This video is and introduction of my motion control slider project, showing the basics of what it can do and how it’s used.

Time Lapse Assembly Video

I figured that I’d record the entire assembly process of the MC slider from start to finish. This video shows about 24 hours of real time in 15 minutes, and in it I discuss some of the problems I faced and design choices I made.

Feedback

If you’ve got any comments or random abuse to hurl, please post it here on the ol’ blog.

Seriously, I’d appreciate any and all suggestions and I’m happy to answer any questions you might have. As promised in the videos, various downloads and a parts list are below.

Licensing

The YouTube videos are Creative Commons – Attribution licensed. You’re pretty much free to do whatever you want with the downloads below, however I ask that you provide proper attribution if you use them in your own project, and that you let me know about it. (Not because I’m some kind of jerk, but just because I’m really curious to see what you come up with!)

Downloads

Parts List

Most of these parts are from Amazon. I’m a sucker for 2-day shipping, and they have [almost] everything I need. The prices below are what I paid; Amazon’s prices can fluctuate widely and rapidly, and I didn’t necessarily look for the best deals on this stuff. You can probably put this together for less money, and if you price these parts out elsewhere, please let me know in the comments!

I also used some CAT 5e cable, six resistors, some miscellaneous nuts and bolts, and of course parts from my Makeblock Kit.

The Makeblock Kit

I bought the Ultimate Robot Kit back in September 2013. It was $288.99 at the time, and I’m kinda proud that I was order #311. (I ordered from them again in February 2015 and they’re up to number 23,705!)

The kit is expensive, yes, but the parts are of good quality and are extremely useful for projects like this one.

For the reason of the cost, I’m not going to say that you necessarily should use Makeblock parts. It’s just what I used. It’s hard to say what the beams, timing pulleys, motor bracket and etc. cost me for this project, but it’s probably about $50 if you were to pick them up individually.

Tools and Supplies

I had to buy a couple of extra tools and other miscellaneous for this project. You might have these already, and there are other ways to skin a cat. I posted this list in case you were curious as to what you saw me using and talking about in the time lapse video.

Video Transcript

This is the original shooting script for the Intro video. I figured I’d just throw it in here rather than re-typing all that same information.

Hi, I’m Scott and I made this video to show how I put together a… uh, where are you going?

{Camera moves back}

In this video I’m going to show you a motion control slider that I put together… This one right here, actually.

It’s based on one of the cheapest sliders available, and it has a stepper motor hanging off of it that’s connected to an Arduino. And that’s in this rather large controller along with the user interface components.

Besides just moving the sled back and forth manually, the controller can record and play back a series of movements and speed changes so it can do its thing without having someone to operate it.

So why a motion control slider? Mostly, I made it for fun. But the original impetus came from my HPRC/Pelican review. Towards the end of that video I had a sequence which showed gear being unloaded from the Pelican.

Check it out in full size. The transitions aren’t very smooth, and there’s tons of artifacts in that last one.

That’s because I don’t have the steadiest hands in the world. I did four or five camera passes, at four or five different speeds. It was one of the last shots of the video, and I got lazy: Rather than reshooting I just threw a warp stabilizer effect on it. That smoothed out the apparent speeds, but introduced all the artifacts.

So instead of using my own shitty hands, I figured I should just build a machine to do the same job.

Here’s a reshoot with the HPRC case being unloaded, using the motion control slider. Much smoother.

Besides shots like that, motion control is just the ticket for stop-motion animation, shooting in different settings for later composition, or just to act as camera operator when you’re a one-man show like me.

Here’s a slow motion shot of me dropping some RAM on the table. It was recorded at 60 frames per second, but you’re seeing it at 24 FPS.

Listen. This was my first real Arduino project, and I’m far from a mechanical or electrical engineer. I’m going to be telling you how I did this, but it’s not necessarily the right way to do it. If you go to s.co.tt/mc-slider you can download my plans, source code, and a materials list and make your own improvements. If you do, please let me know!

First, here’s a look at the slider itself. It’s pretty simple — those aluminum bars, the timing gears and the motor — all the mechanical stuff — are from a Makeblock kit. The switches came from Amazon, and it’s wired up using plain old CAT5.

