Bradbury Engineering and Light Hauling

Change of Plans, for the Better

noreply@blogger.com (Dana Bradbury) — Mon, 21 Oct 2013 12:25:00 +0000

Last post I talked about the jacked-up mill spindle. My plan was to find a used R8 spindle to replace the bent Kwik Switch 200 spindle. Well I found a better option. Midwest CNC is a company right here in Indy, they rebuild Hurcos from this era and are very familiar with my machine. They sold me a used Kwick Switch spindle in good shape, with the bearings.

When I picked it up Mike at Midwest had the spindle on a surface plate and showed how he measures the run out with a 10th indicator. Very cool and the run out was only 2 tenths which is within factory specs for a new spindle. Add to that it has a new nut, so it is kinda brand new to me. Best of all he showed me the process of removing the old and installing the new. Turns out to be pretty straight forward, and well, the job is finished and the mill is up and running. I trammed the head and set the vice last night, then started making chips.

A video just for fun, the replacement spindle going into the quill, lifting it in using the knee.

Triumphs and Setbacks

noreply@blogger.com (Dana Bradbury) — Thu, 17 Oct 2013 13:53:00 +0000

I found some problems with the mill, the spindle is out way out, like .008" out. And not out of round, bent, I think. See the video of the indicator. And the problem gets magnified the further down I go on the tool holder.

My friend Fazni came Saturday at about 3:00. Went through the mill and lathe. Even with the bent spindle it is mostly good news. He really knows how and where to use an indicator to figure what is out and what is not. He used it to check the backlash on the ball screws, there is pretty much no back lash. And he said the ways and screws are in great shape. And the mill overall is a really nice piece of American iron (even though it was cast in Spain!). The only thing wrong is the bent spindle. We are speculating it was crashed hard and that's why it was taken out of service. They probably didn't want to mess around replacing the spindle, so they sold it. That just means we get too!

We talked about several hacks to fix the spindle. But looks like the right way will be to just bite the bullet and replace it.

It has an Universal Engineering Kwik Switch 200 spindle. A lot of people like these. But for my purposes I think it will be easier/better/cheaper to replace the quick change setup with a more standard R8 spindle. And I can sell the quick change tool holders I have to probably more than pay for the replacement spindle.

Once all the mechanical stuff is set we can look at adding the cnc stuff back on. The tight ball screws mean everything, cause we don't need glass scales for position - saving cost and complexity on the controls. I am betting it will be Christmas, maybe spring before all the mechanical stuff is fixed and everything is setup. So it will probably be next Christmas before I get the CNC stuff added. But that's OK, just got to keep plugging away. It was April 2012 when I brought home the first 2 axis CNC table project. Back then I said it would take six months to get running. That turned out to take a year, and it is still not much of a usable machine.

Getting Carried Away

noreply@blogger.com (Dana Bradbury) — Tue, 08 Oct 2013 13:36:00 +0000

I am still getting the shop set up, which has side tracked me from getting the door secured, which has side tracked me from working on the digital control board for the door opener. Life is funny.

What I have been doing is painting, cleaning, running electrical, lights, and such. All stuff that seems a little over the top for what is really a garage, a place to park a car - not a nice car - and a bunch of bikes and plastic kids toys.

But is there a good reason to do all this, I mean is it better to work in a clean organized room. I hope so, cause it seems like I am spending all my time setting a place to do fun stuff and am never going to get to do the fun stuff.

Small rotary phase converter for powering the 3PH machines

Lathe is Finally Home

noreply@blogger.com (Dana Bradbury) — Thu, 03 Oct 2013 10:13:00 +0000

The lathe I bough way back in February is finally sitting in the garage, right acoss from the mill. Hoping to get power to them this coming weekend.

New Home for New Toys

noreply@blogger.com (Dana Bradbury) — Wed, 25 Sep 2013 18:23:00 +0000

I have put off work on the door controller for a bit to finish-up the garage. Exterior is pretty well finished. Inside is getting dry walled today. And power company is coming to connect the meter soon, maybe this week. Soon we'll will have a place to use our new Hurco knee mill, and the lathe that has been waiting in the wings since March. Not to mention the woodworking machines coming up from the basement. Nothing else to say, I'll let the pics do the talking.

