi Justin,
The planets are permanently installed in a cage, shafts welded in place. I placed the sun in place and measured the diameter of the apparent points of contact. There is some backlash, not much. The idea of proportioning distance between shafts sounds interesting. I could measure across the fixed shafts of the planets and then portion by tooth count, two half planets and one sun, needing to compensate for the backlash.
However I think I'm not terribly off since the program creates a gear with dimensions close to what I actually measure.
This might be more of a theoretic problem than practical. The planetary system does not really run for this application, it only tightens a collet!
So unless Art, (or Sir Stevenson), sheds some additional knowledge on the situation, I think an aluminum test gear will be the next step.
---
My rotary is servo driven, 2000 steps per revolution, direct drive to the rotary worm, four degrees per rev, (90 to 1). so one degree requires 500 steps. I'm presently set up in inches with rapids conservatively set at 100 inches per minute on 3 axis. To match that speed on  a 1 3/4" diameter on the rotary I need about 18 revolutions of the rotary per minute. So I need to go almost 6500 units per minute. Now if I cut at 10 inches per minute what does the A axis do? Does it cut at 10 degrees per minute? !!
I don't know, I've only used an A axis to rotate at rapid speed to position for a next cut. 3 1/2 axis?
Ozzie
Ozzie 

[ArtF]

Ozzie:

  Depends on how your set up.. What it SHOULD do is ficure the outside circumferance of the
A axis material, from there figure out linear distance around that cyclinder and then mix feedrate
with the X and Y so that the Feedrate is properly in linear units/minute.

  Mach3 does that if you set a diameter for feedrate....

Art

[ozzie34231]

Hi Art,
I've been using Mach for 10 years Art and did not know that. So I would think that if I have a diameter set for A, and I command "G1A1" a point on the circumference will move one inch, rather than rotate one degree?
If that is so, (a bit OT), all one would need to engrave on a cylinder, (wrap), would be to first "swapaxis" A for Y? And of course set the diameter.
Ozzie

[ozzie34231]

Help!
I just came in from the shop and I saw that what I said above is totally wrong.
So, I need help. Please explain in detail what setting the rotary diameter does in Mach3. I tried the docs, still don't understand.
Ozzie

[JustinO]

Ozzie,
About three quarters down on this page:
http://www.machsupport.com/forum/index. ... ic=23282.0

Is that what you were looking for?

--Justin

[ozzie34231]

Yes Justin, Thanks a lot.
I looked all around for the check box that is in the toolpath setup, didn't find. So then I figured I just had to put a number in the appropriate dro, that seems to be what the docs imply.
Seems like a strange place for that check box to be.
Tomorrow I try cutting some air again. Then maybe some aluminum!
You cool Justin, ,,,cool.
Ozzie

[ArtF]

Ozzie:

Sry, didnt get notification on this thread again...

  The feedrate will be calc'd correctly for the motion of the A, but its still A1 is 1 degree, not one inch. BUT, when you command G1X10A3, Mach will calculate the actual distance moved and ocmpute feedrate based on that, so the Higher the Z , the larger the distance traveled for that A3 and the higher the feedrate will be.

Art

[ozzie34231]

Thanks Art,
That's neat, not my wild hope but cool nevertheless.
I see I'll need a mill about 1.25 mm.
Regarding my post a few back beginning "Well I thought---"; is it unusual that I come up with a module that is not an integer and does not seem an even DP either? Is cutting a test gear my best solution?
Ozzie

[ArtF]

Ozzie:

>>is it unusual that I come up with a module that is not an integer and doe

  Not unusual at all. While gear specs are published everywhere, many are "Tweaked" by their maker in various ways, and that can throw off
attempts to measure them. For example a mod4 profile shifted gear may measure 4.12Mod if you try measuring it and dont know its profile shifted and compensate. It can get very complex to do helicals and bevels in that way.
  As you'll see when you play with GT now, it automatically profile shifts a gleason gear for example and the resulting gear isnt the size
youd think from quick calcs. Such is the way of gearing.

  I do intend to add a screen in the new year that will allow you to measure a gear in various ways , enter the measurement and have GT figure out what kind of gear you may have and suggest how to duplicate it. I can see the facility for such a thing, but in the meantime, have no qualms about making one a weird module.. nothing wrong with that at all. I too look for "magic" numbers when working things out, but sometimes they just arent relevent. :)

Art