My first mistake (and I made a few) was not planning where to mount the interface box, which is why it’s just sorta hanging there. But anyway, the box has a microstep driver attached to it, and all that’s inside are some wiring splices. It interfaces with the controller using an old-fashioned DB-25 male-to-male cable.

The controller is an oversized ABS plastic housing with power and slider connections, power switch, a potentiometer to control speed, a 4-row 20-column dot matrix LCD, and six buttons to control movement, recording and playback.

If you’re interested in details of the design and assembly, I made a narrated companion video which shows the entire process from start to finish. It’s time-lapse so you don’t have to be completely bored out of your mind as I screw around trying to figure out how to put everything together for hours on end.

Anyway, here’s how it works: The camera attaches to the carriage — or sled — whatever it’s called — as usual.

As soon as it’s turned on, the controller goes into calibration mode. At each end of the track are micro switches that are closed when the sled hits them. First it finds one end, then it travels to the other end and back again. The number of motor steps in each direction is counted, and if they differ by more than one percent the controller throws an error. This is to ensure that the whole thing is mechanically sound and that the sled isn’t binding up.

Once it’s calibrated, the sled should never hit either end switch. If it does, the motor will stop immediately.

At this point the stop button is lit and the display says “stopped”. Pressing reverse will of course start the slider moving in the “reverse” direction. When designing it, I first labeled everything “left” and “right”, but I realized that would be confusing if you were standing on the wrong side of the slider. Forward and reverse aren’t much better, I guess, and if I had to do it again I’d label everything right/left but implement a switch that programmatically changes left and right.

So as you just saw, the slider stops automatically when it gets close to the end. Of course you can stop it manually or change direction mid-motion.

The slider pot does pretty much what you’d expect. When it’s set all the way to the right it’s at maximum speed — all the way to the left and it’s at zero. It can be adjusted while the slider’s moving or before hand.

One difference is that if you adjust the speed when the slider is stopped it will update the display with the speed in inches per second. That allows you to match new moves with previous moves. Unfortunately calls to the LCD library are a bit heavy, and they create a noticeable jitter in the movement of the stepper motor. If I could do it all again I’d’ve used either a Raspberry Pi or two Arduinos: One to drive the interface and one to drive the motor.

The “lock” button next to the potentiometer will lock the speed setting so that it remains constant if the pot gets jostled.

I added the “bounce” feature in case I wanted to have the camera sweep the same subject repeatedly. In fact, it’s the way I reshot the case unpacking sequence. It’s also kinda cool. As you can probably guess from the name, it moves the camera between the two ends over and over again.

I guess the coolest parts of this project are the “record” and “play” buttons. When record is pressed and lit up, any change to the camera’s movement will be saved in memory along with a time stamp. Basically it records when forward, stop, or reverse are pressed, and records any changes in the value coming from the speed potentiometer. Since it’s based on time to millisecond precision, you can have camera moves match tightly to a planned sequence.

The record button toggles recording on and off, but hitting play will also stop recording and go directly to playback mode.

Once you have a sequence of movements recorded, pressing play first moves the slider to the same position it was in when the recording started. It then begins the set of movements.

Playback can be interrupted at any time by pressing the stop button.

The decal on the case may not be the best looking in the world, but it beats my original idea of just using a label maker. I designed it in Photoshop and printed it from my inkjet onto a full-page translucent Avery label. If you go this route, I’d recommend getting a white vinyl label — I thought it would be cool to have the grey of the case show through like where it says “s.co.tt” — perfect color match and all — but really it made the rest of the printed sections appear that much duller.

Another couple of things I would’ve done differently in hindsight:

I’d have put the power input on the interface box rather than the controller. After all, the slider is going to stay in place during a shot, but I might want to move the controller around when setting up.

I would’ve also used a lighter, more flexible cable than this 25-conductor behemoth.

I was kinda in a hurry to get this project started, and the project box for the controller was the closest one I could find to the size I needed, but obviously it’s bulkier than it needs to be. If you’re building one, hold out for a better box.