Foiled by 3D Space, Again

noreply@blogger.com (Dana Bradbury) — Mon, 09 Sep 2013 14:10:00 +0000

I etched the first version of my launchpad breakout board. The etching process is a pain, and the results are not great. I want to go back to machining the boards - it works well when everything is setup correctly. But having a physical board in hand and being able to place the components in the real world served it's purpose.

Fig 1 - etched board with components placed

Fig 2 - Lauchpad placed on breakout board

Fig 3 - Kentec touchscreen booster pack place on top of the launchpad

It's hard to see in the photos, and I guess hard to see in 2D, like when designing using EagleCAD, but the stack height is wrong. The big black relay on the upper left side is to tall, it prevents the touchscreen from seating completely. I think it is an easy fix, I have room to slide the whole launchpad/touch screen assembly over to give clearance. Or I can flip the relay to the back side of the board. I think either will work.

I guess what I am wondering is how could I have foreseen this? Should I have built a full 3D model? That would probably have been more work that etching the prototype and finding the error, as I did. But finding the problem at this stage feels dirtier, like a mistake, versus what could have been finding the problem as part of a robust design methodology.

But, there is another problem, one a 3D model would not have revealed, least not to this novice. I didn't realize some through hole components like the relays have to be soldered from the opposite side - they sit flush to the board. So even if I had caught the relay stack height problem I would still be making changes to the board layout and re-making the board. Lets call it Opus 2.

Quick update and some notes to myself

noreply@blogger.com (Dana Bradbury) — Wed, 04 Sep 2013 03:21:00 +0000

I was stoked when I got back from the TI training and setup the touch screen for the door controller. I thought was just about done with my hardware prototyping. Nope. Somehow I could not get cheap-o RC receiver to work with the Launchpad. Did I actually do some damage when I botched the initial touch screen setup?

Let me back up. I was using the RC transmitter and receiver from a cheap RC kids toy as my garage door opener. I just connected the 4 pins from the receiver to four pins on the launchpad a BAM, I had a remote door opener. And it work, no really, there is video to prove it. But some changed, for some reason the Launchpad cuts in and out when I connect all 4 ins from the RC receiver. And they are not, I mean they are 0 volts. When I say cuts in and out I mean the board seems to reset continuously. I put a amp meter across the power jumper and the current just bounces on and off. And worse even with only a couple of RC inputs connected I was no longer able to consistently read ON or OFF on the LP pins. One thing I noticed was the RC rx pins were only reading about 2.5v when they were ON. This seems weird, and I don't think it use to be this way.

Well I kinda wanted to run this problem down, but I also have to keep an eye towards the bigger picture. I really think I prefer having my inputs setup with pull-up resistors like the LP buttons. So just the switch down to ground to read ON. And the RC toy remote was kind clunking and silly looking. So I ordered a generic remote control switch with a nice small remote for the car. It simply trigger a relay, so I will just use that to ground my input. And it is only 1 button, which will simply the connections, the PCB, and the use.

But note to self, I need to better understand pull-ups, pull-downs, and open drain. And I need to better understand how to trigger a pin as high, I mean is 2.5v enough? What happens when you apply 5v?

Anyhoo, I have moved on. The hardware is working and the basics of the code are there. I have all my inputs and outputs accounted for and working on the bench. At this point the project splits in 3 directions:
- Making the breakout board and enclosure for all the controls. Just generallin wraping up the electrical related hardware.
- Coding the final control program - need to account for all combinations of buttons being pushed, door state, stepper motor acceleration, etc
- And the mechanical side, need to finish the drive train design and get everything mounted, motors, gears, limits, locks/latches.

So long.

Touch enabled GUI? Now We are Cooking with Grease!

noreply@blogger.com (Dana Bradbury) — Fri, 16 Aug 2013 13:15:00 +0000

Went to a TI Tiva seminar at Arrow electronics in Chicago Thursday. There I revived a new Tiva launchpad which is the next generation of Stellaris - basically the same thing I have been using with a few upgrades and a new set of libraries. And they gave us a Kentec touch screen LCD boosterpack. The class was I think $30, they provided lunch, a Tiva Launch pad, the Kentec LCD - damn good deal. From what I have seen TI is very good about providing developers the tools and materials they need to well, develop.

