cheap indexer

[jseiler]

My first prototype indexer cost under $100 and some people at the user's group meeting asked about it, so I thought I'd post a few more details.

For an indexer, first you need some way to hold both ends of the stock. Many are using mini lathes, rotary tables and tailstocks, etc. I started off with a hobby lathe designed to be turned with a drill. The headstock is 1/4" shaft. This setup won't turn heavy logs, but its not too bad for the price.

http://www.adjustableclamp.com/sp-71000.htm

Next, you need some kind of gear reduction for increased torque and a motor. Many use belt drive with timing belts, some use expensive gearboxes. I found this 18 to one gearbox with integral motor. Unbelievably cheap at under 20 bux with a 2.2A motor. Its a unipolar motor, so you have to run it half coil on bipoiar drives. With 18 to 1 reduction, its very torquey and I roughly approximate its backlash at less than 25 arcmin. If you run down the length of your turning, then increment the indexer, backlash won't be too much of a problem. You'd need to remove the brake (a couple of screws) and there's a wierd coupling on the output shaft with a friction pin. A nail and a hammer made quick work of that.

http://www.mpja.com/prodinfo.asp?number=16237+MS

Hooking this up to the headstock. I bent a piece of lexan with a torch and drilled some rough holes. This piece of lexan got bolted right to the headstock bracket and used the pretapped holes in the gearbox to mount to. With oversized holes in the bracket, I could slide everything around to get good alignment. I used a lovejoy spider coupling to hook the motor output shaft to the headstock shaft. A few wraps of electrical tape removed the backlash from the coupling.

Grand total, under 100 dollars. I'll try to post pix tomorrow.

John

[GripUs]

Now this could be a major news story. Please keep us updated!

I'll not move from my computer until I get the whole story...well, maybe for lunch


Joe

[jseiler]

Here are some pictures. This was rev 2. I made the bracket a little more stable by bending it twice.

I tried to pull my order history from mcmaster carr to see which coupling pieces I ordered, but that's down right now. Its not a 1/4" on the gearbox side. Might be 3/8 or 1/2, but I can't remember for sure.

John

[jseiler]

I took the indexer apart because I couldn't remember the gearbox output shaft diameter. Its 10mm. Lovejoy couplers are nice because you can specify one side as one size (even metric) and the other side as a different size. The headstock shaft is 3/8", not 1/4". Age + memory = wrong.

John

[Bob_S]

John;
Have you cut any samples yet?
I'd love to see some pictures.
I am thrilled with how simple this seems to be. I think I would still need to get another control board for my shopbot, but I don't see any reason why this won't make indexers available to a lot of us without spending thousands.
Thanks

[jseiler]

Got one pic. I brought this to Pittsburgh as a sample. This is about 6 inches tall. I used a 2x2 in red oak (1.5inx1.5in). I rounded it down to make a blank, then cut this with a 1/8" ball nose. It took a while, but it came out nice. Shined it up with the Beall buff system (which is great for buffing 3d stuff in general).

What I did on my benchtop shopbot is to swap out the y axis with the indexer and update my unit values so save buying another drive board. Using standard 2.0 aspire, I scaled the model so that the y size of the model was the same as the circumferance of the blank. I zeroed to the outside of the blank. Aspire never knew it wasn't cutting a flat model.

Aspire 2.5 should make this a lot easier to think out.

John

[jseiler]

To answer a PM I got privately in a forum, so everyone can benefit...

The lovejoy couplings needed, I got those from use-enco.com. You need 2 hubs and 1 spider.

Hub #1 goes on the headstock. Model #990-4045 (.375 bore, 1.08" overall diameter)
Hub #2 goes on the gearbox shaft. Model Model #327-8227 (10mm or .3937" bore, 1.08" overall diameter). If you change from these lovejoy couplings, you'll need to make sure the bores and outside diameters match up or the spider won't fit right.

For the Spider, you have two choices, Buna-n or Urethane. The buna seemed slightly sloppier to me, but wrapped with a little bit of electrical tape around the whole coupling, it might not matter at all.

Buna-n Model #990-4042 (1.08" diameter)
Urethane Model #619-3475 (1.08" diameter)

Sometimes use-enco.com has a coupon code for free shipping, but I couldn't find one that's valid this month. Maybe someone else knows if there's one out there. Otherwise, shipping will get you for 7-10 dollars depending, almost the cost of the items you'll need.

On the lathe part:
You'll also need to apply a little threadlock on the headstock thingy that bites into the wood because its threaded but doesn't have a lockscrew. A hoseclamp also works to hold it in place so it doesn't unthread itself (this is what I used when I was cutting, but its not shown in the pics I uploaded).

John

[JCTalbert,LLC]

I'm always impressed with the ingenuity of this crowd.... making something from nothing... unreal...just shaking my head at how simple you seem to make this...LOL

[DanielBEE]

I have something that was rigged up long before I had aspire to try making some base fluted base mold that could turn corners.. Anyways it can be manually turned and locked in at like 8 different sides, or has a motor on it that turns at some unknown speed. Is there a way for me to take this "contraption" and somehow sync it with aspire ? I have never attempted or even looked into circle work like this. Where do I begin if I know nothing.

