More help please

[David NobleParticipant @davidnoble71990]

 

I've started to learn to cut gear wheels in preparation to building a clock. I'm having difficulty with the last tooth not lining up. I realise that something is moving but I'm blowed if I can see what it is. I've tried indexing round the whole wheel after cutting just one tooth and everything lines up perfectly. The problem seems to come after I cut 4 or 5 teeth then index round to the end to cut the last couple of teeth.

As always, any advice appreciated.

David

Edited By David Noble on 10/01/2019 09:59:35

[David NobleParticipant @davidnoble71990]

[Michael GilliganParticipant @michaelgilligan61133]

David

A description of your set-up, and preferably some photos, would help any of us attempting 'remote diagnostics'

MichaelG.

[JasonBModerator @jasonb]

Sounds like the blank is being rotated round slightly on each cut either on it's arbor or in what is holding the arbor. When you start put a couple of lines arcoss these places with a sharpie and then see if they move.

[vintage engineerParticipant @vintageengineer]

It can only be your gearing is out or something is moving in relation to your gearing.

[JohnFParticipant @johnf59703]

Ha! Reminds me of my first foray into gear cutting as an apprentice my then mentor Ernie who was showing me how set everything up and make a trial cut came to the last cut and said “this where you ask do they want one big one or two little ones”

Flippancy aside from what David says Jason is on the right track something is moving most probably the blank has insufficient clamping force on its mandrel? As Michael suggests a photo of your set up would probably help a lot.

John

Edited By JohnF on 10/01/2019 11:30:37

[Ady1Participant @ady1]

Fix it in position with a hole and pin?

If it's still out at the end of the job its the equipment

If it isn't it's the workpiece

[David NobleParticipant @davidnoble71990]

Sorry about the orientation! Just using some scrap to practice on, the backing is a piece of ply. The centre is a Morse taper plug made to centralise the wheel.

Edited By David Noble on 10/01/2019 12:40:28

[FMESParticipant @fmes]

Possibly a daft question – any backlash in the dividing head?

Regards

[Yngvar FParticipant @yngvarf]

Also if the table says something like 3 7/16 , that is 7 spaces not 7 holes…..

[Nigel McBurney 1Participant @nigelmcburney1]

Have seen it happen when I was an apprentice,and have had it happen to me once ,the set up was rigid and on a good Adcock and Shipley mill. I had bought a lot of gear cutters at an auction all in good condition except one ,one gear was cut ok, the mating one of different size and change of cutter ended up with a small tooth,the reason after carefully examining the cutter was that the cutting edge on the teeth on one side of the cutter had rubbed slightly against something hard when used previously to my purchase,so as the cutter went into the cut the blunt side pushed the metal rather than cut and the sharp side cut deeper probably only around a thou per tooth which after 40 teeth is around 40 thou (1 mm) at the last tooth,it was soon corrected on the Clarkson cutter grinder but it meant a new blank and if there is any doubt with a new to me cutter it gets sharpened. The slippage error can be more common when cutting spiral gears,I use stepped and threaded mandrels ,with the nut done up very tightly so there is little chance of the gear moving and a peg fitted to the mandrel to engage with the the keyway if a keyway is specified.

[David NobleParticipant @davidnoble71990]

Thank you all. I think I'm getting there. (famous last words) Is it important that there is 0.020" runout on the cutter? The mandrel runs true and the cutter is brand new.

[Pete RimmerParticipant @peterimmer30576]

20 thou runout is like 20 miles for a gear cutter for clock gears.

[David NobleParticipant @davidnoble71990]

Posted by Pete Rimmer on 10/01/2019 18:59:50:

20 thou runout is like 20 miles for a gear cutter for clock gears.

I'll take a look at that then Thank you.

[JasonBModerator @jasonb]

Even if the cutter is running out and cutting an oversize gap it should still index back to the same oversize gap

[David NobleParticipant @davidnoble71990]

Ahh, sorry, when I said 'runout' I was trying to say it was running elliptically not wobbling.

Edited By David Noble on 10/01/2019 19:11:15

[Anonymous]

Posted by David Noble on 10/01/2019 18:53:17:

Thank you all. I think I'm getting there. (famous last words) Is it important that there is 0.020" runout on the cutter? The mandrel runs true and the cutter is brand new.

Ah, I'd place a small bet it's an RDG cutter?

The eccentricity is awful, but as JasonB says it shouldn't affect the indexing. If I understand correctly you've cut one tooth and then indexed around the rest of the wheel without cutting teeth, and once back to the start point the cutter is exactly where it should be. If that is true then the problem is something slipping, almost certainly the gear wheel itself. While it's bolted to the table I'd guess the bolts are not precision fits. So it'll be quite easy for the work to slip a little; especially if the cutter has a slight wobble as well.

