College Engineering gear hobber castings set

Hi,

Has anyone here built up the castings set for a gear hobber that's available from College Engineering ? I heard that it's an improved version of something called the Jacobs hobber which I've seen described as "flimsy" and I'm interested to know if the CE version is better in this regard ? Also, what is the maximum diameter gear size it supports and can it cut helical gears ?

I'm aware of the yahoo group dealing with gear hobbers and know I could email CE with my questions but I was just interested to know the opinions of anyone here who has actually used the CE hobber.

Thanks,

David

Reply to
mangled_us
Loading thread data ...

Not built one but studied hobbing for a long while. I was watching Giles Parkes cutting a 97 tooth gear at the recent show. I'd guess that it's max is about 5" to 6" in diameter. They are lightweight, nothing can get away from mass but they do work. Giles was telling me that the gear he was cutting 97 tooth 20 DP with a tooth depth of 0.108" would need 4 to 5 passes to cut this in cast iron.

They say they can cut helical's but the design without a lot of modification won't allow this. The Jacobs hobber said it could cut helical's and bevels, Jacobs even showed some examples but never any details and other people like Giles and Ivan Law haven't worked this out.

The SMEE has a modified hobber in it's possession fitted with a differential that looks as if it can do helical's but the builder died a long while ago and again no details are forthcoming. It was bandied at one stage that Giles and Ivan would borrow this to see how it worked but ill health on Ivan's wife's part and other things have got in the way.

There was a three part series written in MEW by Harold Hall on building one but it's obvious from the article that it was built to order as an article and never used so no operating details accompany the article.

I don't know where you are located but can you get to the ME show at Lemington? It runs until Thursday and Giles has his on show there under power.

-- Regards,

John Stevenson Nottingham, England.

Visit the new Model Engineering adverts page at:-

formatting link

Reply to
John Stevenson

Not an expert like John but I had a good look at the machine and a chat with Giles at the show today. The machine strikes me as an interesting and useful tool well capable of producing the odd gear within its' range but I don't think it would go the distance for even small time production work. Obviously if time is not to be costed it will get the job done. It also seems that much of the time would need to be spent producing hobs which can apparently be quite challenging for some gear types. I also noticed that CES had a complete machine on their stand marked as "all new =A3390 (I think)", I have no idea if that is for the completed machine or just the basic castings. As John says if you can make it tomorrow it would be worth seeing/talking to the SMEE guys. If not I'm sure you will catch them at Sandown.

Regards

Keith

Reply to
jontom_1uk

I haven't tried the CE castings, but got part way through building the "glue and screw" Jacobs original design.

formatting link

The basic design of the Jacobs original can support cutting of helical gears although his explanation left a lot to be desired. Basically, the infeed angle is set to the pitch angle of the hob plus the helix angle of the gear (for RH helix and assuming RH cut of the hob). This causes at least two problems:

1: The division ratios become a problem as you have to factor by 1/cos(theta), where theta is the difference between the helix angle and the pitch angle of the hob to get it to work at all. This makes for some ugly division gear combinations (if you can actually find a combination that fits within the required level of accuracy). 2: You will encounter problems lining it up again if you have to do a multiple pass unless you lift up and reverse it all the way out each time.

For those and a host of other reasons, I decided not to pursue it any further.

Reply to
Duncan Munro

There was an article in MEW108 by Brian Thompson on electronic hobbing. Unfortunately MEW printed this with some parts missing and some out of context which made it hard to follow. I am using one of these setups and once by the initial electronic mumbo jumbo it's an easy setup.

To do helical's is so easy as you don't need to work out the extra feed ratios, you just plow thru and if you need a second cut it stays in mesh.

Here is it cutting a helical on the horizontal mill.

formatting link
Ignore the hob as my cheap camera won't stop rotating objects. There are absolutely no gears in this setup at all. table feed can be either by hand or the mechanical machine feed, it's not linked to the blank rotation which goes against all normal helical operation.

formatting link
A pair of helical's done by setting the dividing head over to the helix angle of the gear only.

This setup flies in the face of normal gearcutting but it works. Giles Parkes has been here and seen it working so he knows it's described correctly. The number of gear teeth is selected by flicking three thumb wheels on the controller and can do from 3 teeth to 999 at the flick of the three switches.

-- Regards,

John Stevenson Nottingham, England.

Visit the new Model Engineering adverts page at:-

formatting link

Reply to
John Stevenson

formatting link

formatting link
Nice one!

Reply to
Duncan Munro

On Wed, 19 Oct 2005 00:08:08 GMT, John Stevenson

John Do you have access to the correct text with all the bits in the right place?

