Arduino Indexing (Dividing) Head

[jm82792]

Thanks for listing what you bought.
Out of curiosity (being rather clueless yet attempting to learn over the past two years)
how could I cut pieces larger than the collet shaft's bore?
I'm working with wood so I was thinking I could mill out circular plate of corian or delrin,
attach a 1/2" shaft to it one way or another, then from there use the adhesive (the miraculous wonders of double sided tape or hot glue ) of my choice to fasten wood or whatever to the plate. Obviously this wouldn't be very precise but I've yet to make a single machine part for anything, so relative accuracy is to me is better than global accuracy.

The other choice I had for a 4th axis was a 30 watt servo attached to a 1:100 harmonix drive, but I lost the bid due to my lapse in attention, and I WON'T (it was hard to find since it wasn't called by any useful keywords) be finding another one for $50 again any time soon.
To thoroughly ice the cake I saw a youtube video fo a guy using a similar model for milling aluminum

[RCaffin]

I could mill out circular plate of corian or delrin

I have machined Corian. It was a big mistake. It put aluminium hydroxide dust all over my machine, coated the leadscrew and got into the ball nut. NEVER AGAIN!

Delrin is fine.

Can you machine up a proper metal face plate somehow?

Cheers

[jm82792]

I've yet to machine any metal other than engraving brass
There has been always that desire to do so, but I've yet to have a reason to try.
I have some bits that should work, although I don't want WD40 or any cutting oil around my machine since it would contaminate my wood.
I guess compressed air and some patience is needed.

[RCaffin]

I don't want WD40 or any cutting oil around my machine since it would contaminate my wood.
I guess compressed air and some patience is needed.

What metals?
If you are talking about aluminium, try methylated spirits (and asupply of fresh air!). Works well.
Steel - more tricky.

Do you oil your timbers? You could try misting with a little bit of that oil (and lots of air blast). It's just to wet the surface of the cutter so the metal does not stick. The air blast clears the chips out so you don't get 'bumps' from re-cutting shavings.

Cheers

[jm82792]

I don't typically use oils although I'm trying out a tung based oil with a topcoat of finish with some improvements over straight finish.
Realistically I could use a cutting fluid, I'm paranoid about my lacquer or waterbased not laying out right.
Either way I'll make it work, but firstly I'll just try delrin when or otherwise if I'm able to go with what's posted here since I'm still bringing ideas in

[revwarguy]

how could I cut pieces larger than the collet shaft's bore?

If you look at Figure 14 on the website I show a $1.99 arbor which is placed in the collet that allows you to mount round objects with a 1/4 inch or 3/8 hole in the center.

[jm82792]

Okay I see what you mean, I can chuck whatever I can into it that's the proper diameter shaft wise.
This will be my project for early next year I think

[revwarguy]

The Digital Machinist Article is now out, and if anyone is interested the Arduino code for the project is available at Digital Machinist » Home and select the "Interactive...Downloads" menu items, and the right download for the Winter 2013/2014 issue (Vol8 No4). Support for the project is at http://bbs.homeshopmachinist.net/for...ital-Machinist

Thanks,

[RCaffin]

> The collet shaft is like here for $17 and some shipping.
Ah - CTC. I buy a lot from them.

Bearings - I get via eBay. To be sure, they may not be as good as the best Timken or SKF, but for low revs they seem very adequate.

Cheers

[revwarguy]

I've made an enclosure for the unit, shown here:



There is a bit more about the box at Snailworks Electronic Index Build Log

[revwarguy]

There is a new version (2.2) out for this device, available for download on the Digital Machinst website. It has an enhancement and some bug fixes. as well as clearer information about the wiring. More at the website here.

[ZS5AND]

Hi,
I was trying to get the zip file for your very nice project, but cannot access www.digitalmachinist.com/downloads.
Seems like its down. Cannot access the site from here in RSA.
Is there any chance of getting a copy of the zipfile ?

Thanks
AJ

[revwarguy]

That's not the right URL - tryDigital Machinist » Downloads

[ZS5AND]

Thank You
will get to it asap.

Andrew

[revwarguy]

There is a new version, version 2.3 available from the magazine website. This version allows the denominator for the gear ratios to be defined at compile time. One reader emailed me that he had a 4.5:1 ratio, so this allows him to specify a 9:2 ratio. Also, he was from Sweden, so he wasn't used to Fahrenheit, so I added the ability to display Celsius as well (also a selection made at compile time.)

[karlheinz10]

I am new to this, and the project is quite interesting, but I have some doubts:


As does the program to compensate angle errors?
As divides the circumference into equal parts if it is a prime number?
I would like to read a logical and understandable response, I hope not to sound arrogant!

[black7185]

Hi,
How did you connect the wires for the switches mounted on the enclosure to the LCD Keypad shield ?
Did you remove the switches in the shield? Is there a pin out on the Arduino Or shield?

[revwarguy]

I am new to this, and the project is quite interesting, but I have some doubts:

Hi Karl, sorry it took so long to respond.

As does the program to compensate angle errors?

Not sure exactly what you mean by angle errors here, but if you mean rounding errors from fractions that do not result in whole numbers, that is a matter of the precision needed, as is the case in all machining. There are no perfect circles in the real world, only those are are accurate enough for a particular purpose. Stated another way, before you worry about the extent of calculation errors, you have to answer the question, how much accuracy do you need? All physical parts have a tolerance to them. If you machine a cube that is 10 cm on a side, you will always be able to construct a device that will measure some deviation from 10.0000000 cm on a side. The only question is how many zeros you need, and how much you are willing to spend on measurement devices to even see those inaccuracies.

How this applies to the step indexer is a matter of the fixture. You can effectively select the amount of precision you need by setting the gear ratio of the fixture high enough, and then make sure the software knows about it. (How to do this is described on the web page.)

As divides the circumference into equal parts if it is a prime number?

Again, it doesn't matter - how many decimal places do you need, and set the gear ratio of the fixture and software to meet or exceed that.

I would like to read a logical and understandable response, I hope not to sound arrogant!

Not at all, I only hope I have answered your questions!

Good questions.

[revwarguy]

Hi,
How did you connect the wires for the switches mounted on the enclosure to the LCD Keypad shield ?
Did you remove the switches in the shield? Is there a pin out on the Arduino Or shield?

I didn't remove the switches, I thought that would be way too much trouble. I just soldered wires on the switch solder pads on the PC board to wire the panel switches in parallel with them. IIRC, I just used a multimeter to determine which switch pins to solder to.

[RCaffin]

As does the program to compensate angle errors?
There is a bit more to that question than meets the eye. I cannot answer for other controllers, but I do know that Mach3 does all its calculations in either single precision or maybe double precision floating point, inside the PC. This may be a far higher accuracy than some older non-PC controllers can manage. Then it works out how many pulses to output from the result.
BUT - it retains the last floating point value for the next calculation. It does NOT round it off and lose precision over a number of calculations. So if you want to divide 360 degrees into 127 equal steps, just do so. After issuing 127 of those steps you will end up back at 360.0000 degrees. The accuracy of each step will of course depend on the resolution of the drive.

Cheers
Roger