Sorry if this is a bit long winded but the following text file is a series
of posts and answers to queries over the Electronic Gear Hobber I have just
First off I must give credit to the people who have helped me in this
Ironically enough credit goes to our old mate Gareth who noticed this at
the Donnington show two years ago and that set me going. I managed to find
the original builder, Brian Thompson and went up to Scunthorpe to meet him
and have a chat. I also bought one of his old test units to try, pressure
of work and a search for a decent gearbox took until a couple of months
Anyway sorry for the long post but the last two weeks have been very
Test run of the Stevo Eee-lec-tronik hobber.
Managed to get the stepper driven head mounted on the horizontal mill table
in between customers calling and ringing up and generally being a pain.
Finished tidying the wiring for the driver box and connected this to the
head and the arbor driven encoder.
Found some scrap brass blanks and a decent 12 DP gear hob so I turned three
blanks up all the same. Fitted the blanks to the stepper head and set the
table over by 3 degrees 20 minutes which is the lead angle of the hob.
Set the depth, wound clear and started this up, amazing to see it start
cutting and all the while you are expecting this to run all the teeth off.
First one off had wavy teeth because I'd forgotten to tighten the blank up.
I won't mention what happened to the second one After all this is ONLY a
test run, isn't it ?
Turned all three blanks down to get rid of the mistakes and restarted using
fewer teeth No problems this time.
I've put some pics on my web space, 10 in all
[Edit - now 19 at this point ]
Just keep changing the last number from 1 thru to 10
1 and 2 are the head being built up and 3 is the driver box rated at
67 volts and 4 amps.
Pic 4 is an overall view of the setup on the Victoria mill.
Pic 5 is a close up of a 12DP 21 tooth gear being cut.
Pic 6 is a close up of the finished driver box with plugs fitted but before
Pic 7 is similar to pic 4.
Pic 8 is a nice one. It shows three gears I cut tonight, all the same size
OD but they have 20, 21 and 22 teeth on them and they all mesh together
perfectly, only the 21T is the correct standard. These are 12 DP cut in
Pic 9 is a close up of the finish obtained.
Pic 10 is a 20DP nylon gear of 52 teeth. Ironically I cut this for a friend
with a Micron hobber who's missing a 52 tooth gear and he needs a 52 tooth
to make one !!
Got around to doing a bit more thinking and playing tonight.
Next step from doing spur gears are helicals.
Thought this one was easy, just add the helix angle of the hob to the helix
angle of the gear and spin the table that amount.
WRONG - mucho ripped up teeth. Go back to reading the book and it appears
you need a differential feed between the blank and the table. Well if you
need gears this does away with all this He - Lec- Tronik gismo's although
you at least can cut your own special spur gears without a master.
Two nights sleep, long bath and a read of a 1913 gear book and all becomes
Spin the table to the helix angle of the hob, this to all intents and
purposes straightens the hob up so it's neutral cutting. Then spin the DH
over to the helix angle of the gear.
Quick test run and Da-Da:-
helix, shaken not stirred.
it cutting, ignore the funny hob as my digital can't stop moving
Final proof is to spin the DH over the other way by the same amount and cut
the opposite hand. If these are correct they will lie together on a level
Further playing and you get this:-
gears to remove the side thrust associated with helicals.
Last play was some course gears to show the effects of undercutting on low
two outer gears are 8 DP 14.5 pressure angle and 10 teeth, note the
heavy undercutting that occurs. The centre gear is done with the same
cutter, same blank but decreased from 10 to 9 teeth, note the heavy strong
tooth shape produced. All these gears will run together in any order
Further to my playing, well actually testing I did three crossed helicals
tonight, these are helicals where the helix angle is 45 degrees and
depending on whether you run right with right or right with left determines
if the axis are in line or at 90 degrees to one another.
I was setting these up I took a small movie of one being cut. Sorry
about the quality as my digital camera is now showing it's age.
I realise that some of this is a bit overkill, the big toroidal and the
motor but this was done purposely because I wanted to run decent sized hobs
as opposed to small model engine type ones and if I had problems I wanted
them to be of my own making and not have to work out if I had a setup
problem or an equipment problem.
What I have done can be scaled down to suit small work with no problem, I
wanted something that could handle something like change gears on medium
With the exception of the stepper and the gearbox which were bought in
special for the project everything else was off the shelf or scrapbox
related. The driver I used is an API CMD 260 stepper driver similar to a
Gecko but rated at 80 volts and 8 amps fully protected as opposed to
Gecko's 7 amp unprotected. I managed to get 8 of these off US Ebay a while
ago for not a lot of money.
I bought the gearbox in as I wanted a 20:1 box to match up with the maths.
In the end I had to get a 40:1 and gear it up 2:1 to get the required 20:1.
I could have used my big Hoffman but that would have meant altering it's
input 2:1 to get the 20:1 needed but that piece of equipment is in regular
payable use and I couldn't spare the time.
At least this way the hobbing setup is all stand alone and only requires
the horizontal mill as a donor machine, not much of a problem as this
usually only works for about 6 weeks out of a year.
So far this has only been a development project, only one gear cut has been
put to any use. At this stage in time though I can see no reason why it
cannot produce gears.
The control function is that it's all based on a figure of 4,000 more on
The encoder is geared up off the hob arbor to give 4,000 pulses per rev. In
my case as it's a 1024 encoder I needed 125:32 to get the required 4,000.
The stepper is 200 steps per rev so 4,000 / 200 = 20 which is the required
ratio of the dividing head.
In my case I couldn't find a decent 20:1 gearbox so I opted for a 40:1 and
geared it up 2:1.
The pulses from the encoder are fed into a divide by black box where
thumbwheel switched decode what the ratio will be and issues steps to the
motor in the required ratio.
The gearbox is rotating all the while the hob is moving, any variation in
speed is picked up by the encoder and the steps are altered accordingly so
it's always in step.
Simply put its an adjustable electronic gearbox.
4,000 was chosen as above 4,000 pulses you can start having encoder
problems in keeping up.
Below 4,000 and you need to reduce the ratio of the gearbox and you could
have torque problems.
The logic is on the divide by board which is a simple board with about 3
chips, a new board is being done at the moment which hopefully will address
the encoder problem better and require lower gearing to the encoder and the
use of standard 40:1 dividing heads.
Once this is out and tested on my original setup for problems my next stage
from there is to build a new setup that fits onto Bridgeport type turret
mills or any mill where the head can tilt in one axis to accommodate the
hob helix angle.
About the only mill not included will be the round column Mill/Drills.
- posted 18 years ago