I did do some damage when I returned home with my new toys. I took a quick look at the Kentec source code and thought I had a handle on the pins being used. And it looked like they didn't over lap any of the pins I was already using on my door opener prototype. So I plugged it right in and fired it up. All sorts of weird things happened, the screen was flickering, my LEDs on the breadboard were dim and flickering. Yeah, basically I was careless, turns out after reading the Kentec data sheet I was overlapping a bunch of pins. Still not sure exactly what all happened but I did manage to fry a USB cable. Not sure how the cable was the only thing damaged, but so far it is the only casualty.

Anyhoo, started fresh with a new cable, re assigned my pins with no overlap, and started integrating the touch screen into the door controller. Check out the video. Touch a button, open a door. OK really just spin a motor, but getting closer to opening a door.

Eagle Rocks

noreply@blogger.com (Dana Bradbury) — Mon, 12 Aug 2013 13:18:00 +0000

As I have mentioned several times, I am using Eagle CAD to design a PCB breakout board for the Stellaris launchpad. It is a pretty basic board, couple of relays, voltage regulator, and a whole bunch of screw terminals. The board fits the launchpad booster pack format. So my board will sit at the bottom of the stack, it will be a little bigger than the launchpad and have standoff holes for mounting. The launchpad board will sit in the middle and the Kentec LCD touch screen, another booster pack, will sit on top. Least this is my idea and I am getting close to having a board that will do the trick. Also, I am using viewplot to view and print the gerber files that eagle outputs. Gold is the top copper, red is bottom copper, and purple is shows the part foot prints.

Quick update and things I learned

noreply@blogger.com (Dana Bradbury) — Fri, 09 Aug 2013 17:30:00 +0000

I realized I had not explicitly set the clock speed on the Stellaris launchpad (AKA, garage door controller). Not sure what the default is, I think 16MHz. I should go back and check. Anyhoo, I thought I might as well set it to 40MHz like all the example projects in the Stellarisware workshop. So I did. Then I walked away for a couple of days. When I came back to work on other parts of my control program the stepper motor started stalling. I was freaked out at first, having forgotten the clock speed change. Once I realized what I had done I then realized the delay function I am calling to send the stepper pulses to the stepper driver is not am absolute time delay, but a count of some sort related to the clock speed. So when I bumped the clock speed to 40MHz I also increased the frequency of the step pulses - to the point where the driver stalled the motor.

So a small change broke my code, and then I learned something. I fixed the stalling motor by increasing the delay. Then I read up on motor control signals, looks like timers are preferred method. So I am working on that. Also, all these delays, clock cycles, timers, etc really tax my brain. I have using the scope more to visualize the square as I learn and implement new ways to generate the wave.

Looks like with my current settings on the stepper driver somewhere around 1kHz is about the max speed I can drive the stepper. I need to double check this, then back out the max motor speed based on the steps per revolution.

As an aside, I (well we, Kelly and kids are going to the Science and Industry Museum) am headed to Chicago next week for TI Tiva (Stellaris was renamed to Tiva) training. Should be fun, but I am real nervous not knowing anyone, and being a complete novice in the embedded space.

What's your work flow?

noreply@blogger.com (Dana Bradbury) — Thu, 01 Aug 2013 18:25:00 +0000

A physics professor told me once there is hard, then there is complicated. He said magnetism is hard, not complicated. Classical mechanics is complicated, not hard.

I find it useful to keep this in mind, always. I have inputs and out puts mostly worked out for the garage door controller. Now I need a way to actually make the physical connections from my switches, relays, etc to the launchpad control board. And I want to try to make it kinda professional looking so I want a breakout board with screw terminals for the external connections. The idea is surely not hard, just need to have a place for like 12 wires to connect to my controller - not hard but proving complicated.

Complicated cause I have to pick all the actual parts, the screw terminals, the header pins, the actual relays, make sure they are sized for the max current they will be subjected to, get them all lined up for the copper circuit board, blah, blah, blah. Complicated, not hard.

I am close though, I think. I am using Eagle CAD (free PCB design tool) to create the schematic which is a logically drawing of all the parts and connections. Then eagle will take this schematic and create a board based on the actual physical part dimensions. And from there Eagle will run a CAM job to create the geometry files to allow me to cut the printed circuit board. From there I take the geometry into CAM BAM, a free CAM program to generate the G-code to actually tell the mill how to cut the copper. Then set up the PCB in the mill with the right feeds and speeds and hope I don;t break a cutter, again.

