Gear Hobber : Design ideas please ...

Agree. But Joe can't possibly believe Windows is an easy platform to program on can he? He must be comparing Linux with something else entirely, perhaps an embedded environment that has no operating system at all. Programming hardware is much easier when you don't have to share the computer with other users and processes.

Dave

Another thought i had ( yes it is true that you only need two brain cells to think! ) is to use a single pulse generator feeding two identical steppers so in theory they should be rotating at the same speed . One drives the hob the other drives the gear blank but the gear blank mandrel is not directly driven like the hob spindle is .

The gear blank mandrel is driven by a worm wheel that is the same diameter and pitch as the hob and this is driving a gear which is the same tooth count and diameter as the gear to be cut . This should give the correct drive ratio required to cut the gear and the dummy gear and worm wheel would be drawn up in cad then 3D printed , once the hob has cut deep enough to free hob the drive system on the gear mandrel could be disengaged if needed .

Once again just a thought !

If there were going to be missing steps, they would occur on a stepper driving the hob, that has an intermittent load. The gear drive itself should only have friction to deal with, since the gear PD is going to be moving at exactly the same speed that the hob pitch is moving at, so there should be no torque imposed on the blank by the hob. On that basis, it's probably simpler to use a conventional drive and a rotary encoder for the hob and divide that down for a stepper to drive the blank.

Another advantage of the electronic solution is that there are no universal couplings, bevel drives needed, just spur gears (or straight helical gears, once they've been made) to turn the 200 steps/rev of the stepper motor into enough to give smooth rotation of the blank . Also the gears can be whatever changewheels are to hand initially.

Again, the reason for the electronics is to avoid the need for any specific collection of gears in order to create the gears you want to make.

Edited By Mark Rand on 09/08/2018 22:34:32

Missed step detection is, as I recall it, relatively easy with the electronic PLL based control system. Encoder on the output feeding a second PLL and count both output and input steps using a simple grey code divider chains. Pulse outputs should be temporally aligned so any miss is obvious. Recovery isn't quite so simple but it can be done.

System I made way back had this capability, including automatic recovery, but for different purposes. I used both halves of a CMOS Dual PLL chip and monitored the delay between the grey code counts to tweak the output of the voltage controlled oscillator driving the stepper so it could track a moderately unstable "axis". Including ramp up to speed from stationary over a somewhat undefined period of a few seconds varying by up to 50%. The major issue being persuading the PLL to operate over a too wide frequency range.

I'm really annoyed that I've lost the circuit because it would have done for a fair few things. Ideal for Michaels job. Simple too as it was built as part of technology demonstrator for a "£100" sensor and control package for terminally guided submunitions.

On the microchip user side I wonder if there is anything useful in this book :- **LINK** . As I recall it Elektor projects had the reputation of being reliable and well thought out from the "doing what it says on the tin" viewpoint.

Clive.

Hi Michael

Have sent you a PM

John

Hello Michael,

here is a solution that works.

My homemade gear hobber

("https://www.homemodelenginemachinist.com/threads/my-homemade-gear-hobber.30291/"

He cuts straight aluminum gears with a mini mill , a homemade hob , uses a Due and no stepper and the software is included in the posts.

I found this by typing "gear hobbing arduino" in google

br

Walter

Many thanks, Walter ... I will have a look at that.

Joe Noci was very supportive on the electronics side ... but I regret to say that my hardware costs were escalating out of control, and I have shelved the project for the moment.

MichaelG

M choice would have been a Dividing Head, or Rotary Table on a mill, as long as the capital cost for a variety of gear cutters would have been within budget. Having said that a set of eight cutters (probably about £160 ) would cover every tooth count from 12 to a rack.

That way, no worries about the electronics being upset by power surges, stray electric fields, or not being temperature stable.

No wonder I am a mechanical engineer!

Howard

.

I could have done that, Howard ... but I fancied a hobbing machine, for its ‘purity’

MichaelG.

Know what you mean Michael!

A hobber is one of my "love to haves", but not justifiable on grounds of cost, likely lack of use, or availablity of storage space.

Life ain't always fair!

Howard