dividing plate : necessity for gears or just convenience?

[Henk]

Hi all,

Yesterday I went out to buy a rotary table + dividing plate + centre (110 mm diameter).

The seller started inquiring as to whether I intended to do any gearcutting any time soon - which was not the case. Based on that I settled on a one size bigger rotary table (6'' / 150 mm). I think I was glad that he did not immediately sell me what I directly asked for. In all honesty, I expect it will be at least 5 years before I indeed cut any gears.

However, when I got home, I started wondering as to the need for dividing plates. Up to now, I assumed that they were a necessity for getting accurate gears. The way I understand it now (after having studied the table a bit), the real advantage is not the accuracy they provide, but the convenience to ensure you position the table correctly (it offloads your brain so to say), for each tooth position.

The table is geared 90:1, one rev of the dial gives 4 degrees, dial is down to minutes and a nonius (vernier) to ten arcseconds. At the edge of the table (75 mm) therefor the smallest readable dial change amounts to 3.6 micron. Admittedly, I have no idea what accuracy a gear needs - but this is way smaller than at least what I can achieve in positioning two holes w.r.t each other. So that suggests to me that it should be enough for decent gears.

Which leads me to the point that if you are concentrated, you do not need dividing plates.

Is my reasoning correct? If so, anyone out there who was indeed able to do so without messing up?

Moreover, my impression is that it is not very difficult to actually make a dividing plate for a one off job -given that the amount of holes one has to make is typically smaller than the amount of teeth in a typical gear. True or just me jusitifying not buying the dividin plates?

Thanks for any thoughts on the subject.

Henk

 

[b.lindsey]

Henk,
I think you should be fine so long as your RT can mount both vertically and horizontally. You may need the tailstock though at some point for better support.

Bill

 

[gbritnell]

Dividing plates are not a necessity with a rotary table. When cutting increments that don't fall on even angles you have to make up a step chart to keep track of where you are. I believe the Jerry Howell website has a chart for doing such things.
gbritnell

 

[MachineTom]

The plates together with the quadrant speed up a very tedious process. Adding a multi-factored number like 8° 36' and 8" to existing number is laborious, then cranking that out on a rotab, compared to 1 turn drop in the last pin hole, advance quadrant, cut tooth, repeat. The plates are used in conjunction with pin lever and quadrant, something that would need to be made for your rotab to use plates. As well dividing heads are usually 40:1 not 90:1 so plates for a 90 ratio would be needed.

 

[Mainer]

They're a convenience, not a necessity, though it can get pretty tedious without them. You should be able to program Excel (or the OpenOffice equivalent) to give you a chart the required deg-min-sec though.

For example, with a90:1 gear ration on your R/T, to do a 32-tooth gear you need 90/32 = 2.8125 turns per tooth. 0.8125 is 13/16, which with dividing plates would be 26 holes in a 32-hole plate, or equivalent.

In deg-min-sec, I work it out to be 48 deg, 45 min. So, you'll cut a tooth at 0, turn the crank twice plus 48d45m, turn the crank twice plus enough to get to 97d30m, twice plus enough to get to 146d15m, etc.

 

[mklotz]

I've made gears without problems using an RT sans dividing plates. Use the ROTARY program from my page to get a printable chart of the angular settings - example output below...

Number of divisions = 13 

DIVISION   degdec   deg   min   sec
    0   0.0000    0    0    0
    1   27.6923   27   41   32
    2   55.3846   55   23    5
    3   83.0769   83    4   37
    4  110.7692   110   46    9
    5  138.4615   138   27   42
    6  166.1538   166    9   14
    7  193.8462   193   50   46
    8  221.5385   221   32   18
    9  249.2308   249   13   51
   10  276.9231   276   55   23
   11  304.6154   304   36   55
   12  332.3077   332   18   28
   13  360.0000    0    0    0

 

[Henk]

Thanks for the responses.

Bill : Indeed the RT can be mounted upright, as you can almost see in below image. As I do not have an angle plate yet, I even considered using the RT as such when the need is high.

The first time I'd probably want to do gears is for a steam tractor (a project a considerable number of years in the future) - I think that the designer of such also wold have a preference for nice numbers. And if not, that would be a candidate for a modification!

I also had another look at the vernier, and my impression is that they made an error punching the numbers to it (or I simply do not understand minutes and seconds of arc. The division on the dial is 60 for each degree (making it an arcminute as I understand it). So far so good. However, there are three lines on either side of zero, with a max on either side of 60 - suggesting it to mean arcseconds.

When aligning to the 60, then it amounts to only half a displacement on the main dial - meaning 30 arcseconds!

I guess I will just put a piece of sticky tape with other number written onto it over these numbers, unless one of you puts me right - which would be appreciated. I am not too fond of the -30 till +30 range that is then used so I will simply start the 0 on the left hand side, and leave the 60 on the right hand side.

 

[MachineTom]

Your main dial is in increments of 2' min, so the marks of 60 " sec is correct from the "0" mark. So in your photo of 0 to 0 , a bit more would be 20" than 40" than 60" or one min, at that point a line will be on the 60, and the "0" (wisdom line) will split between the O and first tic on the dial. Thats plus one minute to the table reading.

 

[Blogwitch]

Your RT looks to be either a Vertex or one of it's clones, so there is no need to worry over dividing plates.

If some time in the future you decide that you would like to use them, third party sets are availble to fit your table at very reasonable cost.

The top one on here, or a bit further down for more expensive Vertex ones.

http://rdgtools.co.uk/acatalog/Dividing_Head_Accessories.html


Bogs

 

[mklotz]

Just ignore the vernier. Arc seconds are extremely tiny things.

1 arc second is 5 microradians. That means that if you make a 1 arc second error pointing your rifle at a target 100 yards (3600 in) away, you'll miss by a whopping 0.018 in.

There's nothing you're going to be doing with a rotary table where you will have to worry about arc seconds.

 

[Henk]

MachineTom: You are very polite : Apparently I do understand what an arcminute and an arcsecond is. I am just not able to count the number of divisions between 0 and 30..... Silly me.

Bogs: The label strongly suggests that it is indeed a Vertex.

Marv : The nothingness of even 20 arcseconds indeed is what let me to wonder that maybe you do not really need a dividing plate performance wise.

So for now that finalizes my questions - until I actually start using the RT - which will be a few weeks as my mill is in repair.

 

[Mainer]

It's a vernier for subdividing the minute reading. As Marv says, forgetaboutit. I would be somewhat surprised if the r/t could in fact reliably measure that precisely.