A picture is a worth a 1000...

Milestone

noreply@blogger.com (Dana Bradbury) — Tue, 23 Jul 2013 11:34:00 +0000

Got the the big NEMA 34 stepper turning, controlled by the Stellaris launchpad, and triggered by the cheap kids RC remote control. That's all, just a quick video.

First oscilloscope use, yeah baby!

noreply@blogger.com (Dana Bradbury) — Fri, 19 Jul 2013 13:04:00 +0000

Quick video and pics, finally had a chance to use my Rigol digital storage oscilloscope. And already learned something.

So I got the inputs mostly sorted out last night, then wanted to get the output for the stepper motor working. I have done this in the past with the nerdkit, just write a loop that turns on and off a digital output pin and use that output as the input to a stepper drive. Most stepper drives just take a square wave (at least that is what I was thinking) as an input. Every time the drive sees a logic high it advances the motor one step. Here is the funny thing, in my previous nerdkit code I was turning on the the input for 5ms then turning it off. In my mind that was a square wave. And with a 5ms period there was no way I could see from the led I had in line that it was not a square wave. I mean the led was always on because the frequency was sufficiently high.

It wasn't till I was playing the period on the scope that I saw the my old way was really just creating a signal that was 0v most of the time with real quick jumps to 5v (see first pic). Turns out it didn't much matter cause the stepper drives are really triggered to step when the signal rises. That is, a clean square wave is really not necessary, I think. But is was cool that the scope revealed this misunderstanding in my head. And so I modified the code to delay 5ms both while the output was on AND off (see pic two). If that doesn't make since, don't worry. I am not really explaining it well, partly cause I am still trying to make sense of it all.

Really what I was trying to say was using the scope in this case was surly not critical but it gave a visual into what the electronics were doing and gave me a snippet of insight I hadn't had.

Who would have thought GPIO was tricky?

noreply@blogger.com (Dana Bradbury) — Thu, 18 Jul 2013 17:40:00 +0000

I Think I finally have the General Purpose IO pins worked out for the garage door controller. Couple of things confused me - actually still not 110% clear on most of this.

First, there are 40 break out pins on the Lauchpad, plus 6 breakout pads for external power. Trying to figure which ones I could safely use for my inputs and outsputs was, well, not obvious. Finally found this nice pinout picture that breaks down which pins can be used for what - http://austinblackstone.com/handy-stellaris-launchpad-pinout-infographics/. So, for now I am using 4 pins on Port D for the inputs form a cheap RC car receiver. The pins map to forward/backward and left/right on the remote control. And I am using 2 pins on Port B for limit switches.

Next problem was how to configure the inputs. The example buttons on the LP board are setup to ground the input pin, so they are configured with pull up resistors. I loosely understand this means that when the input pin is not grounded it will be pulled up to Vdd, or 3.3 volts. In logic terms it will be pulled high. So to trigger the input, we ground it and the pin goes to 0 volts and logic low, or logic 0. Not sure I fully understand the pros and cons here, but I get how it works and I can copy the code for my limit switches. I just configure Port B pin 0 and 1 with weak pull up resistors and connect my limit switch to ground the two pins when activated.

But what about the 4 inputs from the RX IC chip I hijacked? When the RC control buttons are pressed the corresponding IC pins read 3.3v. So I need my 4 pins on Port D to read logic high when I give 3.3 volts to the pin. I guess it turns that I just need to not configure the input with pull up resistor. Easy enough right? Well, not quite. The code to configure the input works OK, but when the pin is given 3.3v meaning I press a button on the Remote the value I read from the pin is 2,4,8 etc. I little googling later and it turns out I have to account for "bit masking". I still have to read up this, but I guess depending on the pin, a logic high could be one off 2,4,8 - multiples of 2. This is some kinda trick to allowing reading multiple pins at once, I think - Still working to understand it.

