DIVIDING PLATE ..DIY ..Help...!!

[Steamchick]

Hi Indian,
Please can you add a picture of the back cover of the Zeus book? - It may help for those of us without the copy you have.
Thanks,
Minh-Thanh, without CAD I find this of little use, as I can't make a picture (but can copy yours!), but you seem to have a method that works for you. (What is pamme? - or panme? you seem to use 2 spellings?).
Is this "by eye" - or is it by indexing on x-y coordinates? I wonder if it as accurate as indexing off gears? (Indexing by gears systems had been used for a century of more before Mr Babbage invented his calculating machine).
Of course, without the change wheels on the lathe an alternative to gear indexing is needed. - So I should like to understand your method. It looks - from my interpretation of the picture - as if you are indexing by using a stop on the workpiece and indexing 9.23mm at a defined radius? - I can't figure how to do this on a blank after machining only the first tooth gap?
K2

 

[indian]

sorry have tried photos and scanning but the condition is awful but to the rescue comes YouTube . just do a search for bolt hole calculation there are several and among the best IMHO is Joe Pie ,tom stikkelman or r. dean Odell

 

[minh-thanh]

Hi K2
I read your posts in this thread but they don't fit what I have : I don't have odd gears , but your comments will help someone so I'm silent .
my fault : PANME
This is my idea

 

[BaronJ]

Hi Indian,
Please can you add a picture of the back cover of the Zeus book? - It may help for those of us without the copy you have.
Thanks,
Minh-Thanh, without CAD I find this of little use, as I can't make a picture (but can copy yours!), but you seem to have a method that works for you. (What is pamme? - or panme? you seem to use 2 spellings?).
Is this "by eye" - or is it by indexing on x-y coordinates? I wonder if it as accurate as indexing off gears? (Indexing by gears systems had been used for a century of more before Mr Babbage invented his calculating machine).
Of course, without the change wheels on the lathe an alternative to gear indexing is needed. - So I should like to understand your method. It looks - from my interpretation of the picture - as if you are indexing by using a stop on the workpiece and indexing 9.23mm at a defined radius? - I can't figure how to do this on a blank after machining only the first tooth gap?
K2

Hi Ken,

Free Download !

Check out "https://www.presto-tools.co.uk/Downloads" Counsellor

 

[fvd]

Back cover of the zeus book

From the book Precision hole location, by J. Robert Moore (1947):
39 holes, page 320 and 321.

 

[ajoeiam]

Hi K2
I read your posts in this thread but they don't fit what I have : I don't have odd gears , but your comments will help someone so I'm silent .
my fault : PANME
This is my idea
View attachment 159971

Almost looks like someone is trying to create a vernier protractor.

https://ia904708.us.archive.org/11/...changability in toolmaking and production.pdf

If someone wants the complete reference that mr fvd referenced - - - here it is. (above)

Looks like in attached pics

 

[trlvn]

Stepping back a little, this is for a gear hobber, right? The 39 tooth plate will create a 39 tooth gear? So any angular deviation in the plate will be faithfully transferred to the gear? (As compared to a dividing head with a 40:1 worm gear. There, only 1/40th of any deviation in the plate ends up in the work piece.)

If correct, then accuracy in creating the plate is exceptionally important. Really, there are only 2 basic approaches:

1) Turn the work to the necessary angle and drill.

2) (Use trig to) determine the x y positions of each hole. Move the work and drill.

Even without a DRO, I think 2) will yield greater accuracy. Use a spreadsheet to determine the absolute position of each hole. Taking care of backlash and carefully counting turns, it should be possible to place each hole within a couple of thous (0.05mm) of the desired location. Tedious: for sure. Accurate: better than trying to interpolate factions of a degree over and over.

My $0.05 worth,

Craig
(Using a CAD-produced template for hole locations is, in effect, just another way of using trig. But with human error introduced in the centre-punching process.)

 

[indian]

I have posted but its not appeared whats going on?

 

[minh-thanh]

Stepping back a little, this is for a gear hobber, right? The 39 tooth plate will create a 39 tooth gear? So any angular deviation in the plate will be faithfully transferred to the gear?