I also screwed up the size of the interface box: Originally I bought a stepper motor driver that would fit in there, but it was completely the wrong type. That’s why the microstep driver is piggybacked on to the outside. The connections from the motor and the micro switches to the interface box could also benefit from removable connectors: RJ-11s or RJ-45s would work great. Not using them was just sheer laziness on my part.

I couldn’t find a reasonably priced timing belt that was long enough and that shipped from the US. So I just stapled a couple together. It’s not as bad as it sounds — the staples have actually been holding up pretty well.

The stepper motor can be pretty loud, depending upon the speed. I had it wrapped in foam when I was shooting the opening of this video, but that’s not a good solution if you’re using it a lot — the motor will overheat pretty quickly.

And again, I should have implemented this with two Arduinos or a Pi. In order to have the motor actuate smoothly, I had to set the microstep driver to 3200 pulses per rotation, which is 16 times the motor’s spec of 200 steps per rotation. Asking the Arduino to process user IO as well as accelerate, move, and decelerate without jitter is a challenge. I had to make some compromises, like only updating the LCD when it’s stopped, and reducing the resolution of the speed control and button poll intervals.

Despite all that stuff I’d do differently, I still think this project came out pretty well. There are things I’m happy with, like the layout of the controls, the simplicity of the external cabling, and of course it’s overall functionality; It does what I wanted it to do from the start without too many compromises.

If you want to make your own version of my Motion Control Slider, you can find everything at s.co.tt/mc-slider like circuit diagrams for the controller and interface box, a parts list, the controller decal template, and the full assembly video.

Thanks for watching, I hope this gave you some good ideas to create your own motion control camera slider.

“An automatic garage door opener makes you feel like you’re working in a futuristic wonderworld”.  – Frank Ormand, Pretzel Magnate

With the purchase of Amanda’s new Volt, I had to get the garage ready for a permanent resident.  She had kept her old Saab in the driveway, so I pretty much had the run of the place until now.  Protection from the elements aside, the garage is just a logical place to stick an electric car whilst it’s charging.

And my father raised me on automatic garage door openers, so I figured installing one would be the right thing to do.

Here’s a before picture of the garage:

Nothing to laugh at, I suppose.  But the stupid door would just sit there unless you went up to it and moved it around manually.  Clearly an unacceptable situation.

But wait.  Did you see the problem in that picture?  I sure didn’t.  Not even after looking at the situation real closely, like in this picture:

As anyone would, I went to my local Home Depot, and purchased their top-of-the-line Chamberlain Whisper-Drive Whateverthehell for $250.  Like an ass, it was after I got home that I started to check clearances around the door.

Do you see that 2×4 running horizontally above the door?  In the picture, you can see that it’s shaved down on one corner.  The joist across from it is likewise shaved down, only on the corner facing the door, so you can’t see it.

Why were they shaved down?  Because that’s the only way the motherfreaking door will actually clear them.  The top of the door travels up in between those joists as it rises, by about 3 inches.

Meaning I couldn’t mount the damn guide track for the opener underneath those beams.

Fortunately after some head scratching — “Should I lower the track and get a low-clearance kit?  Get rid of the overhead storage completely?  Return the opener and let Amanda open her own damn door?” — I came up with a solution:

Use a Sawzall (yes, I have an actual Milwaukee) to cut the ever-living crap out of those joists. It took some temporary bracing and a bit of cursing, but I managed to get a nice channel cut out for the opener, without losing my all-important storage area.

The middle of it is hung from the ceiling, which you can see a bit better here:

The structure is hung by 1×4 maple (yes, I decided to go fancy) which is bolted to 90deg angle brackets, which are then lagged into the ceiling joists.  It’s pretty damn solid, fortunately.

Here’s one more picture from the door’s end of things, just for fun:

Between figuring out a plan, removing (and later re-wiring) the lights, restructuring the overhead storage, installing and configuring the opener, getting power to the opener, and about 4 trips to Home Depot, what should have been a simple project ballooned into a 9-day adventure of anger and confusion.  (OK, so 2 weekends and some evenings in between, and I was calm for most of it).

Fortunately the whole thing works just fine, and I managed to avoid any major blunders along the way (except assuming that this would be a simple project from the get-go, of course).

Oh, and the garage door is probably 50 years old and will fall apart under the stress of constant use, so I’ll probably be replacing that soon.  (Hooray?)

