Sunday, March 23, 2014

3D Printed Tape Measure

Going off the success of my 3D printed dial calipers, I decided to try to print something even more elaborate. But what to print? I contemplated several options but ultimately decided to print a tape measure.

Originally I didn't think a tape measure would be that interesting... I mean, it doesn't even have gears. Once I started piecing it together in my mind and determining the acceptable "cool factor", I realized that the parts count alone was skyrocketing. My calipers had 9 pieces, this tape measure would have well over 100... Now things were getting interesting.

I decided to attempt this based on the parts count and the fact, that if successful, I would be able pull out over 4ft of tape from something about 3" sq. Also, I had no better ideas at the time.

I designed the tricky parts first and printed little test pieces here and there to validate the design. Right around the time I starting adding all the cutouts in the main body (for cleaning purposes), it started to look pretty cool.

The entire tape measure would be printed already assembled, wound up (pictured above), with support material. The support material is necessary to support all the complex geometry and separate pieces. Support material will be in every void not already occupied with the printing plastic (UV hardened resin, actually). This support material needs to be cleaned out after printing, which is what all the cutouts are for. For me, not being artistic in any way, adding lots and lots of cutouts in a way that didn't look stupid was the hardest part of the design.

Now for some tech specs! There are 114 parts in this tape measure, 52" of tape links labeled every 1" with graduations down to 1/8", a lock for the tape, a belt clip, and a fold-out crank with free-spinning handle to reel the tape back in. Measurement accuracy over the 52" is about 1/16". The belt clip, lock, and crank all have flexible features on which stress analysis was preformed to be sure they wouldn't break. The printed material is fairly brittle which means if it breaks it does so like glass. Unless the stresses in the material are low and are distributed properly it will break.

The crank has an over center cam-like feature and hard snaps in both the open and closed positions to make it bias towards those. The tape lock knob has a detent (visible on the backside) to make it lock in the unlocked position so it doesn't rattle around and lock up the tape when not in use.

I would really have liked to print a one-piece flexible tape or even a spring retractable tape but due to limitations in printing technology and material I couldn't do that. I was also resistance to turning this into a giant research project. Springs are one of the things that are difficult to print, especially in a brittle material. Springs with a preload are, as far as I know, impossible to print.

In this case, if I were to print a retraction spring, it would have to be a very large spring that would have to work over 10+ rotations and not be susceptible to material creep. Not to mention it would have to provide enough force over those 10+ rotations to reel the tape back in. That is not an easy problem to solve.

The tape itself can also be considered a spring; assuming its the same curved cross-section profile as a normal steel tape. If the tape was printed outside of the tape measure body, out of a flexible material, and that material wasn't very creep sensitive it would be possible to print a one piece tape that was rigid when extended outside the tape measure body. The printer required to do this would have to be very large though since the tape would have to be fully extended during printing.

The other, simpler, option is to just print a simple ribbon of tape out of a flexible material. This tape would not be rigid but it would be one piece and could be printed coiled up inside the tape measure body. The down side to this method is that it requires the use of 3 different materials during printing: rigid modeling material, flexible modeling material, and support material. This is currently possible but all but a few printers in the world cannot do this. Moreover, my preference is to keep my designs uni-material (not including support material) so they can be printed on a wider range of printers. 

All that being said, I still think it would be cool to further investigate some kind of spring retraction option.

For the curious, this tape measure was printed on an Objet Eden 3D printer.

And lastly: Yes. This is in Imperial units. Why? Because this is America. We are simultaneously innovative and stubborn. Mostly though, its what I am used to. Although on the small scales I greatly prefer SI units. ... Ok, I just looked up Imperial units and came across some of the lesser known ones... All I can say is wow...

3D Printed Dial Calipers

3D printing initially interested me because of its ability to create physical parts very quickly with nearly any geometry. By the time I had access to a 3D printer the ability to print virtually any shape had already been well proven and had even become common place. I was then introduced to the idea that multiple parts could be printed together, assembled, and captured. This may seem like a new concept but it is merely a new way of looking at 3D printing. The printer doesn't care how many pieces its printing, or even if they are connected.

I had seen adjustable wrenches printed already assembled. In the same fashion, I designed a c-clamp to try my hand at this concept. The camp worked perfectly. So then the question became "What's next?"

Dial Calipers. Yes. That sounded more than complicated enough with its gears, dials, and half dozen moving parts. I guess the irony of 3D printing a precision measurement tool with, what is normally considered, an imprecise manufacturing method was just too good to resist. But just to clear, 3D printing accuracy can range from +/-0.001" to +/-0.015 or more depending on the printer. Its typically not as bad as most people think... But it can be.

