Free hobbed worm wheel

[A GParticipant @ag]

Has anyone made a large (around 12" dia) free hobbed worm wheel?? I'm after one for an indexer and would like to know what the accuracy of one free hobbed this size is, or am I better buying a rotary and using the worm and wheel out of that? I want to use it for indexing when gear cutting

[A GParticipant @ag]

[John ReeseParticipant @johnreese12848]

Have you checked the vendors of stock gears? They might be expensive but probably less than buying a rotary and tearing it apart.

[John ReeseParticipant @johnreese12848]

Just spent several minutes on the web searching for stock gears of the size you wanted. This is the closest I could find. http://www.martinsprocket.com/products/product-search?Website_Code=GR5

[A GParticipant @ag]

When looking at UK suppliers, it's around £250 for a 6" worm and wheel plus the VAT!!! I can get a 6" rotary table for around £150. Ideally I'd find a second hand 12" plus diameter rotary table and use the bits out of that if I can't get the required accuracy free hobbing

[HopperParticipant @hopper]

Why would you need a 12" diameter worm wheel? It seems overly large.How big are the gears you want to index? Surely, for a positioning-only application such as gear cutting indexing, a smaller wormwheel would do the job? GH Thomas claims an accuracy of 1/1000th of a degree for his Versatile Dividing Head, using about a 3" wormwheel and a secondary 2" wormwheel for indexing prime numbers etc.

And why buy a rotary table and pull it apart? All you need to do is add the readily available indexing discs to a rotary table and you can cut your gears with that.

[Paul LousickParticipant @paullousick59116]

I have been looking for a used 10" or 12" rotary table but cannot afford the price tag. Up to $1000. I believe that all of the big ones are built much better than the cheap Asian clones. I would not like to destroy one to make another.

Why not just use one" as is" as Hopper has said. this is what other people do.

Paul.

[Michael GilliganParticipant @michaelgilligan61133]

A G,

Jesse Ramsden [to whom we owe a lot] effectively free-hobbed the very large wheel on his Dividing Engine. … have a look at the link in my post on this previous thread.

That said … Like others; I am a little confused as to what exactly you want to do.

MichaelG.

[JasonBModerator @jasonb]

I think I have seen "jubilee" clip material fixed around a wooden disc that works much like a worm & wheel

[Michael GilliganParticipant @michaelgilligan61133]

Posted by A G on 04/04/2016 00:28:11:

When looking at UK suppliers, it's around £250 for a 6" worm and wheel plus the VAT!!! I can get a 6" rotary table for around £150. Ideally I'd find a second hand 12" plus diameter rotary table and use the bits out of that if I can't get the required accuracy free hobbing

.

• Why are you surprised by the price of a 6" worm and wheel ?
• 6" rotary tables very rarely have 6" wormwheels. [photos in this thread]
• Maybe your best compromise would be the 8" table from a BCA … these do have extermal teeth, and are beautifully made. … You may hit lucky and find one with a battle-scarred table-top at a modest price.

MichaelG.

Edited By Michael Gilligan on 04/04/2016 08:25:10

[BWParticipant @bw]

Posted by Hopper on 04/04/2016 05:15:16:

GH Thomas claims an accuracy of 1/1000th of a degree for his Versatile Dividing Head, using about a 3" wormwheel and a secondary 2" wormwheel for indexing prime numbers etc.

Am fascinated by this – just raised that very same point in another thread – I have read the GHT books and think they are great – am in awe of the man.

Anybody got any thoughts on the plus/minus associated with that 1/1000th of a degree ?

Bill

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Bill Wood 2 on 04/04/2016 10:11:20:

Anybody got any thoughts on the plus/minus associated with that 1/1000th of a degree ?

Bill

.

First thought … It's a resolution, not accuracy

[but I think you already knew that]

MichaelG.

[Anonymous]

I've free hobbed a worm wheel with a home made cutter, although not 12" diameter:

I haven't measured the accuracy, but I suspect it is poor compared to that needed for a rotary table or dividing head.

