An indexer for drilling fusor grid holes

Stuff you made and how to make cool stuff

An indexer for drilling fusor grid holes

Postby Doug Coulter » Tue Sep 23, 2014 4:59 pm

Yes, I know, a few of you here have said this won't work. But no one stepped up to make me grids, either, so...I mad a jig that's just gotta be better than the cam-timing degree wheel magnetically stuck to my lathe chuck - I'm lucky to get half a degree accuracy with that; which is the important parameter here, all the holes could be a bit off center and not affect anything much, as long as they are in a circle and nearly concentric with the middle - spacing is much more important for this lens.

Here it is.
http://youtu.be/dLRHJM6Vtzk


I found out (again, duh) that a 1/4" drill doesn't make quarter inch holes. They come out a bit oversize in my case - and yes, I tried shimming and other foolery, but that piece just wasn't going to work out, so I made another. But a "D" size is a few mils too small, so with that, and a 1/4" ball driven through the hole, I managed a tight fit on the precisely .250" motor shaft, in my 1.25" OD by 1.75" long 6061 t6 stock. After a little foolery to get the thing chucked up again with perfect hole concentricity (well, about 2 tenths), I was able to machine the rest of it in a couple chuckings and fiddlings. I added a setscrew for the motor shaft, not sure why - it's a very tight fit unless the aluminum is really hot - and 4 more to capture and center the grid-end, which for this design is in the range of 3/4" OD by 1.25" cross section (starts out as a square cross section, the last step is hand rounding that after all the holes exist).

According to various sources, I can expect around 3% of a step accuracy in position with a common stepper. This was one of the nicer ones in my collection, so hopefully it's better, but that would be good enough anyway (for now). This is a 1.8 degree/step motor, I'm full-stepping it as I only really want 45 degree hole spacing just now. It's rated at 4v, 1.1 amp, and I'm running it at about half that (limitations of the wall wart that also drives the arduino circuit). I built a total-overkill driver, with 4 IRL-1104 fets, and a current limiting resistor (Dale, 1% 6 ohm, 25w). Like I said, utter overkill. And no worries - the wire is an adequate fuse should something go wrong with fets of this spec...and they are easily driven straight off the arduino digital outputs with the stepper library. What's shown in the movie is just a demo - I will add some software to parse inputs from a PC for commanding it, maybe add a pushbutton or switch or two if I want stand alone manual uise, this isn't rocket surgery after all, and a PC control is probably overkill for this use...but I'll do that anyway, since it's not that hard to add. It'll be more trouble to make the USB-B and wart jack square holes in the right places in the aluminum box this is going into.

I haven't made any grids with this yet, but it's sized to duplicate the one we are now setting new records with, but make it more accurate than I did with that degree wheel and the lathe toolpost drill. The proof will be in the pudding...on which I'll report when I have something to report.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: An indexer for drilling fusor grid holes

Postby Jake Gray » Wed Sep 24, 2014 1:04 pm

I think that is really slick. Should be as good or better than a good dividing head. I like spider for centering your grid pieces as well. Are you using any tricks to start the holes?
Jake Gray
 
Posts: 25
Joined: Wed Sep 15, 2010 4:29 pm

Re: An indexer for drilling fusor grid holes

Postby Doug Coulter » Wed Sep 24, 2014 4:42 pm

It\s definitely better/more-repeatable than my cheap Indian circle table, and mounts the grid-ends better. Whether it's better than I'm doing now, by hand with a racer's cam degree-wheel on my lathe chuck (at least it's big = 16" diameter) we'll know soon.

The trick for starting holes..is to simply have a flat surface. I don't do the chamfering/rounding over of the grid ends till after drilling for that reason. The graphite I'm working with is soft as chalk, but less abrasive - the drill goes right in without any noticeable skipping around. The main source of inaccuracy now might be flex in the bit - they made the .020" part way too long for the holes I'm drilling, and it can kinda get into wave mode at high rpms. What I do now is "tune" the speed for minimum deflection, but at the wrong speed it can easily be 1/8th wide in a blur at the drill bit tip.