This dial caliper was designed to open to 4" with 9 different moving pieces that would all be printed together, already assembled. It would have a working thumb wheel, inside/outside jaws and a depth gauge. I also included a cam lock for the dial so the 'zero' position could be adjusted. Since this would be 3D printed I didn't see any reason to put the graduations below 0.005". In hindsight, graduations down to 0.0025" wouldn't have been completely pointless.

While cleaning away the support material I had my doubts as to if this would even work. It took a bit of work to free up all the various moving pieces but they all came free.

Fresh from cleaning, I went to show the calipers to one of my friends. He looked at me with half confusion and half pity with a look that said "...why did you do this?.." Then we had a little conversion about how the calipers would be horrifically inaccurate which ended with him handing me a random piece of steel rod stock he had been machining. He asked me to measure it. I zeroed the calipers, measured the diameter, and read off the number: 1.997". He then measured it with a proper set of Starrett digital calipers: 1.9975". He just looked at me and said "How did you do that?" Admittedly, I was a bit lucky there. Further testing showed accuracies in the +/-0.0025" were more common.

I want to be clear that this is just a proof of concept and is not in anyway intended as some functional production model. There are gaps between all the moving parts on the order of 0.01". The dial has 0.05" of backlash due to slop in the gear train. The plastic (actually, resin) that its printed out of is wildly temperature sensitive.

Accepting these facts, its a pretty cool device. Everything works as intended, its no bigger than normal calipers, and it looks cool. Every time I see this I'm reminded how steampunk it looks. Maybe someday I'll make a version out of stainless steel and brass.

STL files on Thingiverse.

Telepresence Robot

One night I was watching an episode of The Big Bang Theory where Sheldon made a telepresence robot to avoid physical interaction with people. As I watched that scene I thought to myself "I could do that." So I did.

My main motivation in building this was to use it to telepresence with my family back in the north east. I figured this would be a lot more fun and interesting than just the usual Skype session. 

The design is pretty simple: a motorized base with a laptop up top and a webcam. The base has two geared brushed motors which are controlled by the laptop via USB. I used a Maestro from which lets you control standard r/c equipment from a PC. I then wrote a little C# windows app which connects to the Maestro and then could control the motors. This app also could connect to another instance of itself over the internet in either a master or slave mode which would allow external control.

The webcam was mounted on a simple gimbal which allowed pan/tilt control for general looking around. All the video and audio communications were done through Skype.

The coolest part of this project was getting to play with the robot controls and creating new operating modes, different controls methods for user interaction and etc. Piloting a robot from 3000 miles away was also pretty cool, although internet latency really sucks.

I almost put a pair of nerf guns on this to add a little more fun to the mix. If I ever do another telepresence robot I'll have to add some offensive capability to it.

4ch Plane with Retactable Landing Gear

This is the more mechanically complicated plane I've ever built. It has a wing span of 9.5", weighs 4.25g, has full flying rudder and elevator, and retractable landing gear.

Flight times on this airplane are in the 7-10 minute range with a 30mAh single cell lithium polymer battery. The frame is all made from carbon fiber rod or tube in the 0.5mm - 1.5mm range. The covering is OS film and is 0.5-1 micron thick.

The propulsion system uses a 4mm 10ohm motor, geared down 6:1, driving a 3.25" carbon fiber propeller. The servos that drive the rudder and elevator are each 400mg and are my own custom design. Oh, and the tail wheel is steerable.

Ground handling on this plane is amazing and arguably the most fly part of not-flying it. Landing gear is very very uncommon on micro planes. So being able to do things like take offs, landings, and touch & goes just makes this plane too cool. When it does get in the air it flies very slow. Although, it has enough thrust to get going pretty fast or climb in altitude. The thrust:weight ratio is over 75% as I recall.

The landing gear retracts mechanism also uses a custom servo of my own design. It is designed to retract the landing gear parallel to the fuselage and deploy them at an angle. The entire mechanism as show below weighs around 0.25g.

Here is a video showing a close up of the airplane with all the moving bits in action. The next video shows it in flight.

Micro Servos

When I started to take a break from microscopic flying vehicles in 2010 and migrate towards some larger, easier to work with, models (5-20g) I realized that magnetic actuators just weren't going to cut it for control (rudder, elevator, ailerons). Magnetic actuators are heavy and draw power continuously when deflected away from neutral. I needed servos. Servos would provide lots of pulling force for their weight but at the cost of mechanical complexity. Since servos at the weight I needed didn't exist or were out of my price range I decided to further over complicate the situation and build my own.