If correct, then accuracy in creating the plate is exceptionally important.

39 can also be 35 , it is used to cut odd gears .
Yes , accuracy in creating the plate is exceptionally important. I have made some gears myself, it is not precise but good enough for model engines, But to determine the precision needed for a homemade DIVIDING PLATE and make a good enough gear I don't know yet


Lots of interesting information
Thanks for the comments!!

 

[Steamchick]

the 39 toothed gear will need multiples of 3 and 13.
A 35 toothed gear will need multiples of 5 and 7 - So NOT the same? - Or did I go o the wrong school?
I missed the point of your comment here Minh Thanh, about using a 35 toothed gear?
That's why I explained the use of the 65 toothed gear in the set-up using change wheels to index the lathe mainshaft. it is divisible by 5 and 13, so can - with another pair - the 30 and 90, or 25 and 75, have a 13 and 3 in the gear chain so becomes an accurate integer stepping device for pre-setting the blank - without resorting to x-y coordinates, setting "by eye" to a dividing plate print, etc. - Thus eliminating a source of inaccuracy or error.
A dividing head is a very accurate set of gears that does exactly the same indexing as using change wheels on a lathe, but with reduced errors due to back-lash, I guess?
K2

 

[minh-thanh]

K2 !
39 or 35 makes no difference in the way the gears are made. It just depends on the diameter of the dividing plate.

 

[indian]

sorry but zeus book is in poor condition and does not scan well but go to YouTube and look up coordinate hole drilling. Joe pie etc ,3 of them are quite good.

 

[timo_gross]

Stepping back a little, this is for a gear hobber, right? The 39 tooth plate will create a 39 tooth gear? So any angular deviation in the plate will be faithfully transferred to the gear? (As compared to a dividing head with a 40:1 worm gear. There, only 1/40th of any deviation in the plate ends up in the work piece.)

If correct, then accuracy in creating the plate is exceptionally important. Really, there are only 2 basic approaches:

1) Turn the work to the necessary angle and drill.

2) (Use trig to) determine the x y positions of each hole. Move the work and drill.

Even without a DRO, I think 2) will yield greater accuracy. Use a spreadsheet to determine the absolute position of each hole. Taking care of backlash and carefully counting turns, it should be possible to place each hole within a couple of thous (0.05mm) of the desired location. Tedious: for sure. Accurate: better than trying to interpolate factions of a degree over and over.

My $0.05 worth,

Craig
(Using a CAD-produced template for hole locations is, in effect, just another way of using trig. But with human error introduced in the centre-punching process.)

The approach of using a micrometer screw for the rotation is probably not useless. A sine bar is using same principle.
Provided we are able to determine the distance between the micrometer tip and the center of rotation. (and provided the micrometer tip is precisely spherical)
Basically it is a mixture 1) turn the work and 2) use trig to determine the rotation angle by translating a distance.

In 1946 trigonometry was (I guess) still heavily relying on look up tables an the like, today I just ask an apple, a raspberry, a banana or (the espresso machine?) 19 years ago I would have asked Mr. Packard or Mr. Casio.
I do not see any issue in finding out the amount of angle or distance to move, I see the problem in actually moving with sufficient accuracy.
The next problems will be to drill a hole in an exact position (without wandering) and to drill the hole to exact dimension.

My understanding of post #25 is. A 90° angle (or two parallels), Five holes drilled (corners of a square with known side length + center for pivot) will allow me to rotate the square to any angle, using a standard micrometer screw, or gauge block stacks. I would be able to adjust angles quite accurate.

Thinking of it, only one pin and the center hole will be sufficient. Using four pins will require less "data" points.
It will be a pain to make a new gauge block stack for each hole to drill, but it will probably work.


Greetings Timo

p.s. I am not 100% convinced that makes sense

 

[TSutrina]

A method I heard about but never used myself is to get yourself a holed metal strip as used in packaging. The distance between the holes is remarkably precise. Get yourself a lenght that includes the number you want, say in your case 39 holes or 78 holes. Now calculate the outer diameter of a disc that fits your required number of holes and fit the holed strip flush around it. Index on the holes. An older and wiser member of our horological society used this method to cut gears for remarkably precise clocks.