And one final problem I have not fully resolved. I may mangle some terms here, but this is power management in my brain. The board bus voltage is 5v, provided by USB, and Vdd, or the controller operating voltage, is 3.3v. So I am not sure how to handle power and ground between components. For now I have the RXIC connected to the 3.3v terminal of the lauchpad, and both boards share a common ground. All supply voltage comes from the USB cable. This works OK, but at some point I need to drive the coil of a 5v relay and I don't want to tax the USB cable, so I need an external power source. I found where this possible, but then I think I lose the ability to connect the USB debugger. So still got some things to figure out here. My bigger problem is how to supply the stepper driver, it takes step and direction from the controller but gets power from a 48v external power supply. Not sure if I can just tie the grounds together and hope for the best?

More to come.

Open Seseme

noreply@blogger.com (Dana Bradbury) — Tue, 09 Jul 2013 14:15:00 +0000

We are in Denver, CO for the next couple of days - arrived by way of Austin, TX and Carlsbad, NM. We'll be home Saturday then work on the Garage will resume. What remains is the metal siding and a proper garage door.

I have always pined over the garages doors in automotive service bays cause they let so much light in. So I set out to find one for our garage. Found exactly what I was looking for at home depot. But the price tag was sky high.

Once upon a time in a land far away I was going to be ag-engineer, work for John Deer or Cat, maybe work on some kinda fancy new robotic tractor. Fast forward 20 years and I am sitting in a cube in Carmel, IN writing Java for some nondescript corporation. It's a damn good job by most measures, just not where I thought I would be.

So there are companies, it turns out, in Indy who do the kinda engineering I think I could really get into. I think most people would call it controls engineering, maybe systems engineering. These are the guys who make your cruise control and anti-lock brakes work. They do the same kinda work in airplanes and jet engines and yes in tractors too. Modern tractors are full of controls and are approaching the point where a driver is not really needed.

My problem has been how to transition out of web development into controls engineering without basically starting my career over. And I think I am onto a path. My hobby micro controller programming (AKA Nerdkits and Arduino) has led me to more sophisticated embedded programming - using C++ on 32 bit ARM micro controllers. I actually did a little side project for a company to trouble shoot a bug in an embedded medical device. It was really cool and gave me the bug to learn more and figure out how I can do this kinda thing for living. Next month I am headed to TI training in Chicago to learn more about the Cortex M3/4 controllers and embedded controls.

Back to the garage door, I couldn't afford what I wanted and I was looking for a platform to learn more about embedded controls. And at the same time I don't really want to limit myself to embedded controls. I really want to transition to embedded software then into more hardcore controls engineering. So I need a project that will encapsulate a lot of disciplines. So I am building my own garage door complete with translucent panels for the light I crave, electric door opener, wireless remote, touch screen operation from inside the garage, and maybe an android app - least that is my plan.

What I have so far is a welded steel door frame running on an exterior metal track, a block diagram of the system, half baked Solidworks Assembly and a bunch of parts and pieces coming from China. I hope to do a post at least once a week, maybe more, till I finish the door as a way to motivate myself to treat this home brew barn door as a real engineering project.

Coming Full Circle

noreply@blogger.com (Dana Bradbury) — Wed, 22 May 2013 15:03:00 +0000

I am getting back into feedback control systems. This was one of the topics in engineering school that really fascinated me but that was covered at such a theoretical level it was always kinda voodoo to me - plus the teacher sucked. Well I am learning control theory and control systems don't have to be mystical. There is a great YouTube series I am working through - http://www.youtube.com/user/ControlLectures?feature=watch

Anyhoo, I am not getting into this randomly. My brother is getting into ROVs and AUVs, he is working in naval salvage Actually the biggest navel salvage ever attempted. He is like 2nd in command, you know, el hefe - heavy duty.

So I was looking into these under water robots when I remembered that this is where I got started with mirco controllers. It was like circa 2007 when I had the idea it would be fun to build a model submarine, use some radio controlled motors and such, build it from PVC pipe. Well, that led me to Google, and yeah lots of people had the same idea. Still nothing new under the sun. Well one search led to another and soon I was turned on to the guys over at Nerd Kits. The rest is history, but it is funny that I am coming back around to using micro controllers to control a hobby AUV/ROV, and will probably need a feedback control system to bring it all together.

More to come, I hope.

