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

[minh-thanh]

DIVIDING PLATE ..DIY ..Help...!!
I am finishing the gear cutting tool, it can cut spur gear and Helical gear
I am making a dividing plate to cut gears
The outer diameter is for cutting even numbered gears, the inner is for odd numbered gears
The outer diameter is relatively easy to divide by angle = 9 , but with the inner diameter, the angle of each part will be 9.23 degrees
The question is: How to divide accurately with an angle of 9.23 degrees ?
Do you have any suggestions ?
Thanks

 

[Jasonb]

Do you have a DRO on your mill if so just enter the number of hole sin the PCD function.

 

[minh-thanh]

I don't have , I wish I had one

 

[Richard Hed]

I don't have , I wish I had one

Do you have CAD? If so, simply print the pattern on paper and transfer to metal.

 

[timo_gross]

How many holes? 39?

https://www.blocklayer.com/circle-centers

(not sure about the auto generated subtitles)

Can you describe more, what you want to do? It might give someone more idea what makes sense.

Following advise is just 2nd hand (seen somewhere) or untested ideas.

1. paper as suggested by Richard: I saw someone: (guess where)

• using a Laserprinter to print out the pattern from the CAD program. A3 paper from office or copy shop can add more acuracy.
• he glued the paper on a flat wood or plastic
• carefully drilled the center and attached it to the lathe head stock spindle.
• He used a thin wire and a magnifiying glas to find the next position.
• Drilling holes with a motor in the tool post.

2. You can order the hole plate from a PCB maker. The 80 mm x 80 mm PCBs are cheap and there is no rule that you need to use them for electronics.

3. I guess it requires absolute care and precision.

• put some sort of clamp on milling machine that allows to clamp the disc and some fence parallel to the x-travel.
• drill and ream two holes in exact the center and at the radius. (using the most care you can and only x-travel)
• some fixing of the center hole to keep exact position is required
• calculate gauge block stack required. (use the first hole with a pin inside & center hole like a "in situ" sine bar)
• drill 2nd hole (rinse and repeat until full circle is complete)

Greetings Timo

p.s. if you can find a 39 tooth gear somewhere and a small piece of matching rack, you could glue the gear to your workpiece and use the rack as a "stop block". That is how they did it on my tool grinder.

 

[Ken I]

Lift the XY co-ordinate for each hole from your CAD model and pitch it out - if you can rely on your screws.

You can copy your model centre to 0,0 on your drawing then "list" the selected holes and "print" the list.

Since working towards the lash can be problematical you can set your 0,0 in the bottom left "corner" then all dimensions are with the lash.

Regards, Ken I

 

[stanstocker]

I'm assuming your plate either has a center hole or can have one for mounting. Cut a piece of masking tape exactly the circumference of the plate, and mark out the number of divisions. Apply to the perimeter of the plate and rig up a pointer. Rotate to align the pointer with a mark, hold and punch or drill. Rinse and repeat. It's going to be a long say in the shop...

If doing high counts, consider making a large disk wrapped with a stick on metric measuring tape and rig up a pointer. It's just a plod to get it done, have a list before you begin and keep track. I've seen masonite disks used on the outboard end of the lathe spindle with a diameter of 36 inches. Wrapped with an imperial tape, this gives 576 distinct locations, and you can easily interpolate to twice that. double that if the tape is in 32nds. A one meter diameter gives 1000 distinct "steps", again interpolation will get you to 2000 locations readily.

Generating plates from first principles is sort of a pain. You can always scribe a circle of known circumference and set a pair of dividers to allow marking out the number of divisions for that ring of the plate. This can be even worse that all the shenanigans of mounting an outboard large disk and an arbor for the workpiece in the lathe and deciding on punch or drill from a tool holder on the cross slide.

William Smith wrote a series of articles in Home Shop Machinist about clock making where he went in excruciating detail on all of this. He also produced a video and books on tooling the shop for clock making.

Once calibrated for correct size, you can print "tapes" up to around 11 inches long (longer on the diagonal) with a printer and wrap them around your dividing plate to be. You could also print out a drawing of the plate to make and just stick it on the stock. Depending on how much use this tooling will get, it might be reasonable to just print a drawing of the plate to a disk and spray it with lacquer or clear coat. Direct index on that with a pointer.

Many folks have a variety of gears on hand, just the wrong pitch or material. Doesn't mean you can't mount them up somehow and use them for indexing.

Instead of all this, I'd spend my time putting together a stepper based indexer, either with a reasonably large ratio timing belt and micro stepping to get decent accuracy, plop a stepper on a small rotary table (this group has posts on running these with Arduino controllers), or buy the Sherline automated rotary table. All the stuff that comes along with making and using plates gets very old very fast. When my clock restoration business started getting quite busy I stood up a small 4 axis Sherline based CNC to let me make gears. Cranking through making a 96 tooth gear is really neat the first time. By the fifth one you are looking for a better way

Good luck with it all however you proceed!

