Speaking of making gears.....

[Paul Alciatore]

YES! This is what most of the descriptions of the button method do NOT ever tell you.

The circle and the involute ARE DIFFERENT curves.

The button method may be OK for slow speed and/or low power gears, but if you are not in both of those areas of operation (slow AND low power), then you can have problems. The buttons make gears with ROUND faces, not involute faces.

And that video actually tells you almost nothing. You need the carefully calculated numbers. And the cutters produced are much like the numbered gear milling cutters, only good for a small range of tooth counts. The diameter and spacing of the two buttons are both quite different for a 10 tooth gear and a 100 tooth gear of the same DP. One tool, button or commercially purchased and numbered involute tool, will not cut both of those tooth counts. And for both types of cutters, the errors get greater as the tooth count gets lower.

If you want an easier way of making gears then look into hobing them. One hob cuts all tooth counts for a given DP. Well, almost all.

Browne & Sharp cutters vs those made with round buttons? My money is on Browne & Sharp.

The round buttons are an approximation. They generally do work, I understand, although I have not used them.

Whether the approximation is better or worse than the use of a Browne & Sharp type cutter over a range of tooth count, I do not know. There is a difference between a circle and an involute curve, but the difference can be subtle in specific cases.

[djc]

...if you use the right table---and there are a few out there

You have to remember that all the tables were produced in the days pre-computer, where the maths was cranked out longhand.

The tables differ because the assumptions behind them differ and there is more or less simplification of the maths from beginning to finsihed product.

These days, there is no need for tables. You can work out arbitrarily many points that are on the actual involute curve of the specific gear you are making and then calculate a best fit circular arc for that involute. Your program will output the centre and radius of the arc to use.

That is what Mr Bilhardt did in 1990 on his IBM PC. Computer power and programming language ease of use have developed since then.

He had a particular reason for it - CNC production. There is no G-code for an involute move. G-code interpreters know straight lines and circular arcs. So at the machine control level, you have to pick one or the other to drive the cuttter around an involute curve. A straight line approximation will have gazillions of segments, making the code extremely long. A circular arc approximation of the involute is one line of G-code.

Even the Van Keuren handbook, a masterpiece of its day, is obsolete now. Have a look in the back of it for pages and pages of trancendental equations on over-wire measurement of screws. Someone on PM forum has written a PC program that does all that work in the blink of an eye.

[RSG]

That's pretty cool BCRider!

I made a copy of his fixture for my cycloidal cutters for the Tourbillon build and was going to use his button idea if I wasn't successful grinding the correct rad in a piece if HSS.

[J Tiers]

YES! This is what most of the descriptions of the button method do NOT ever tell you.

The circle and the involute ARE DIFFERENT curves.

The button method may be OK for slow speed and/or low power gears, but if you are not in both of those areas of operation (slow AND low power), then you can have problems. The buttons make gears with ROUND faces, not involute faces.

And that video actually tells you almost nothing. You need the carefully calculated numbers. And the cutters produced are much like the numbered gear milling cutters, only good for a small range of tooth counts. The diameter and spacing of the two buttons are both quite different for a 10 tooth gear and a 100 tooth gear of the same DP. One tool, button or commercially purchased and numbered involute tool, will not cut both of those tooth counts. And for both types of cutters, the errors get greater as the tooth count gets lower.

If you want an easier way of making gears then look into hobing them. One hob cuts all tooth counts for a given DP. Well, almost all.

Browne & Sharp cutters vs those made with round buttons? My money is on Browne & Sharp.

You are "somewhat wrong" with that description.

The button method DOES NOT USE a "circle" as the approximation as actually cut. You have to remember the tilt of the circular button. That then becomes not a circle, but an ellipse.

An involute is more similar to an ellipse than a circle, so tilting the buttons then improves the match of the generated shape to a true involute

[Rich Carlstedt]

.................................................. .......That is what Mr Bilhardt did in 1990 on his IBM PC. Computer power and programming language ease of use have developed since then.
He had a particular reason for it - CNC production. There is no G-code for an involute move. G-code interpreters know straight lines and circular arcs. So at the machine control level, you have to pick one or the other to drive the cuttter around an involute curve. A straight line approximation will have gazillions of segments, making the code extremely long. A circular arc approximation of the involute is one line of G-code.
Even the Van Keuren handbook, a masterpiece of its day, is obsolete now. Have a look in the back of it for pages and pages of trancendental equations on over-wire measurement of screws. Someone on PM forum has written a PC program that does all that work in the blink of an eye.

djc- i suggest you read my post (#16) on another thread here at HSM

https://bbs.homeshopmachinist.net/forum/general/2039505-what-year-were-you-first-exposed-to-numerical-control

The GE 2000 CNC control in the 1980's (!) could do an involute curve in one line of code !
Didn't take thousands of g code lines or MasterCam to produce the curve that is needed today. -But we sometimes think today is better than yesterday