What do you use that for?

noreply@blogger.com (Dana Bradbury) — Mon, 20 May 2013 16:51:00 +0000

Nothing much new on any specific project, still working on getting the laser running "good". Talking to the guy in Istanbul who sold it. Getting tech support from someone in another country, in a different time zone, speaking a different language, is both frustrating and damn cool.

This post is just a couple quick thoughts.
- What do you use that for?
- Do you really need 4 routers?
- What do you need a lathe for?
- What is that thing?

I hear these kinda things from friends and family, a lot - use to hear it from Kelly but she gave up asking a long time ago. I have alluded to this kinda questioning in past posts. I am always just so confused when confronted with this line of questioning. It is not that I can't answer, it is more that I just don't typically find this way thinking useful.

There are two problems with this mind set. The first is more ideological (philosophical?) You don't need a reason to explore something that interests you, or something that you could learn from. Damn near everything I tinker with falls in this category. I don't have am immediate need for a 1800 pound metal lathe. Or a very low power inaccurate laser engraver, or a 1944 Farmall H. But they all get my juices flowing, and I learn something new every time I play with them. We, you know us humans, learn by experience best, not by rote. So every experience shapes and changes our view. John Dewy wrote about his idea, or at least that's what I have taken away from for what little I have read. And I am still reading him. Reading Dewy is I guess no different than playing with my pottery wheel. I throw a little clay then go do something else. Then come back to the pottery wheel.

But the other problem here is more pragmatic, or concrete. Lets assume you understand I like to build stuff. So when you build things you have tools and materials. Typically we work with the tools we have and the materials we are comfortable with, or that fit with our tool set. If you have a table saw, jointer, and planer, you probably work in wood. And if you need a work bench you would probably fashion it from wood. Likewise if you have a welder, grinder, and cut off saw, you probably work in metal. You get the idea, and this is OK. But if you can expand your tool set, or your material choices then you might find yourself making something new. So while I might not have needed a lather for the last 10 years it doesn't mean I won't find a use for it, and I will probably start to finding ways of making stuff with the lathe, just cause now it will be another tool in the tool belt.

That's all I have to say about that.

Burning Ring of Fire

noreply@blogger.com (Dana Bradbury) — Tue, 14 May 2013 03:47:00 +0000

Last post I mentioned maybe mounting a laser diode to the engraving table. Well that's what I did. I got a laser module from KaleCNC in Turkey - 2 watts 808nm (infer red spectrum). Here is a short video burning some paper. I have a lot to do to make this thing usable. Need to focus better and work out feeds and speeds. But this is a good start.

Going Round in Circles

noreply@blogger.com (Dana Bradbury) — Thu, 18 Apr 2013 12:02:00 +0000

It's not just a phrase. I actually have the two axis engraving table going round in circles. Check out the video and pic below.

Almost exactly one year ago I started tinkering with a little two axis engraving table. It is a nice table with ball screws and some really beefy stepper, but no controls. My goal then was to control the table with MACH3, a PC based CNC program. I could have gone straight for a 2 axis stepper driver and been off and running. But instead I went down a bunch of more complicated paths, mostly just cause I wanted to learn a little about motors and electronics. Some of the attempts are here, here, and here. Well there were a lot of other things that happened over the last year, but 3 "heres" seems like a good round number

Anyhoo, I finally I felt I had learned enough and it was time to get real. So I ordered a 3 axis driver card from Ebay - real nice little card, has optically isolated parallel port breakout, spindle control and inputs for an e-stop and limits switches. So after a year of jacking around I got a board for $40 from China and in one night had it connected to my steppers and mach 3.

Next step will be to tidy up all the connections, make them more permanent, then figure out what to mount to the carriage. I am thinking a laser diode to cut paper and engrave wood would be fun to try.

Made a Little More Permanent

noreply@blogger.com (Dana Bradbury) — Thu, 11 Apr 2013 04:48:00 +0000

In a previous post I mentioned using an easy driver and the Nerdkit to control a stepper motor. The last couple of weeks I have been trying to tidy-up some old projects. So one of the things I did was move the stepper driver from the breadboard to a a more permanent prototype board. And I added the bigger bad-er 4 line LCD. So now I have a complete package to install in a machine with one stepper motor. This is not just a random thought, I actually have a plan for a micro controller stepper driven cross cut sled for my table saw. But that is a whole other story. Check out the pic, video, and good music!