 

[Steamchick]

A good question. Just like post #5 Timo has already explained.
My Dad asked my brother to use his computer (CAD?) to divide a circle by the number of odd teeth he wanted, then after it was posted to my Dad he did exactly what Timo suggests: Using a pointer and magnifying glass, set-up for drilling the late as Timo explained.
Then when he made the gear it was adequately good.
For "regular" gears he simply used a gear on the back-gear of the lathe (Now I wonder how many of you understand or have used a Back-gear?) with a dog to lock the gear in place while he cut the gap between the next gear teeth on his blank.
Ha also explained that you can use the gear train for the lead-screw to get a multiple of the indexing you want, by careful selection of the change wheels, and a locking dog. For 39 teeth, you want a 65 tooth change wheel with intermediate gears to give the 39/65 = 3/5 ratio. - say a 60 and a 100 tooth gear? Or a 30 and 50 toothed pair? (But his small 1920s lathe was not fitted with change wheels!)
Just a matter of working out the maths per gear position for what you want? Does your lathe have change wheels?
K2

 

[Steamchick]

Or maybe I have "screwed" that one up?
Should it be put a 65 tooth gear on the mainshaft, then a 25 tooth gear next, and finally a 75 tooth, and index the 75 tooth gear by 5 teeth at a time so the mainshaft does 1/39th of a revolution?
Naahh, that's wrong too!
I'm sure someone can figure it out.. I'm just thick this afternoon.
K2

 

[minh-thanh]

Thanks for the comments !
I am planning to divide the degree and drill each area to limit the error
I will divide it into 3 areas : from 0 - 120 , 120 - 240 , 240 - 360
Start drilling from 0 --> 120 , 120 --> 240 , and from 240 --> 360
If the deviation is 0.05 degrees ( A pretty crazy number.)then the final deviation is 0.65 degrees at 120 , 240 and 360 degrees ,
If I drill from 0 --> 360 then the deviation will be : 0.05 x 39 = 1.95 degrees - quite large .
Also, I think I'll need to print a larger template so I can divide each angle into smaller
Still thinking,,,,,

 

[timo_gross]

Instead of all this, I'd spend my time putting together a stepper based indexer, either with a reasonably large ratio timing belt and micro stepping to get decent accuracy, plop a stepper on a small rotary table (this group has posts on running these with Arduino controllers), or buy the Sherline automated rotary table. All the stuff that comes along with making and using plates gets very old very fast. When my clock restoration business started getting quite busy I stood up a small 4 axis Sherline based CNC to let me make gears. Cranking through making a 96 tooth gear is really neat the first time. By the fifth one you are looking for a better way

Good luck with it all however you proceed!

I guess a sherline divider will cost quite some "Dongs" The Arduino solution can be financially interesting, provided there are some existing mechanical parts.

Or maybe I have "screwed" that one up?
Should it be put a 65 tooth gear on the mainshaft, then a 25 tooth gear next, and finally a 75 tooth, and index the 75 tooth gear by 5 teeth at a time so the mainshaft does 1/39th of a revolution?
Naahh, that's wrong too!
I'm sure someone can figure it out.. I'm just thick this afternoon.
K2

65 / 25 = 2.6 .... 2.6*75 = 195 ..... 195/39 = 5

Suddenly I also get confused, but it should be correct. Conclusion yet another method. 3d printed change gears for manual indexing!

 

[minh-thanh]

"Dongs"

 

[ajoeiam]

Thanks for the comments !
I am planning to divide the degree and drill each area to limit the error
I will divide it into 3 areas : from 0 - 120 , 120 - 240 , 240 - 360
Start drilling from 0 --> 120 , 120 --> 240 , and from 240 --> 360
If the deviation is 0.05 degrees ( A pretty crazy number.)then the final deviation is 0.65 degrees at 120 , 240 and 360 degrees ,
If I drill from 0 --> 360 then the deviation will be : 0.05 x 39 = 1.95 degrees - quite large .
Also, I think I'll need to print a larger template so I can divide each angle into smaller
Still thinking,,,,,

View attachment 159954

Its not the easy route but you could plot the location of each hole use trigonometry.
Its much more tedious but it does work.

 

[clockworkcheval]

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.

 

[Mike N]

Build a Arduino Index head. Just tell it how many holes you want and push a button. If you search on this forum, you will find out how to build one. If you use the correct sketch code it will be very accurate.

Arduino Rotary Table for Dummies​

 

[skyline1]

Conclusion yet another method. 3d printed change gears for manual indexing!

I have done this, It worked really well. I have also 3D printed index plates for an indexing head, so you could probably 3D print the actual plate or a template to drill a stronger one.

Best Regards Mark

 

[Steamchick]

Hi again. (You'll be fed-up with my confused brain by the time I get it right?).
I have slept, and my head is a bit clearer now.
Fit a 65 tooth change wheel onto the Mainshaft. of the lathe.
We know 5 teeth will give 5/65 = positions for machining a13 tooth gear.
But you want a 39 tooth gear, so we need to index the 65 tooth gear by 1/3 of 5 teeth.
If you set a pair of gears together, a 75 and a 25, when you index the 75 toothed gear by 5 teeth the 25 toothed gear indexes by 1/3 of 5 teeth.
So couple the 25 toothed gear onto the 65 toothed gear, then index the 75 toothed gear by 5 teeth to get the mainshaft to index by 1/3 of 5 teeth, which gives 3 x 13 = 39 index positions.
You can prove this by turning the 75 tooth gear 15 teeth, when the 65 tooth gear should index by 5 teeth - and it will be an exact indexing.
QED?
Then you are using precision machined parts to copy by mechanical indexing, rather than relying on Optical setting to a template.
That is what I call a "machining" solution, not a "best by eye" solution.
Someone clever can even calculate the tolerance "drift" or variation and any clock maker will explain which is the better method - I hope? Another thing I don't know.
Always learning...
K2

 

[minh-thanh]

Hi All !
Thanks for the comments
I am thinking about this machining method
I will use a panme to adjust, calculate the ratio between mm and 1 degree from there deduce 0.2 or 0.25 degrees is how many mm of panme.
For example: 1.5 mm of panme then the angle will rotate 2 degrees, if I need 0.23 degrees then the panme will be 0.173
if 2 mm of panme then the angle will rotate 2 degrees then 0.23 degrees will be 0.23 mm of panme . if it is 9.23 then it will be 9 degrees plus 0.23 mm of panme
Adjusting on the pamme is relatively easy and it will give a fairly accurate angle

 

[timo_gross]

Y

 

[indian]

Back cover of the zeus book