Been there, done that with RDG cutters, not that it really mattered as I was only interested in seeing if I could cut helical gears the old school way. I've also had a gear slip when cutting. That's why there's a green clamp in this picture – equals no slipping:

Andrew

[JasonBModerator @jasonb]

That then raises the question, should involute cutters be used for clocks?

As I said early on the slippage is small so not easy to see, draw lines across table to ply/mdf and mdf to workpiece and take a few cuts then look to see if the tell tale lines have shifted out of line.

Always try and turn the hand wheel in one direction and add a bit of friction with a table lock.

[David NobleParticipant @davidnoble71990]

Posted by Andrew Johnston on 10/01/2019 20:16:33:

Posted by David Noble on 10/01/2019 18:53:17:

Ah, I'd place a small bet it's an RDG cutter?

You would be right

All this is hugely helpful, thank you.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by JasonB on 10/01/2019 20:23:55:

That then raises the question, should involute cutters be used for clocks?

.

It's a matter of a-pinion, Jason

Seriously though: This is worth reading **LINK**

https://www.csparks.com/watchmaking/CycloidalGears/RichardThoen.xhtml

MichaelG.

.

Note: the page that links to that ^^^ contains good information about traditional cycloidal wheels

https://www.csparks.com/watchmaking/CycloidalGears/index.jxl

Edited By Michael Gilligan on 11/01/2019 00:04:20

[Paul KempParticipant @paulkemp46892]

David,

When I cut my traction engine gears which were quite big and a mistake would be costly in both time and material I gave myself a check to make sure there were no issues. Clamped a square to the table against the edge of the gear (gears were parallel to the table) and went round and scribed a line at every division. I then clamped a scribing block to the table lined up with the first line and checked each time I advanced the gear that the tip of the scriber lined up exactly with my previously scribed lines. Probably took an hour on the biggest gear to set up but it saved me 2 mistakes due to brain fade over the 6 gears. However focussed and organised you think you are it's very easy with the boredom of winding handles to get distracted

Paul.

[Russell EberhardtParticipant @russelleberhardt48058]

Posted by Andrew Johnston on 10/01/2019 20:16:33:

Ah, I'd place a small bet it's an RDG cutter?

Most clockmakers stick with Thorntons cutters, expensive but good. I buy mine from Meadows and Passmore and never had a problem. Most clock designs for amateur construction specify cycloidal cutters. You might have difficulty sourcing involute gear cutters in the appropriate MOD.

Russell

[David NobleParticipant @davidnoble71990]

Thank you all for this info' I am getting there an allied question is 'what is the recommended way to work out the tooth depth'? I have had a look on-line and found various answers. Some give a fixed depth based on my cutter of MOD 0.75 and some vary the answer by using number of teeth.

[Anonymous]

Posted by David Noble on 11/01/2019 11:06:41:

Some give a fixed depth based on my cutter of MOD 0.75 and some vary the answer by using number of teeth.

Which just goes to show you can't always trust the internet. The real skill is in determining who is an expert, and who is an ex-spurt.

I'll give the cutting depth details for gears defined by diametral pitch P, as I've got the numbers in my head. Shouldn't be difficult to translate to MOD defined gears. For involute gears the depth of the cut is determined solely by the diametral pitch; it is not dependent upon the number of teeth. Logical if you think about; otherwise a large gear and small pinion may not mesh properly because the tooth depths are different.

The depth of cut is the sum of two parts, one above the pitch circle and one below. Above the pitch circle is the addendum, defined as 1/P. Below the pitch circle is the dedendum, also defined as 1/P. However, also below the pitch circle, some clearance is needed at the bottom of the space. The clearance is arbitrary, but two common values are 0.157/P or 0.25/P. The total cut depth is the sum of the addendum, dedendum and clearance, so either 2.157/P or 2.25/P. A value of 2.157/P is the more common. The total depth of cut should be marked on a quality cutter. If not then assume the lower value and either measure the tooth width at the pitch circle with a gear vernier, or see how they mesh compared to the calculated centre to centre distance. As an aside my experience of gear verniers is that they're rather difficult to use, especially as the teeth get smaller so I no longer use mine.

Andrew

[JasonBModerator @jasonb]

I tend to look at HPC who as well as selling gears have a handy table of how to work out various things gear related.

So for your MOD 0.75

Addendum is 0.75mm

Dedendum is 1.4 x 0.75 = 1.05mm

Total depth is the two added together so 0.75 + 1.05 = 1.8mm

Or you can add the depth constants together first to get 2.4 and use total depth = 2.4 x MOD

NOTE they do give different depths depending on DP or MOD with smaller teeth not being cut deeper

 

Edited By JasonB on 11/01/2019 12:15:11

[Author]

Posts