I can't see the point of buying a rotary table, dismantling to get the worm and worm gear, and then making ……. a rotary table. I've used a rotary table to cut gears:

The rotary table is 12" diameter and the gear OD is 14.8". I can't remember the exact diameter of the worm wheel in the rotary table, but it was about 4". After cutting the complete gear the cutter went through the first cut with almost no sound of re-cutting. So even with a relatively small worm wheel the alignment on the OD of the gear must have been very close.

I find it difficult to believe that GHT was achieving an accuracy of 1/1000 of a degree. My commercial (Elliott) dividing head claims a maximum error of 1' 30" of arc; that's about 1/40 of a degree. If nothing else how would you measure that sort of accuracy? On a 3" diameter 1/1000 of a degree is about a quarter of a tenth of a thou. Given that the resolution of a measurement system needs to be rather better than the quantity to be measured one is looking to resolve a few microinches.

Andrew

[Michael GilliganParticipant @michaelgilligan61133]

For those who might not be familiar with GHT's design:

He achieved high resolution by cascading two worm reductions.

High Resolution is easily claimed

Accuracy and Precision [or in modern terminology low 'uncertainty of measurement'] is a little more challenging.

MichaelG.

[John Stevenson 1Participant @johnstevenson1]

Once you get below 50 microns though or claim to it's all dick slapping .

[Jeff DaymanParticipant @jeffdayman43397]

"Once you get below 50 microns though or claim to it's all dick slapping ."

Excuse me???

BTW Lots of people around the world work well below the 50 micron accuracy level. Most shops have no need to, but there are plenty that do.

[John Stevenson 1Participant @johnstevenson1]

True Jeff and they probably have the facility to accurately measure this but how many in a home shop for all what they say can repeatedly work to it AND measure it ?

Who works in a climate controlled shop ?

When was the last time a home shop guy sent a set of micrometers or other measuring gear away to be calibrated ?

When was the last time you even saw a set of model drawing with a tolerance on them ?

Go onto CNC Zone and see all the posts where some Herbert has got a Machine Mart micro lathe and is converting it to CNC and announces that if he micro steps the drivers to 254 steps and on his 0.1mm pitch ACME screws he will have a resolution of 2 microns ?

Yer, whatever.

Dick slapping.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Michael Gilligan on 04/04/2016 12:15:09:

… He achieved high resolution by cascading two worm reductions.

High Resolution is easily claimed …

.

Each reduction is 60:1 [combining to give 1/10° per tooth] and the 'micrometer drum' has 100 divisions.

[quote] A drum on the worm spindle having 100 divisions would thus enable us to work to 1/1000 deg. which is an order of accuracy well beyond anything we would be likely to want — or be able to use ! [/quote]

Of course: The actual 'uncertainty of measurement' will depend mainly on the accuracy of manufacture and set-up … but the principle is sound, and GHT's [quoted] claim seems modest enough.

MichaelG.

[BikepeteParticipant @bikepete]

Large worm wheels are also often used on telescope mounts, so small ads sections on astronomy forums or the like might be worth watching if you're looking to buy. Also astronomy suppliers do them: look e.g. here and you will find a 14" worm & wheel set for £250 (details here). Also worth a read is this on DIY worm wheel cutting from Chris Heapy which was I think written with telescope mount applications in mind.

Edited By Bikepete on 04/04/2016 16:23:04

[HopperParticipant @hopper]

GHT also warns against using free-hobbed wormwheels for dividing purposes because the teeth do not necessarily all come out the same thickness. He advises instead to use a commercially available helical gear of about 4-degree angle or so.

Or you can use a straight-cut gear and mount the worm at the 4-degree angle to mesh with the straight gear.

[AjohnwParticipant @ajohnw51620]

The 60T gear and worm are still available from ME suppliers. The alternative is an ordinary gear and angling the worm rather than slanting the teeth on the gear – more easily done on a dividing head. I do have a rotary table where the worm is held in a carrier and swung into mesh rather than the usual eccentric cam arrangement. Might be easier to mount a worm at an angle in a table that way.

The telescope wheels are usually made of aluminium but the same home made technique could be used. Gash the blank for the correct tooth count and then use a hob or a tap to form the shape, Then lap worm and wheel. I don't see why this shouldn't work out in materials other than aluminium especially if a tapered tap is used such as the ones used to make deralin nuts.