I think you fellows are getting way outside the box when it comes to gear design and home shop work
Let me explain because some here feel that the true involute is the perfect answer for perfect gear operations...IT is not !
When gears run together and say the driver ( Master) is perfect in "constant" revolutions, the follow gear is not a perfect revolution replica !
it's speed varies from EXACT revolutions to the Master and the variance is related to pitch used and tooth form
Lets look at TWO Masters
So you have a big 6 Pitch Master gear and the same diameter 20 Pitch Master gear and they both drive slaves the same diameter
Which Slave runs smoother, or should I say " With Less Variation" to rotation ? why the 20 Pitch of course
The reason ?
Well, two smooth tired wheels engaged with each other will have 100 % replication of rotation
But that relationship is limited by friction and to overcome slippage, gears were invented.
The smaller the gear teeth (Physically) , the closer to the ideal replication of movement occurs ...
BUT if you want more horse power, you need to make a choice
besides material , you have two choices, make the gears wider ( called face) or go with a larger tooth
But as you get larger teeth, you encounter greater speed variation...so thats how gears are designed.
The Involute is not perfect even though it is close
so worrying about how close your button is to a true involute is nice...but over done.
As Jerry said, the oval cut is closer than a round cut and even a Hobbed cut as Paul mentioned using ,
is still only an approximate relationship to what two smooth wheels do.
Rich

[BCRider]

That's pretty cool BCRider!

I made a copy of his fixture for my cycloidal cutters for the Tourbillon build and was going to use his button idea if I wasn't successful grinding the correct rad in a piece if HSS.

Nice! What sort of radius were you shaping when you did the forming cuts on the HSS blanks? Or did you rig up a way to swing the blanks past a wheel to achieve the radius with control?

Actually that last thought gives me an idea It would not be overly hard to rig up an arm on a pivot and clamp the square blanks at the end and use that to form the larger sizes for high count gears where the button method is not really practical. I'm thinking on gears with 50 and more teeth where the curve is very subtle.

[BCRider]

Just to add that this topic sure did open up the flood gates. It has our usual bickering going on but I've learned more about making up single point cutters for gears and making gears in general than some of the other sources I've read on a casual basis. So well done all of you!

I still have to go and read the links given earlier. And still need to watch the other half of Huub's video. Real life you know....

[RSG]

Nice! What sort of radius were you shaping when you did the forming cuts on the HSS blanks?

First I used a Cycloidal gear generator to address the correct shape for the cutters, then laser cut the outline on the top of the pieces of HSS, then ground them by eye. I tested them, then made the final tweaks including stoning them before running them. Pics below.

[BCRider]

Bringing this back up because Clickspring just posted a short on how to make form relieved four tooth cutters. And I instantly thought about the two button cutter trick in combination with the form relieved cutter idea.

(2) Making A Form Relieved Pinion Cutter - YouTube

If we combine the round button option with his relieving method we end up with four tooth, or more, gear cutters that would fit on a custom made arbor. Anyone tried this idea before?

[Rich Carlstedt]

Bringing this back up because Clickspring just posted a short on how to make form relieved four tooth cutters. And I instantly thought about the two button cutter trick in combination with the form relieved cutter idea........ (2) Making A Form Relieved Pinion Cutter - YouTube.............
If we combine the round button option with his relieving method we end up with four tooth, or more, gear cutters that would fit on a custom made arbor. Anyone tried this idea before?

Lots of guys have done that including our own Sir John Stevenson
From the archives of Metalweb news

https://raremetalblog.com/metalwebnews/howto/gear/gear1.html

there are many pics that john published showing the offsetting of the blank like the URL you provided shows

rich

[DaveR]

Bringing this back up because Clickspring just posted a short on how to make form relieved four tooth cutters. And I instantly thought about the two button cutter trick in combination with the form relieved cutter idea.

(2) Making A Form Relieved Pinion Cutter - YouTube

If we combine the round button option with his relieving method we end up with four tooth, or more, gear cutters that would fit on a custom made arbor. Anyone tried this idea before?

Starting at post 1140:

https://bbs.homeshopmachinist.net/forum/general/60104-what-did-you-machine-today/page29#post1717662

[Paul Alciatore]

Sorry, but I couldn't let you get away with that.

One line of HIGH LEVEL code hardly tells the whole story. That one line probably called up a machine code routing that had quite a few lines in it.

And, on top of that, there is no guarantee that it was actually generating an involute curve. Since you never saw the actual code (the machine language routing), it may have just pasted together two or three circular arcs for a somewhat close approximation.

djc- i suggest you read my post (#16) on another thread here at HSM

https://bbs.homeshopmachinist.net/forum/general/2039505-what-year-were-you-first-exposed-to-numerical-control

The GE 2000 CNC control in the 1980's (!) could do an involute curve in one line of code !
Didn't take thousands of g code lines or MasterCam to produce the curve that is needed today. -But we sometimes think today is better than yesterday

...<snip>....
Rich

[RB211]

I just 3D print them if they are a conducive size for the process.

[Paul Alciatore]

Yes, I saw that method of providing the relief some years ago, probably here on this BB.

Here are two, slightly longer Clickspring videos that show the process with the button cutters:

https://www.youtube.com/watch?v=XJxxIRCe4B4

https://www.youtube.com/watch?v=GlAm9eva_fQ

If I were to make a cutter with the button method this is how I would provide the relief. It really is clever.

Bringing this back up because Clickspring just posted a short on how to make form relieved four tooth cutters. And I instantly thought about the two button cutter trick in combination with the form relieved cutter idea.

(2) Making A Form Relieved Pinion Cutter - YouTube

If we combine the round button option with his relieving method we end up with four tooth, or more, gear cutters that would fit on a custom made arbor. Anyone tried this idea before?

[BCRider]

Apparently this is one of those older tricks that I had not seen. Or passed by and forgotten.

It's still pretty cool. And might serve others that have not seen it before. Or perhaps like me had seen it but it passed by way long ago?