I started by designing a simple linear servo (lead screw design). The picture above is of my 400mg variant which has a throw of about 0.2", resolution of 0.0015", and can pull about a 25g load. The motor is a brushed 3.2mm diameter coreless design by Shicoh. The gears used are module 0.15 and can be gotten from The threaded rod is 0000-160, which means it has 160 threads per inch (TPI). Threaded rods in this size (and 120, 90, etc TPI) can be purchased in 1ft lengths from

All the control software is done on a PIC microprocessor and the feedback is done magnetically with a hall effect sensor. Here is a video of my first version which is a little heavier and has a bit longer throw; 450mg and 3/8" respectively.

Naturally, once I had this servo design done and working I had to see if I could make a smaller and lighter one. After some thought I figured that if I took apart the 3.2mm motor, threw away the case, and rebuild it as a stepper motor I could cut the servo weight almost in half. I also compacted the design to further reduce the weight. Here you can see all the parts for the servo nicely arranged on a penny.

This servo functions very similarly to the large one described above. The only difference is that I wired the brushed motor as a stepper motor by soldering directly to the 5 contacts on the motor's commutator. This allowed me to get rid of the steel case and use the commutator as the main structural component (backbone) of the servo.

This servo weighs just 190mg, has a throw of 1/8", and a resolution of 0.003". It can pull with a force of about 4g. Overall length is about 3/8". Here is a video of it in action:

These two servos can run at voltages between 3-5v and are controlled with a 1-2ms pulse as is standard for all hobby r/c servos.

Sub Gram Ornithopter

I can't help myself. Every time I build something I invariably think of ways I can build it smaller. I built a large ornithopter, now I have to build a small one.

This ornithopter comes in at 920mg, just under 1g, and has a 2.5" wing span. It has 4 flapping wings. The two wings on each side flap towards each other and then away from each other. This produces both thrust and lift. 

It is powered by a 8.5mAh single cell lithium polymer battery. Flight times are around 3 minutes per charge. Controls are done through an infrared receiver system which can command the throttle and rudder. The rudder is actuated with a small magnetic actuator.

The motor is a single phase brushless design that is geared 10:1 to flap the wings. The frame is all carbon fiber rod between 0.01" (0.25mm) to 0.02" (0.5mm) in diameter. The covering on the tail is OS Film while the cover on the wings is some thin grocery bag plastic film.

This is very weird to watch fly. The wings are a blur and the whole vehicle looks like its being propelled by a large dust ball. It flies quite slow and looks very much like a large insect. I wouldn't dare fly this outside.. I'm positive a bird would make a meal out of it.

VTOL Ornithopter

Now this is a bit of a terminology cluster... VTOL stands for Vertical Take Off and Landing; and an Ornithopter is any flapping winged thing that flies (aka, birds & insects).

This hot number was pieced together from an Air Hogs Avenger back in October of '08. The Air Hogs Avenger was a nice little 4-winged ornithopter with a little too much thrust-to-weight to remain unmodified. I gutted the system and used the wings, gearbox, and motor to make my own, cooler, ornithopter.

I made the airframe out of carbon rod and added elevator control. I also had to replace the radio system to support throttle/rudder/elevator control. Total weight, with battery, is 10.5g and the flight times are over 10 minutes.

This vehicle has the ability to take off and land on its tail, almost hover in place, cruise in level flight, and do loops. It is very responsive and aerodynamically quirky (not in a bad way). Flight stability is great, it can be flown with only rudder control by a novice or flown mildly acrobatically by someone with a bit more skill. Its a really great, weird, vehicle that I have been flying since I built it. I really enjoy cruising this around.

The elevator and rudder control were originally done with magnetic rotary actuators but I have since replaced them each with 400mg linear servos of my own design. 

390mg Plane - My Lighest Plane & Unoffical Record Holder

To date, this is the lightest plane I've ever built and it held the unofficial record of the world's lightest plane for about 4 months back in July of '08.

This plane is my crown jewel of micro plane accomplishments... despite the mountain of headache it took to build. It has a 3.1" wing span, a 1" chord, and is 3.5" long. Flight times were surprisingly high, around 4-5 minutes, considering the size of the motor. An infrared transmitter/receiver system was used to control the throttle and rudder. The rudder control is done though an electromagnetic actuator.

The frame is made of carbon fiber rods ranging in size from 0.01" (0.25mm) to 0.005" in diameter. All the rods less than 0.01" diameter were hand sanded down to size. The covering on all the surfaces is OS film (DuPont).

All the electronics in the IR receiver - microprocessor, FET, and IR detector - were sanded down to reduce weight. The aluminum that encases the 8.5mAh lithium polymer battery was trimmed away to get the battery weight lower. Even then, the battery weight was 75% of the plane's weight.