Regards

Charles

Reply to
Charles Ping

That's pretty likely to be enough for me, I'll ask CES for the actual spec before I decide to buy though.

I don't need it for production work so I don't think the production rate is much of an issue.

Hmm, I really do need the facility to cut helicals...

I'm on the South coast and it's a bit too far to go unfortunately.

Thanks for the reply,

David

Reply to
mangled_us

I saw that, it is an interesting otion anyway...

OK, that looks like just the thing. I already have a rotary table so istm that the work involved in attaching a stepper to that is going to be easier than building up a hobber from castings to say the least. I don't have a horizontal mill but presumeably a vertical mill will do as well so long as the cuts are sensible ? What size stepper would you recommend for use with a 90:1 geared rotary table ?

Thanks again,

David

Reply to
mangled_us

Here's a text file that explains a lot of the operation parts of the process.

formatting link
It was done a while ago and some things have changed but the basic's are still relevant. The magic number is still 4,000 but it's now possible to use both channels of the encoder to reduce the gearing. With the stepper drivers we have today this number is still valid. To use a high count rotary table you will have to gear it down.

An alternative is to use a proprietary gearbox such as I have done as a low cost method.

-- Regards,

John Stevenson Nottingham, England.

Visit the new Model Engineering adverts page at:-

formatting link

Reply to
John Stevenson

Thanks, that's very interesting. Can you tell me what level of torque is needed to turn the dividing head ? I want to choose a stepper and though obviously going bigger is safe I don't want to spend more than necessary.

David

Reply to
mangled_us

Sounds like a case for electronic synchronisation :-)

David

Reply to
mangled_us

Reply to
giles

Hello Giles and welcome to Usenet!

I suggest that we continue our email correspondence herein, and we might find that others have the answers at their fingertips. (I have undertaken to provide Giles with an Excel spreadsheet to calculate his gear ratios for helical hobbing. If there are some who already have such a thing, you know what to do!)

Points upon which I am uninformed.

  1. What is the reason for having a differential in the gear train for the hobbing of helices? Is it to combine the separate motions arising from the revoultion of the hobs and also from the transverse motion of the hob into the rotation of the blank?

  1. If the transverse moti> It seems that the largest rotation of the vertical slide is limited to

Reply to
Polymath

Now that we have the flexibility of stepper motors and widespread computer availability at our disposal, is there perhaps a much, much simpler way to address the generation (not forming) of any gears?

Already there have been articles on the adaption of electronic techniques to simplify gear trains for hobbing (especially when prime-number gears were to be cut for which the corresponding prime-number master wasn't available). But is it possible that even this approach is unnecessarily complex?

I have a proposal, and it is based on the idea that the need to cut gears by the model-maker (but perhaps not the professional) is few and far between, and that therefore there is nothing lost if it were to take an hour, say, to cut a whole gear.

The idea is based upon the generation of teeth by using a rack, and realising that the whole principle that applies for a many-toothed cutter that is the rack could be applied to a single toothed cutter, and then applied many times over, with suitable increments in rack position, rack traverse, and blank rotation. (And hence providing the necessary flexibility in a number of dimensions to be able to generate helices)

Only one cutter is needed for the whole range of involute gears from 12 to rack, and that same cutter can be used for gears of differing DP's. That cutter is also simple to manufacture, either as a wheel or as a vertical milling bit as it has straight sides at the pressure angle.

How to vary the DP? Simple - a small enough cutter for the smallest DP you wish to work with can then cut either flank by two separate generating actions?

Takes l> Hi,

Reply to
Polymath

Been done. I have a vertical shaper with exactly that arrangement on it. Unlike the idea above I only use one tool per DP, after all it's only a simple Vee tool and easy to grind. It is slow but remarkable accurate and if the machine is sturdy enough as mine is then the quality is also very good.

I'm working off and on with cutting cycloidal gears by the same rack method. Making progress but slow going as it has to fit in with other work.

-- Regards,

John Stevenson Nottingham, England.

Visit the new Model Engineering adverts page at:-

formatting link

Reply to
John Stevenson

This method has been described in the model engineering literature. The teeth on the hob are circumferential (not helical), and can be used without relief.

Cutting teeth is tedious, practically the same as using a single milling cutter. Advantage is of course that you only need one hob to cut all gear sizes of a given pitch and pressure angle. But accurate indexing is required for decent results.

It does a good job. I've cut a bunch of 32 DP gears for Maudsley's reversing gear on a steam engine.