I can stabilize this with a finger tip, but "who knows" how much I'm bending it off course doing that. A drill jig, unless made out of tungsten carbide itself, would be a wast of time, since these drill bits side cut quite well - and are WC themselves. It's one more issue to solve. Perhaps I can just find a vendor that makes these bits with only about .125" skinny part (the back end is 1/8th to fit standard collets). I'm only drilling holes about .11 deep at most, I don't need a .4" long skinny part like what McMaster ships for these.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: An indexer for drilling fusor grid holes

Postby Doug Coulter » Fri Sep 26, 2014 7:42 pm

A little progress on this. After scrounging around, I didn't find a good power supply that would fit in one of my die-cast project boxes, not waste a ton of power/hear and so forth, but I did remember I have a ton of old AT and ATX power supplies, mostly good, from old computers or upgrades. Bingo, efficient switcher, and since the motor ran fine at around 1.8v (rated at 4v, the engineers must have been thinking darlington drive transistor and a 5v supply with losses), I figured the 3.3v output would do fine - the drop in the drive fets may as well be zero (look up specs on IRL 1104 - utter overkill but I had them in stock, so they're free).

Some folks may say this sloppy build - not even a box but a good ole board - shows lack of pride in workmanship. Some will say it's overkill. I suggest either one look at Pareto optimization and then get back to me. Or as Robert Heinlein said "too fussy wastes man-hours". I had all this stuff laying around. No, it's not optimal if one were going to build thousands, or if one wanted to program it for different angles at the job - I'll actually have to fire it up hooked to a computer and take, dunno, 30 seconds or so to change that. Whoop. de. do. Mounting, wiring up and writing code for an at least 2 digit thumbwheel switch would probably take longer than I will ever spend just changing the number in the code and re-uploading it.

Not done yet is extension cords for the motor windings and the pushbutton to make it move. I'll get that done tonight, and later, show the whole mess leaning on or laying on the lathe ways with the motor/jig in the chuck and the toolpost drill in action.
That is, when I have something to drill. Since we moved to 3/4" grid OD, I decided to quit wasting my 1" stock and just buy some 3/4 from McMaster, which will happen real soon now, since I've finally got enough other stuff I need from them to make shipping worth it. Early next week - they ship fast.

But here's what it looks like now. Just screwed to a board so it'll lay on the lathe ways, or lean on them.
GridJigProto.JPG
Proto now tested. This motor is much stronger than I expected, and faster - but I slowed it down because it seems to look cooler that way.


The entire mess draws 12 watts from the wall, about 7 of which go into the motor, the rest is power supply "tare" and the arduino uno.
For which this is the code, folder and all, for your sketchbook:
StepperDriver.zip
Unzip in your sketchbook directory
(1.1 KiB) Downloaded 398 times


Note, I always just go ahead and move the serial speed to 115,200 baud...on just about every project. You rarely run into a "real" serial port that might barf due to buffer overflow anymore, it's all USB dongles, and they have plenty of buffer for most any project, and this one doesn't push it, it just types one short line when it moves the motor (debug). You don't need a computer present for this to work, once built.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: An indexer for drilling fusor grid holes

Postby Jerry » Sat Sep 27, 2014 2:34 am

That's why you use the Teensy instead. None of that crappy USB to serial stuff. Real USB, a fraction of the latency of an arduino. And cheaper than an Arduino. And you are supporting an american company. They also run Arduino code too so there is no real difference in programming, thought he Teensy 3.1 has 16 bit ADCs (13bit practical)
Jerry
 