[jseiler]

@Jim

This was quite easy. The first one I made, the bracket flexed because I used 0.09" lexan. The next one I used the same lexan, but folded twice which really stiffened it up. If I were to do it again, I'd probably do it out of aluminum and cut the holes needed with vcarve/aspire to perfect size, then bend it with the proper tools. However, the lexan works suprisingly well and you can look through it to see where you need holes drilled. I just marked those spots with a sharpee and drilled them. Simple as pie.

@Daniel

Can you post a pic of what you are using now (maybe in a separate thread)? I can't quite picture what your apparatus looks like.

More info...

There are 6 wires on this motor, making it a unipolar motor. On one coil, you have green and white-green stripe. The common for that coil is the white wire. On the other coil, you have red and white-red stripe, and the common is black. For my old shopbot drive, a bipolar drive that only uses 4 wires, I used white and white-green for the A coil and black and red for the B coil. The motor is spec'ed at 2.2A and 3.3V. I have no idea what its inductance is so I can't offer any guidance on how much voltage it will take off a really hot drive system, like a gecko 540 etc. I used 28V and 1.5A, without incident....your mileage may vary. Maybe someone who has more experimenter experience with steppers may be able to make some better guesses. I just presumed the motor was a lot like my pk266A which seemed to be good enough.

Legaleeze...for educational purposes only, blah, blah, blah, if something bad happens to your drive due to my ignorance or your own, I'll move before you can sue me. I'm also poor, that's why I make $100 indexers Be careful out there.

I got the wiring configuration using an ohm meter. I got about 3.3 ohms on my cheap meter between the ends of the coils and about 1.7 ohms between commons and their ends.

John

[jhop]

Thanks for posting. I might have to give this a try.

Jim

[dewalt58]

I got the gear box on order with MPJ, sounds good to me for a an R&D project to test the wrapper with. I'm thinking more in the lines of an alum angle to bolt to rather then acylic, just hopeing there's not too much backlash in the gear box. Question John....how do you set your "steps per inch" on the rotory stepper? Is it by the diameter of the part? Diameter at the bottom of your design?


AL

[jseiler]

@AL

I do mine differently than many will because I am replacing my y axis drive with the stepper motor. To do this, you must essentially remap circumferance to y distance. I generally set up my model such that in the aspire model in the y direction works out to be the same as the circumferance of the rounded blank. That just makes it easier for me to make sure the math is right and the model joins at the 0 and 360 point correctly.

If you are doing the same thing (swapping out an axis), then you have to figure out how many steps per inch, which is a little tricky and will depend on your particular motor drive. With my old shopbot box, using the y in the normal linear mode, my unit value is 2000. This corresponds to a 5 tpi ball screw, 200 steps per motor revolution and 2 pulses per motor step (half step mode). 5*200*2=2000

When cutting something round, I have to change this to:
200 steps per motor revolution, 18 motor revolutions per gearbox revolution, 2 pulses per motor step, 1 revolution per 360 degrees and 360 degrees per circumferance . Circumferance = (2 * pi * blank radius).

For my setup, this works out to be:
radius =1, unit value = 1146
radius = 2, unit value = 573
radius = 3, unit value = 382

Here's an example with numbers.

For radius =1
200 * 18 * 2 * 360 / (360 * 2 * pi * r) = 1146
(if you notice, the 360's cancel out, I just left them hopefully because it makes the explanation clearer).

For everyone that has a separate A output driver calibrated in degrees, you have to figure out how many pulses you have to output per degree of motion, I guess.

I suppose it should be:

200*18*2/360 for a half step driver configuration. Since I don't have an A axis drive yet, I'm not 100% sure this is right. I'll have to burn that bridge when I get to it.

You'll also have to set your accelations and maximum velocities accordingly. Since I was using this in a mostly indexing mode, I was very conservative with acceleations and feed rate. I picked a y max velocity that clocked out to be about 10 rpm, or one complete gearbox revolution every 6 seconds. It probably could have gone quite a bit faster. I picked these numbers by experimenting and I didn't write them down in my notes. Theoretically, the gearbox shaft output torque should be about 1600-2000 in oz....not bad for less than 20 bux.

Its easier than it sounds, and its really easy to verify you got everything right...send the command to turn one circumferance (either distance or 360 degrees) and make sure it actually does the dance.

John

[BSImages]

I am truly impressed.

[dewalt58]

Thanks John! For the info....got the MPJ stepper gear box yesterday and I must say its a sweet deal for the money! Should work well for a small home made wrapper! Do you think it will turn a 4x4 stick of wood? Will be a while before I can get into designing a mount for it, more then likely first part of next year after the holidays. Looking foreward to making it dance for sure!!
And the color code for the wires check out ok!!

AL