Thread: 0.3 Module Gear Cutting Query

Originally Posted by cba_melbourne

A very worthwile project is to fit a stepper motor to your rotary table with an electronic divider box. It will let you choose any number of divisions (and much more). You then can then sell your old division plates.

I bought the circuit board for the control box from Steve Ward in the UK (http://www.worldofward.com/rotarycontroller/). Here the price list: http://www.worldofward.com/rotarycontroller/shop/
The software is free. You additionally have to buy a box, a stepper motor, a stepper driver and a power supply and some connector/cables from eBay. And an Aluminium bar offcut to make an adaptor to mount the stepper motor onto the rotary table. Its a nice DIY project that will give you many years of use. Once you have made it, you cannot understand anymore how you could waste time setting your rotary table with division plates.

Easier for some than others Chris but I like the idea. I had a go at making a motor speed controller from a Jaycar kit and all it made was smoke.

Bob

This might be of use to make a 107 point reference circle

http://www.cgtk.co.uk/metalwork/reference/divider

Then print it and stick it onto a piece of mdf and use that as your dividing reference as this link below explains - hes a forum member called nearnexus. It wont be NASA accurate but maybe good enough ?

http://users.picknowl.com.au/~gloami...25rev7-n2.html

Bill

Here's another idea. My colleague Martine drew and printed on a sheet of paper a 600mm diameter circle the circumference of which is divided 107 times.

IMG_4836 (Large).JPG

The first part of my idea is to glue the sheet onto a 12mm MDF backing which would have a steel hub at its centre facilitating mounting on probably my 6" Vertex rotary table. (The Vertex is lockable whereas the Schaublin dividing head's quill is not.) The next part of the idea is to mount my centring scope rigidly on the mill's table and use the scope to align the intersections of the printed lines. Then the gear cutting could commence....

024 (Large).JPG

A 0.5mm error at a radius of 300mm equates to about 0.03mm at 18mm radius. With the centring scope's 20x magnification less error would hopefully be achievable. Hopefully .

For context here's a photo of a similar microscope - NSKT-100-R-500x500.jpg

Originally Posted by steamingbill

This might be of use to make a 107 point reference circle

http://www.cgtk.co.uk/metalwork/reference/divider

Then print it and stick it onto a piece of mdf and use that as your dividing reference as this link below explains - hes a forum member called nearnexus. It wont be NASA accurate but maybe good enough ?

http://users.picknowl.com.au/~gloami...25rev7-n2.html

Bill

Hey Bill, I must have been one finger typing when you posted this!

Hi Bob I'm sure this crowd will make you one. Don't know the price or shipping but... http://www.pearlengineers.com/spur-gears.html Phil
ps they are in India though.

The final gear diameter is around 32-33 mm, I guess the template diameter is 6 or 7 times that, so you should potentially get 6 or 7 times the accuracy of your centering scope on the finished part.

How wide are the printed lines?

Originally Posted by RayG

The final gear diameter is around 32-33 mm, I guess the template diameter is 6 or 7 times that, so you should potentially get 6 or 7 times the accuracy of your centering scope on the finished part.

How wide are the printed lines?

I can't really say until I try to measure them with a calibrating slide Ray.

In the olden days when AutoCad was used in this office, printed line width matched pen width, ie your 0.5, 0.35. 0.25 and so on mm Rotring technical pens or whatever your choice of pen was. Now they use Microstation and Revit. Line widths are 0, 1, 2 ,3..... I asked Martine for the finest line.

BT


An update from Martine - Turns out she used Coreldraw - Line weights in corel are: hairline, 0.1, 0.2, 0.25, 0.5, 0.75 etc... or also totally customisable.

‘hairline’ on print - isn’t specified but is thinner than 0.1 mm

Bob there are many different types of 3D printer technologies around, and unfortunately most not associated with the field just bundle them all in to "3D printing". It's a little like saying a "metal working machine", as it really could mean many different things in reality.