My hob was made of O-1 tool steel, machined in the lathe with appropriate "threading" tool, heat treated, then finish ground on a Quorn T&C grinder. Obviously this method leaves small flats on the tooth flanks; nonetheless, the teeth mesh and run smoothly and quietly.

I can certainly recommend this for model / home use.

Regards, Wolfgang

Polymath wrote:

Reply to
wfhabicher

Wolfgang. Not quite the same thing, you are describing hobbing with annular grooves and cutting one tooth at a time, with the adjacent teeth making an approximation of the involute shape. Like you say it has small flats.

What our learned friend proposes is a modern version of the Sunderland gear planer when the cutter [ in this case a short rack ] planes backwards and forwards over the work. At each stroke the cutter moves along slightly in mesh with the gear blank which rotates. In theory you get a many sided involute that in practice is a curve.

After the Sunderland has cut about 4 or 5 teeth it backs off, index's round by 4 or 5 teeth and repeats the operation until complete. These machines have been around since the late 1800's but work very well.

Substituting the rack for a single vee tool, i.e. one rack tooth, this works the same but takes a while longer.

Using Mach3 which is a windows 2000 application and a special wizard screen you get to fill in various boxes like depth of cut, DP, number of teeth and home points and it then automatically generates the code to run three axis. One axis controls the feed past the gear, one controls the infeed for depth and clearance when backing off and the last controls the rotary table carrying the blank.

The cutting action is done by a mechanical ram and the feeds are only applied when the tool is at the top of the stroke, no feeds are carried out whilst cutting.

-- Regards,

John Stevenson Nottingham, England.

Visit the new Model Engineering adverts page at:-

formatting link

Reply to
John Stevenson

John,

Thanks for the enlightenment! I guess I missed the bit about the cutter moving along the gear blank pitch line incrementally, ie. "rolling" with the blank.

I have thought of introducing this motion during hobbing with the concentric hob.....this would reduce the size of the flats. Indeed, for a small # of teeth I index the blank by 1/2 tooth and raise the cutter by a commensurate amount. It improves the tooth form and increases the # of flats.

As I said earlier gears produced this way run well together and with commercial gears of same pitch and pressure angle.

I am not well versed in CNC machinery but, I would think that an ordinary shaper, equipped with stepper motor driven cross feed, similarly driven index head, and limit switches on the ram travel, with appropriate controller, could produce good gears using simple, rack-type cutters.

I would really like to try this, especially with rack-type cutters made from mild steel commercial racks: At the cutting edge machine a 2 degree back-rake angle, and in the cutter holder introduce 1 degree or so of clearance angle to prevent rubbing during the cutting stroke. Case harden the rack, and machine. Should be OK for brass, bronze, and leaded steel gears.

Could anyone here perhaps direct me to a book or web page that would shed some light on the controller requirements, suitable for "rolling your own"?

Thanks,

Wolfgang

Reply to
wfhabicher

Taking a few incremental cuts on each tooth will improve the shape of the tooth.

It's not a new solution at all. There were commercial attachments made around the turn of the 20th century that bolted onto a standard shaper for just this purpose. Model Engineer ran an article in September 14th 1950 by 'BaseCircle' describing this attachment. There was a later article on the 20th January 1989 covering the same ground.

The controller is easy, any CNC controller like Mach3 or Turbo CNC that can run 3 axis will do this. The code is quite simple.

You start off with the tool central on the blank and just touching, this is X0.0, Y0.0

The code only runs when the tool is at the top, limit switches control this.

N001 X0.0 Y0.0 [Sets zero ] N002 G00 X-20.0 Y0.0 [ Moves tool clear of blank at rapid move] N003 G00 X-20 .0 Y-3.0 [ moves tool in to depth ] N004 M03 [ starts stroke motor] N005 G01 X20.0 Y-3.0 A20.0 F50.0 [ Moves tool past the blank by 40mm whilst revolving it thru 20 degrees at 50mm/min] N006 M01 [ stop motor] N007 G00 X20.0 Y1.0 [ move tool clear on depth ] N008 G00 X-20.0 Y1.0 [ Move to start point] N009 G00 X-20 Y-3.0 [ moves tool back in to depth ] N010 G00 A 360/N [ rotate for next tooth ] N011 [ Repeat from line N004 ] Etc.

This code is quite long winded and done this way to explain the code, many moves like G00 are modal and do not need to be repeated, same for X and Y values that don't change. Line N010 will be a fixed value for the number of teeth required, so

20 teeth will be 360 /20 = 18.0

-- Regards,

John Stevenson Nottingham, England.

Visit the new Model Engineering adverts page at:-

formatting link

Reply to
John Stevenson

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.