With Warm Weather Comes Outdoor Projects

noreply@blogger.com (Dana Bradbury) — Wed, 03 Apr 2013 13:02:00 +0000

It would be easy to assume lack of posts means I am not doing much. Turns out the opposite is true. I have been ramping up on several new projects and trying to complete several old. One the outdoor front the big thing is the new garage. Foundation is poured, the slab will poured by the end of the weekend, then framing can begin. I got the bright idea, again, to do some custom form work for the foundation walls. They cameout OK, but not great. And it took more time and material than I planned. This is my last DYI concrete project. They never go as well as I think, always cost more and take to long, and never come out as good as they should.

On the pottery front not much happening. I have not thrown in more than a month. Several projects are waiting on the garage completion including a place to setup the pottery kiln. Then we hope to start firing and glazing the 40+ items setting on our shelves!

What else is waiting on the garage? Well the lathe, it is cleaning up well but I kinda back-burner-ed it until I have a permanent home. And the tractor is waiting. What, you didn't know we have a tractor - Hell Yeah! It is a circa 1950 Farmall H, and it is bad ass. Our neighbor pretty much donated it, he got it hung up in a ditch and no longer needed it. No, it does NOT run. But it is really fun to sit on, just ask Isaiah. Thanks to friends Josh and Jugg for helping to pull it ( and my stuck truck ) out of a muddy field to bring it home.

And on the electronics side a lot is happening too. I have been doing some embedded C++ work on a ARM Cortex M3 microcontoller, pretty neat, but it's a steep learning curve. One the one hand this chip is way more powerful than the 8bit AVRs I have used. But on the other, a lot of the stuff is the same. It is kinda kicking my ass, but it's fun.

To keep me motivated and because I want to learn more about embedded controls, and well cause I like buying things, I have ordered a bunch of toys. I bought a digital oscilloscope, a bunch of nerdkits parts, and several stepper motor driver cards. I have three micro controller projects started that have been lingering - some for multiple years. I hoping to bring them to a close and hoping in at least one case use the new scope to trouble shoot a I2C communication problem I am having between the nerdkit and a little digital altimeter. It's not that I need an altimeter, but understanding and being able to troubleshoot I2C problems is fairly import to using many of the commonly available digital sensors.

So that's an update for March, hoping, as usual, to post a little more frequently.

Too many toys!

noreply@blogger.com (Dana Bradbury) — Thu, 28 Feb 2013 13:39:00 +0000

The Sherline/LabVolt lathe conversion is finished mostly. I mounted and configured the spindle encoder and it works, sort of. It is reading half the RPMs I think it should. I suspect I either have a pulley ratio defined wrong, or the motor control is not setup correctly. But I have kinda tabled that problem for now. I still need to button-up the case, but I was able to run through the full CAD/CAM workflow, I drew a profile in CamBam, generated the toolpath, posted the g-code, and ran it through MACH3. It actually worked, crazy. Now what can I actually do with this little CNC lathe? Pretty sure no one reads this blog, but if anyone needs a small part cut, shout out, maybe I can help!

In the last post I mentioned the Hardinge CNC lathe project for Fazni. Not much new there, we have all the "parts", just need to find time to do the conversion. But, now I have a whole new Hardinge lathe project. OK, Actually it is a Feeler 618 project, it is a clone of the venerable Hardinge HLV. I bought it on Ebay, in unknown condition, it was in Orlando FL, and weighs in at 1800lbs - BAM!

Lets call the picture below the official before picture. And lets also hope that at some point in the near future it is all cleaned up and running, in our new garage, so I can post an after picture. Else, I will have to post a scrap yard picture, and that would break my heart.

The story of how I got the lathe up to Indy will have to wait for a separate post. At present, it is sitting at Mike's shop where I am slowly (slow progress seems to be a recurring theme) assessing it's condition and cleaning it from top to bottom, inside and out.

In case you didn't know, this lathe is a real bad boy. It is widely considered the best "tool room" lathe ever made. And it is really the only lathe I have ever used. Back when I was an intern at a medical device company I used this same model nearly everyday. And it is so nice to use.