The motor and propeller were custom designed to power this plane. The motor is a single phase brushless design swinging a 0.6" prop. A video of it can he seen here.

Weight Breakdown:
Airframe + Act: 37mg
8.5mAh Li-Po + On/Off Switch: 290mg
IR RX: 25mg
BL Motor: 30mg
Prop: 5mg
Landing skids: 3mg

All up Weight: 390mg

Original Post: 390mg Plane on RcGroups

560mg Plane - Unoffical Record Holder

During the time when I was really into building these small planes ('06 - '10) there was a little unofficial competition of sorts going on between the 5 or so people in the world that were building planes this light. Contrary to what one might think, building a small plane isn't as hard as building a light plane. So, there was always a little friendly competition to see who could build a lighter plane. This plane, for about 3 months back in November of '07, had that record.

The wing span on this plane was 2.75", the chord 7/8", and the length was about 3.25". The air frame is all balsa with OS film covering. Flight times were about 4 minutes. The controls were done through an infrared receiver which controlled the throttle and rudder.

Planes this light become a lesson in frustration. For example, all the "large" components of the receiver (microprocessor, IR detector) were sanded down to the copper die to reduce weight. The 3.2mm diameter brushed motor was completely disassembled and rebuild without the steel can to save even more weight (and thus killing motor efficiency). Propeller design becomes very critical at this point because there isn't a lot of motor power to work with at this point.

WS: 2.75"
Actuator: 3mm ID, 160ohms
Motor: Lightened 3.2mm Shicoh

Component Weights
Airframe + Actuator: 45mg
IR RX: 50mg
Lightened 3.2mm Motor: 130mg
Prop: 20mg
8.5mAh Li-Po: 295mg
Misc: 20mg

Total: 560mg

Original Post: 560mg Plane @ RcGroups

840mg Biplane

I built this back in May of '07. With this plane I wanted to focus primarily on small size. I was still going for something lighter than my previous plane but I wanted to make an airplane that even I thought was small - a tall order considering I had been building micro r/c planes for year or two at this point.

I chose a biplane design for this plane because it would allow me to keep a high wing area but keep the wing span small. The wing span on this plane is 2.75", the chord is 7/8", and the length is 3.75". The frame is all made of balsa wood and covered with OS film. Flight times were in the 4-5 minute range.

It is powered by a 3.2mm diameter brushed motor and a 1" propeller. The infrared receiver is capable of controlling the throttle and the rudder. The rudder is driven with a small electromagnetic actuator. The power source was a single cell 10mAh lithium polymer battery.

Weight Breakdown:
Airframe + Actuator: 75mg
IR Receiver: 65mg
3.2mm Motor: 275mg
Prop: 10mg
10mah Li-Po Cell: 395mg
Misc: 20mg

Total: 840mg

About a year or so after I built this I replaced the balsa paddle prop with a slick carbon fiber one that one of my friends designed. Not only did that prop increase the flight time but it made the plane basically silent. When it got within 5-10 feet of you it sounded like a mosquito about 1ft from your ear.

Original Post: 840mg Biplane @ RcGroups

My First Sub 1g Airplane - 0.98g

To start off my blog I would like to start by going over my first chain of wildly ambitious engineering projects - micro r/c planes...

This was the first r/c plane that I built that weighed in at under 1 gram. I built this back in April of '07. It has a 5" wing span, 2" chord, and its about 6" long. The flight times were in the 4-5 minute range on a full charge.

It was constructed of carbon fiber rod from 0.5mm to 0.25mm in diameter. The covering on the surfaces is called OS film and is made by DuPont. It is on the order of 1 to 0.5 microns thick.

The electronics consisted of a 10mAh Lithium Polymer battery (1S), a 3.2mm diameter brushed motor driving a 1" diameter propeller, an electromagnetic actuator to control the rudder, and an infrared receiver which was capable of throttle and rudder control. Here is a full breakdown of component weights:

Weight Breakdown:
Airframe + Actuator: 150mg
10mah Li-Poly Battery: 400mg
IR RX: 120mg (76mg sensor, 44 mg receiver)
Stock 3.2mm Motor: 270mg
Prop: 10mg
Glue + Wires + Misc: 30mg

Total Weight: 0.98g

I went on to make some power plant improvements to this plane later on. I replaced the battery with a lighter variant and replace the brushed motor with higher power direct-drive and geared brushless versions. This brought the weight down, increased flight time, and increased the power:weight ratio. I also replaced the IR RX with a newer, smaller, lighter version. All this brought the weight down to 670mg. The wing loading was low enough where I could fly over my head and the plane would gain altitude due to the heated air rising from around my body. 

Original Post: 0.98g Plane @ RcGroups