May not find this metal strip, however; printers that use punch holes to drive the paper would work also. Not very strong but accurate. Make your own strip with this as the template. Get a punch that matches the diameter of the hole lift it off the metal and punch on a bar of metal a group of holes. The reason for the group is that the error is going to be the same for two next to each other or two many holes a part. Drill and dowel say a few holes with some relationship to use other holes to check pick out the best and have a few pairs so that you can fill in the holes. Just a thought.

 

[TSutrina]

A way to get accurate holes to make your own plate is to realize that the error is the same when for a hole that are on a 6 inch disk or A 3 yard disk. Drill a hole through the three tightly aligned. Need to match a dowel center punch and dowels for alignment. Now with a dowel in a plug tightly fit in the plate center, at the center of a disk. put the three sticks on that dowel and drill the desired hole or holes for multiple circles through the two sticks and the disk with a drill press or something that is at right angle to them all. Then drill a hole through the three sticks at the end to match the ones that will be drilled in all the bars. Measure the distance between centers for this farthest hole, may get it off the stick itself. Then use a table or a calculator to calculate the sine of half angle needed in on the disk ( distance time sin (pi / the number of holes on the disk circle). Double the distance and drill two bars with two holes at that distance for the dowels or a shoulder bolt . Tap over the holes not use on both sticks leaving on set. place a dowel in the first hole and one stick. punch the center mark in the second stick. Drill it. put the second stick on a dowel through that hole and drill a hole for the third stick puting it on a hole. The objective is to always us a bar to drill a new hole and the other bar to cut the error in the two holes in the disk in half. So a 0.002 error in holes. is cut to 0.001 at the disk and at 36 inch with the disk hole at 2.5 inches is reduced to 0.00014 plus the error carried over by the pinned holes in the plate. more bars and more holes in the plate further reduce the error and the error will decrease with more sticks faster then linear. So a guess that 4 sticks would ~ 0.0005. angular error which would also have both plus and minus errors. so not accumulate to the full count of holes.

 

[BaronJ]

Hi Guys,

One dodge that I use, particularly for stamping graduations on a wheel or disc, is to use a salvaged plastic gear from a printer or photocopier, as an index. Many plastic gears, paper shredder ones tend to have low tooth counts, whilst ones from printers tend to have high tooth counts. The biggest one that I have is 82 teeth and the smallest only 8.

 

[ddmckee54]

minh-thanh:

Do you know anybody with a 3D printer? I'm talking about a filament printer, not a resin printer. Although both would probably work since it looks like the diameter of your outer hole circle is 50 mm? It would be a fairly simple task to design and print a template for the holes in your indexing plate. Then just use a transfer punch to mark the hole locations.

For limited usage you could probably even just print the the entire indexing plate. It would be useable, just not all that durable. Most FDM printers claim accuracy to within 50 microns, and resin printers are a lot more accurate than that.

Don

 

[TSutrina]

For you that have a 3D printer and think it is accurate. Mount a brush toy motor with a small drill of center point and mark up the Dividing plate. https://www.thingiverse.com/thing:4858421 That is the hardware. To get the printer to move correctly is what I have been looking for. Lots of video on making pcb but none on the programming of the g code. PCB free software can make a drill g code. But it isn't for a 3D printer exactly.

 

[TSutrina]

For you that have a 3D printer and think it is accurate. Mount a brush toy motor with a small drill of center point and mark up the Dividing plate. https://www.thingiverse.com/thing:4858421 Another posting is https://hackaday.io/project/174733/instructions That is for an ender and has some information on Gcode.

 

[timo_gross]

For you that have a 3D printer and think it is accurate. Mount a brush toy motor with a small drill of center point and mark up the Dividing plate. https://www.thingiverse.com/thing:4858421 Another posting is https://hackaday.io/project/174733/instructions That is for an ender and has some information on Gcode.

I think G-code would be the least to pose a problem. Hand coding the drill program should be straight forward enough.
There are lots of obstacles and challenges more serious than the G-code.