Posts: 573
Joined: Sun Jul 18, 2010 12:07 am
Location: Beaverton, OR

Re: An indexer for drilling fusor grid holes

Postby Doug Coulter » Sat Sep 27, 2014 11:01 am

I'll look into those, Jerry. I'm not a huge fan of arduinos generally, it's just that I had some 10 uno (seeduinos) I got from yes, China, for something like $50 - including shipping, laying around. I've been using them for data aq too - via oversampling and sigma-delta math, I'm getting ~14 bits of "goodness" out of even those 10 bit a/d's. I vastly prefer the chipkit (PIC based) stuff, probably as much from habit as anything, since I've been doing pics for decades now, and like the IO.
Even with the arduino, it's the PC tools that limit the baud rate, actually, most terminal emulators will only go 115,200, even though my perl code can go lots faster, and I generally use human-readable ascii and a terminal emulator as part of the debug process. At any rate, that's all the arduino/chipkit ide's will let you have as well. Hasn't been an issue yet for what I'm doing with these guys. I was just warning potential users of my code that they'd have to set the default up to 115200 instead of the 9600 it (the IDE) comes with. Java...don't get me started. This dev environment is about the lowest thing you could possibly call an IDE, at least running in linux with the open java runtime, but probably even if you use Oracle's. Click on an error in the compile window and go to the error? Nope. Add multiple files for better organization? Possible but difficult. Adjustable syntax highlighting so it looks like my other C tools? Haven't found a way yet. Easy for noobs, very lame to an experienced coder.

I had a bit of a setback last night...I discovered it's a bad idea to knock over a beer onto the lashup, and found that old PC power supplies don't like beer (though the rest was fine with immersion in it). The next old one I pulled form the junkbox didn't like loads on the 3.3v supply (and yes, I had the sense wire properly hooked in) so I had to go to 5v on the stepper, which makes it pretty hot (everything else is ice cold - and the drop across a turned-on driver fet measures zero - it's something under 5mV).

So, today I get to make the extension cord for the motor, and one for the pushbutton that will have some magnets so I can stick it to some handy place on the lathe. I'll simply use really high gauge wire for the motor and accept (want) some voltage loss there where any resulting heat is spread over 6' or so of wire, rather than concentrated on a diode drop or resistor.

I'm kind of in a hurry to finish this and get my desktop cleared so I can get going on Philipp's MCA design - he sent me one, without other distractions. Part of the optimization...and maybe it'd be worth the time to put a beer holder that wouldn't allow bottles to tip over on my bench somewhere...that little oops wasted a lot of my time. I couldn't find out what was wrong with the beer drinking supply - nothing burnt, no bad fuses - it just quit working after a nice fat arc showed through the holes in the case. I'll maybe scrounge some of the ferrites out of it. At least they are cheap (free in my case, I have a large box of these computer supplies that I'd like to be able to close the lid on). Sometimes my local computer store even gives them away.
They're too big, too noisy for a lot of things, so finding a use for them at all seems worthwhile.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: An indexer for drilling fusor grid holes

Postby Doug Coulter » Sat Sep 27, 2014 3:11 pm

OK, here it is in action, sort of. I don't have any stock the right size to drill with this at the moment, but I will soon. I used #26 wire to make a 10 foot extension cord for the stepper to get an additional bit of series resistance without having any hot spots. The motor now only reaches about 110f (in 76f ambient) at equilibrium, and while that wire has to be getting warm, the heat is so spread out along the length you can't feel it. I decided against making a fancy extension cord for the go button, since accidentally hitting it would break the drill if it was in the hole, it's not that hard to reach when you want to as is.
http://youtu.be/sqvYT_IsJeQ

Now to get some 3/4" graphite to machine into grid ends and make a new, more accurate grid. The proof will be in the pudding.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: An indexer for drilling fusor grid holes

Postby solar_dave » Sat Sep 27, 2014 4:09 pm

Doug have you tried a #000 center drill to start your hole, the pilot is 0.020 and is about 0.030 long. Some good magnifying goggles might be in order.
Dave Shiels

My TED 5000 power monitoring
http://phx-solar.no-ip.info:8081/Footprints.html
solar_dave
 
Posts: 108
Joined: Tue Aug 21, 2012 12:33 pm
Location: Phoenix, AZ

Re: An indexer for drilling fusor grid holes

Postby Doug Coulter » Sat Sep 27, 2014 8:02 pm

No. Didn't know they existed. Great find! I will most definitely try one. Heck, they aren't even expensive!
Thanks! I already have 5 and 10x gunsmith goggles.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: An indexer for drilling fusor grid holes

Postby Jake Gray » Mon Sep 29, 2014 12:21 pm

Jake Gray
 
Posts: 25
Joined: Wed Sep 15, 2010 4:29 pm

Next

Return to Machining and Fabrication

Who is online

Users browsing this forum: No registered users and 3 guests

cron