The most common type of 3D printer at the cheaper end of the market is thermoplastic FFF (also called by the trademarked name FDM), but SLA (the original 3D printing technology) is also becoming more popular as the resin becomes less expensive. SLA printers expose the resin to UV light cure the resin. I very much doubt an FMA printer would be able to print that gear directly, but SLA may. However I'm happy to print you a dividing plate if you want to go down that route? Send me an email with the details and I'll model it and get it over to you.

I don't have anything engineering related to hand that I think would be useful as a demo, but here's a little guy that I printed before work for one of my daughters, sitting on a 20 cent piece to give you a sense of scale and resolution. That's about as close as my iPhone will focus, and sorry but it looks like it's focussed on the coin and not Marvin. It's obviously just a useless widget for a 4 year old, and no particular care was taken in printing it, however I think a printed part would be adequate for something like a dividing plate that would be used just a few times, and you could just run a drill bit or reamer through the printed holes to get the size correct for your indexing pin. While I've never done a dividing plate, it's this type of hack work I use the printer a lot for. You will of course still be left with finding a cutter.

I can print nylon and it is a very useful engineering material that many poo-poo unfairly. I will certainly be re-making some of my primary lathe gears in nylon to replace some metal ones as the nylon runs much more quietly. However the CAD program I currently use isn't yet set up to model gears accurately. Once it is I will print them and then probably recut them with a conventional cutter to maximise the accuracy.

Marvin 2.jpg

Originally Posted by Anorak Bob

Here's another idea.

I'm not saying that wouldn't be "close enough", I doubt you even need all the teeth, does the gear turn through a full 360? I doubt it would be as accurate as dialing in the numbers. Papers made out of wood, at the level you're talking about it changes size with humidity*.

The good thing about making a plate is the improvement in accuracy you can using the worm. Though you'd want to be sure you can fit 107 holes on your max dia and then you have the worm as an unknown.

Just use excel to print of a list of the angles you need. Use your shaper, it will make grinding the tool easier

You know its going to be fun which ever way you go.

Stuart

*though maybe drafting paper is better......

Originally Posted by Stustoys

I'm not saying that wouldn't be "close enough", I doubt you even need all the teeth, does the gear turn through a full 360? I doubt it would be as accurate as dialing in the numbers. Papers made out of wood, at the level you're talking about it changes size with humidity*.

The good thing about making a plate is the improvement in accuracy you can using the worm. Though you'd want to be sure you can fit 107 holes on your max dia and then you have the worm as an unknown.

Just use excel to print of a list of the angles you need. Use your shaper, it will make grinding the tool easier

You know its going to be fun which ever way you go.

Stuart

*though maybe drafting paper is better......

Bob, Stuart,

Excel? that's the hard way... Do you have a DRO that can do PCD?

Ray

Originally Posted by Pete F

Bob there are many different types of 3D printer technologies around, and unfortunately most not associated with the field just bundle them all in to "3D printing". It's a little like saying a "metal working machine", as it really could mean many different things in reality.

The most common type of 3D printer at the cheaper end of the market is thermoplastic FFF (also called by the trademarked name FDM), but SLA (the original 3D printing technology) is also becoming more popular as the resin becomes less expensive. SLA printers expose the resin to UV light cure the resin. I very much doubt an FMA printer would be able to print that gear directly, but SLA may. However I'm happy to print you a dividing plate if you want to go down that route? Send me an email with the details and I'll model it and get it over to you.

I don't have anything engineering related to hand that I think would be useful as a demo, but here's a little guy that I printed before work for one of my daughters, sitting on a 20 cent piece to give you a sense of scale and resolution. That's about as close as my iPhone will focus, and sorry but it looks like it's focussed on the coin and not Marvin. It's obviously just a useless widget for a 4 year old, and no particular care was taken in printing it, however I think a printed part would be adequate for something like a dividing plate that would be used just a few times, and you could just run a drill bit or reamer through the printed holes to get the size correct for your indexing pin. While I've never done a dividing plate, it's this type of hack work I use the printer a lot for. You will of course still be left with finding a cutter.