Sorry if you were expecting a how-to, but I hope this gave you some ideas if you’re stuck in a similar situation.  (Plus, I just didn’t think to take pictures during the project).

I’ve only recently begun doing my own maintainance and repairs on my Jeep XJ 2000. I’ve hit a few snags and problems along the way, and this page details a solution for the most frustrating I’ve encountered so far. I went to replace both my front rotors, and they were both well beyond stuck to the hub.

I’d been hammering, pushing, pulling, and spraying the old rotors on one fine Saturday for over 3 hours (4 hours if you count the time I took to go to the auto parts store and stop for a beer). Finally it was close to getting dark when one of my neighbors came along and advised that I should rent a rotor puller from a rental place nearby. The place was closed by then, so I told him I’d just kludge something up from what I’d had on hand. He laughed at me and went away.

Below is what I came up with, and it worked like a charm. I’m sure I’m not the first one to do this, but I was pretty pleased with the results, and so I share it with you.

Safety

Needless to say, I claim no responsibility for what happens if you attempt to use the information on this page. This could be dangerous, it could break your tools and/or your car, so emulate it at your own risk.

That having been said, I did this exact procedure for both my front rotors, and absolutely nothing went wrong. It worked great both times, and except for a couple of bent screws on my C-clamps, nothing was damaged. (They were cheap clamps anyway).

This will quite probably damage your rotor. I was completely replacing mine, so I didn’t care. If you are going to have the rotor turned, you might not want to do this, or at least put padding on the clamping surfaces and in front of the hub where the rotor will land when it pops off.

Also, many thanks to all the helpful people on jeepforum.com. I’ve hardly looked at my Haynes manual since I found that site.

What You Need

• 4 C-Clamps
• 1 Bottle jack (came stock w/my Jeep)
• 10 Ft (appx.) strong rope
• 1 Stuck rotor

A hammer and some Liquid Wrench type thing will probably help as well.

The Steps

Step 1 – Spray some penetrating lube/rust remover so it soaks in between the rotor and hub and let it sit for a while.

Step 2 – Attach the first two c-clamps to the area where there is no backer plate (where the caliper sits). Make sure they are gripping the rotor ONLY.

Step 3 – Attach the second two c-clamps such that each one is directly opposite one of the ones you’ve just put on. C-clamps should be spaced as evenly as possible, and tightened well. You may have to jam the second two clamps between the backer plate and the rotor. I didn’t have much trouble at this step, and the backer plate wasn’t damaged by bending it out of the way.

Step 4 – Zip tie the bottle jack to the top two clamps such that the head of the jack rests on the hub.

Step 5 – Close-up of the above. Note there’s a hole in the base of the jack that the zip ties are running thru.

Step 6 – Tie off the end of the rope to one clamp and run it over the base of the jack to the opposite clamp and tie off there. I did 2 turns of rope between each pair of c-clamps for added strength. Repeat for the other pair of clamps so that it looks like this. Make sure the rope is snug so that it holds the jack in place.Also, make sure the jack is as perpendicular as possible to the rotor so that it doesn’t pop off sideways!

Step 7 – With the rope snug, cut the zip ties. You don’t want them snapping off under tension. If the rope isn’t snug enough to hold the jack in place, crank it out a little bit.

Step 8 – Begin cranking out the jack until the ropes are very taut. Keep an eye out for rope fraying, clamp or jack slippage, or knots coming undone. Stand to the side of the jack/rotor, never stand in front of it! It can pop off with great force, and you don’t want it landing on your foot!

Step 9 – As you increase the tension a bit at a time, keep giving the hub/rotor shots of Liquid Wrench (or similar). Give the rotor a few taps from behind, it may pop off with ease. Don’t whack the rotor/jack/etc hard. You don’t want the jack coming loose and flying off under tension!

Step 10 – POP! The rotor flys off. Picture shows exactly where it landed. This one required a LOT of tension; in fact, the jack wouldn’t extend any more.

Step 11 – Closeup of the assembly. The jack and the clamps are undamaged. Again, this may damage your rotor severely, so either don’t do this, or put padding all around if you plan on having the rotor turned and reused.

Step 12 – The Result.