Lapping might best be done in 2 stages. A cast iron worm first and then the actual part. Cast iron laps are often used on glass as it holds it's shape well. Might be needed might not. Lapping can get pitch errors to very low levels. It's even been done on spur gears used in telescope drives but care as to be taken that the same teeth don't keep coming into contact. Different base primes tooth counts should help with that – maybe more than one of them to be sure used alternatively.

Using a gear and the worm in a frame is probably the easiest option. No guarantee on the other way given or implied – just seems feasible to me. I even bought a tap from China but the idiots sent me a left hand one and changed the listing when I complained.

Adding on the GT arrangement – the accuracy depends on the main wheel. As he points out the accuracy of the division plates or 2ndry worm and wheel is improved by the tooth count factor of the main wheel – which could also be lapped lightly.

John

–

Edited By Ajohnw on 05/04/2016 12:06:03

Edited By Ajohnw on 05/04/2016 12:08:10

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Ajohnw on 05/04/2016 12:03:36:

Adding on the GT arrangement – the accuracy depends on the main wheel. As he points out the accuracy of the division plates or 2ndry worm and wheel is improved by the tooth count factor of the main wheel –

.

Unless you are actually indexing small angles … in which case that 'improvement' [averaging effect] is not relevant. i.e. the improvement depends upon the use of multiple turns.

MichaelG.

[Michael GilliganParticipant @michaelgilligan61133]

I was mildly amused to find this in Hemingway's description of the kit:

[quote] Prime numbers or indeed any possible division outside the range of the two plates are indexed by means of the “Micro-Attachment” giving capabilities down to 1/100th of a degree! [/quote]

That's a whole 'order of magnitude' coarser than GHT's claim.

Typo? … or does 'capabilities' have some special meaning?

MichaelG.

Edited By Michael Gilligan on 06/04/2016 07:56:01

[AjohnwParticipant @ajohnw51620]

Posted by Michael Gilligan on 05/04/2016 14:03:53:

Posted by Ajohnw on 05/04/2016 12:03:36:

Adding on the GT arrangement – the accuracy depends on the main wheel. As he points out the accuracy of the division plates or 2ndry worm and wheel is improved by the tooth count factor of the main wheel –

.

Unless you are actually indexing small angles … in which case that 'improvement' [averaging effect] is not relevant. i.e. the improvement depends upon the use of multiple turns.

MichaelG.

He assumes the main wheel is perfect and is using the fact that say the error in a plate is 5% it will be 5%/60 as far as the rotation of the main wheel is concerned. The plate has X holes so when one full rotation is completed the max tooth error is 5%/60 with a correct tooth count plus any error from the main wheel which are likely to be very very small. Where things start getting dubious is when what ever is being done with it is much much larger than the diameter of the main wheel as the errors will be multiplied. They will typically be reduced when things like vernier scales are made as they are likely to be smaller than the main wheel.

Personally I also like his idea of having a long plain bore in the spindle but I don't think he gives the reasoning behind that. Some one gave me his thick posh book some time ago. It's interesting to read but not sure the rolls are worth making compared with another much simpler design.

John

–

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Ajohnw on 06/04/2016 10:28:41:

Posted by Michael Gilligan on 05/04/2016 14:03:53:

Posted by Ajohnw on 05/04/2016 12:03:36:

Adding on the GT arrangement – the accuracy depends on the main wheel. As he points out the accuracy of the division plates or 2ndry worm and wheel is improved by the tooth count factor of the main wheel –

.

Unless you are actually indexing small angles … in which case that 'improvement' [averaging effect] is not relevant. i.e. the improvement depends upon the use of multiple turns.

MichaelG.

He assumes the main wheel is perfect and is using the fact that say the error in a plate is 5% it will be 5%/60 as far as the rotation of the main wheel is concerned.

.

Accepted, John … 

I suppose, at the small angles I was citing, we can just take the 60:1 reduction ratio as a lever.

MichaelG.

Edited By Michael Gilligan on 06/04/2016 12:42:20

[Author]

Posts