I can print nylon and it is a very useful engineering material that many poo-poo unfairly. I will certainly be re-making some of my primary lathe gears in nylon to replace some metal ones as the nylon runs much more quietly. However the CAD program I currently use isn't yet set up to model gears accurately. Once it is I will print them and then probably recut them with a conventional cutter to maximise the accuracy.

Marvin 2.jpg

Hi Pete,

Thank you for your generous offer.

Michael pointed out earlier that 107 divisions is out of the realm of dividing plates and me thinking the Swiss might have had something clever awaiting me in my set of plates, found only disappointment.

An intense discussion with our mathematician Martine revealed that while a plate could possibly be made with indexing holes to enable a division of 3 degrees 21' 52.15", multiples of that division get messy. But a 107 notch direct indexing plate similar to that illustrated in the dividing head photo I posted earlier in this thread, might be achievable with a printer.( Might not too!) Rather than a full plate, could a notched ring be made allowing it to be glued to a largish diameter metal disc thereby reducing the amount of printable material needed?

Bob.

Originally Posted by Stustoys

I'm not saying that wouldn't be "close enough", I doubt you even need all the teeth, does the gear turn through a full 360? I doubt it would be as accurate as dialing in the numbers. Papers made out of wood, at the level you're talking about it changes size with humidity*.

The good thing about making a plate is the improvement in accuracy you can using the worm. Though you'd want to be sure you can fit 107 holes on your max dia and then you have the worm as an unknown.

Just use excel to print of a list of the angles you need. Use your shaper, it will make grinding the tool easier

You know its going to be fun which ever way you go.

Stuart

*though maybe drafting paper is better......

Stu,

The Swiss Sihl Utoplex tracing paper I draw on is more moisture sensitive than any of the white printer paper we have at work.

Originally Posted by RayG

Bob, Stuart,

Excel? that's the hard way... Do you have a DRO that can do PCD?

Ray

Ray.

My inclusion as an addressee was obviously an error!!

Ned.

Hi Ray,

Originally Posted by RayG

Excel? that's the hard way... Do you have a DRO that can do PCD?

I mean for just dialing the deg:min:sec on the rotary table to make the gear.

Originally Posted by Anorak Bob

An intense discussion with our mathematician Martine revealed that while a plate could possibly be made with indexing holes to enable a division of 3 degrees 21' 52.15", multiples of that division get messy.

But the divisions on the plate would not be off by more than 5".
What did the mathematician say happens when you use that plate to index the gear through a 60:1 worm?

If you have something that a plate will fit on and you wound like someone to PCD a plate for you, I'm happy to do that.(not sure it would be more accurate though)

Stuart

Originally Posted by Anorak Bob

Hi Pete,

Thank you for your generous offer.

Michael pointed out earlier that 107 divisions is out of the realm of dividing plates and me thinking the Swiss might have had something clever awaiting me in my set of plates, found only disappointment.

An intense discussion with our mathematician Martine revealed that while a plate could possibly be made with indexing holes to enable a division of 3 degrees 21' 52.15", multiples of that division get messy. But a 107 notch direct indexing plate similar to that illustrated in the dividing head photo I posted earlier in this thread, might be achievable with a printer.( Might not too!) Rather than a full plate, could a notched ring be made allowing it to be glued to a largish diameter metal disc thereby reducing the amount of printable material needed?

Bob.

G'day Bob, yes a notched ring could be made, no problem, but I wouldn't be too concerned about the amount of material, at least if I'm printing it for you.

I'm not sure why it's not possible to do an indexing plate. Keep in mind that I'm not a mathematician, and I've also been working since 0330 this morning (sorry if I woke you up this morning), so my mind is a bit mushy at the moment, but I can't see why an indexing plate wouldn't work. I am limited to 200 mm x 200 mm, so what I would suggest is doing the indexing plate in multiple radii circles to make the indexing plate physically smaller. So to get 107 indexing holes you could probably get by with 3 circles, 2 of 53 holes, and the third with the one odd hole. Even though they are the same angular distance apart (3 degrees 21' 52.15"), the fact that every alternate hole is offset will prevent the holes overlapping each other. A third circle will need to used to pick up that odd hole, as 107 is a prime number.

Hopefully that explanation makes sense.