Now you may be wondering what I could possibly do with a lathe like this? Well you're not alone, everyone I have told about it has asked the same question. But it's just not a useful question. I mean, by analogy, I have taken a lot of math classes, including the same calculus series every undergraduate engineering student takes. And I have never, out side of a classroom, had to integrate a circle to find the area, or take the derivative of a polynomial function to find the instantaneous rate of change. That doesn't mean the math classes were useless, or that I should skipped them. And that's all I have to say about that.

Big Lathe Might Live - Teaser

noreply@blogger.com (Dana Bradbury) — Tue, 05 Feb 2013 19:49:00 +0000

Nothing new on the the Sherline lathe, still need to install the spindle encoder. And we went to Siesta Key, FL for a week, that was nice! But just cause I was out of town didn't mean I couldn't keep working on basement projects.

Ok, in this case it is not my basement but Fazni's shop south of Bloomington, IN. He has a nice Hardinge lathe with an old (1985) CNC controller. Because the little lathe conversion went well we want to convert his big lathe. This will give it a new lease on life and make it much more useful to his work. So this post is really just a teaser, couple of pics below.

Basically we ordered similar stuff from CNC4PC only this time we are dealing with servo motors versus steppers. So it is a little more complex, but not too bad, I hope. Parts should be in any day now, hope to get into it next weekend. Stay tuned.

Close up the cross slide and servos. The smaller motor things are resolvers and tachometers. They will be replaced with quadrature encoders, on per axis.

Everything except for the power supply is being replaced in this bulky old box.

Little Lathe Lives

noreply@blogger.com (Dana Bradbury) — Thu, 24 Jan 2013 12:06:00 +0000

Quick update on the Labvolt (Sherline) lathe conversion:

- Boards all mounted
- Wires all pull and connected
- New front panel made with E-stop and reset switch - Mike cut them on the BOT, made two in case I decide to retrofit the 5400 mill.
- Everything except the spindle encoder is connected and configured in Mach3.
- I had one weird problem where I can't enable backlash compensation without stalling the stepper drives. Got to track this down at some point, but for now .001" backlash is OK
- Ran a program using one of the Mach3 turning wizards. Worked great and the part which had a .250" diameter measured .252 on the dial calipers. That's pretty good for not having done any real tuning yet.
- So two things remain, mount and configure the spindle encoder and permanently mount the parallel cable through the case. These will wrap up the hardware side. Then I can get into the software side and see about how to get parts out of CAD (Rhino or Solidworks) into CAM (camBam) then into Mach3.

Here is a video of me cutting using the manual jog controls:

Pottery is Piling Up, What We Going to Do

noreply@blogger.com (Dana Bradbury) — Thu, 24 Jan 2013 05:32:00 +0000

I am roughly 1 year into making pottery on the home brew kick wheel and I am starting to feel less like a complete novice. Really most of my time on the wheel has been since October - lets call it 4 months of somewhat consistent throwing and a little glazing.

I have not mastered centering but I am over the hump. I can usually center the lump in two rounds of kicking. Kick it up to speed, cone up. Kick it back up to speed and center down, then fine tune, might need one more round of kicking. It was a huge relief to get to this point.

Next problem for me, and yeah I think this pattern I am unfolding is true for every one starting out, is pulling a wall. Initially I lost about half of the attempts. Now I hardly ever lose the clay at this stage.

So what I am on to now is being able to really make the clay do what I want, and then be able to repeat it. So far not much luck. I have made nice pieces but not because I planned them. And I still lose some lumps if I get too high or get too fancy, like trying vases. To gain control and learn the repeat throwing part I started just focusing on mugs. Only problem is mugs need handles "Need" meaning if you want the average Joe to recognize your work as a mug and not for example as a small flower pot. And I am just not ready for handles. Sure I could learn, but right now handles are not a priority.

So next ides was/is to try tea bowls, more formally Chawan. Made 9 attempt a couple nights ago and had some decent results. Two pics below, first is close up of several things that sort of pass for tea bowls, second is a shot of what is piling up in my basement waiting to fired - about 30 pieces and that is just what I think is worth firing. At some point I am going to need to fire my work more quickly or in bigger batches - Ellen Cornelius has helped but I don't want to wear out my welcome. When Spring comes some kinda kiln is going to be in the works, not sure what style but it should be fun.