Notes on Final Project

Below is a sketch of the spinning machine final project.

An extension of this idea is the binding machine, which wraps two or more core yarns with a lighter binder yarn at the desired rate of turns per unit length.

Table of components of Final project and their priority:

• 1 = Get it done this week. Essential to success of project.
• 2 = Would make project much better. I want to get this done.
• 3 = Would be nice.

Notes:

• Line 1. I have made some of these parts, I have the bearings and tubing to make the rotating parts. The DXF file. Challenges will include making a mechanically controlled friction device, and finding a nice way to couple motor and spindle.
• 
• Line 2. Flyer. Made a start at casting a flyer a few weeks ago. I will continue, perhaps with poyurethane mold to give more stiffness, and cast epoxy. This is the largest unknown among the priority 1 components.
• Line 3. Motor control. I have a DC brushless motor with a manufactured PWM control board. The speed is controlled by the supply voltage (3-10V). I can figure out how to set up a PWM to effect this control This is not the most desirable because the control board is not made by me.
• Line 4. Motor control. If I can find a workable brush DC motor, I will control it with an AVR and the 4953 H-bridge driver with programmed PWM. This is the best simple alternative, preferable to above.
• Line 5. Motor control. I really like these Hall sensor Brushless DC motors, and wonder if I can get a reasonably priced version somewhere. It would be fun and pretty straightforward (?) to program a driver.
• Line 6. Foot Switch. This wireless foot switch is needed to integrate input devices and communication into the final project. I'll do it.
• Line 7. The binding machine, pictured above, is actually more exciting than the spinner to the people that love making things with fiber. There are now more than one electronic spinning wheel on the market, but none that double as a binding machine. This would be an easy extension. I'll push to get this done by the final week.
• Line 8. Battery. The spinner could run for hours (see earlier calculation) on a laptop-sized battery. Could include a battery by:
  • Ripping one out of an old laptop and installing it together with charging electronics. I like this option.
  • Programming an AVR to charge the battery. This should not be too hard......(?).
• Line 9. Display. Another programming job. This would be nice for users, and should be in some version, but maybe not this one.
• Line 10. Carrying case. I really would like to do this. Modelled on a Dred's Earphone case that I saw recently, a zippered composite carrying case.
• 
• Line 11. Computer monitor. I don't really need a computer monitor of the spinner at this stage, but would like to show data aquisition and display from a DIY sensor using matplotlib. This would actually help to make the case that what I've learned might be useful in undergrad lab teaching.

Progress on Project, line by line. Week of Dec ~4 - ~10.

Line 1: Parts and mechanical design of spinner.

Dec. 2-4. Make acrylic parts and put spinner toghether with motors.

Dec. 5 Design friction mechanism and adjustment

• Below is a sketch of a lever arrangement. I did not do this, but made a friction adjustment using a more traditional string around the axle of the spool.
• 
• 
• Here are photos. The friction mechanism is hard to capture. Need to make some drawings for the documentation.
• 

Above I show the 12V brush DC motor from the inventory. Below is an alternate motor - a brushless Hall sensor DC motor. Interesting to compare.

Dec 6-10. Put together a working prototype, using an older flyer.

Line 2: Flyer

• The original flyer from a previous version. See above. This flyer could be statically balanced, but was dynamically unbalanced due to the arm extending only on one side. It may be desirable to make a symmetric (two-armed) flyer.
• The 3D-printed, coated version from casting/molding week
• . I may modify this to use on the machine. I do not think it makes sense to cast a version of this. Need to get a better idea of what works first, using a faster turn-around mtehod. I like the idea of a multilayer object.
• Design and planning for a layered structure, cut from 2D sheets:
  • cut on laser cutter form acrylic and perhaps sheets of composite material 9for more stiffness.
  • slider for yarn guiding, 3D printed.

• Creating sheets of composite material for stiff rotating beam:
  • Cut out sheets of fabric - four layers of burlap, and about 12 layers of cheescloth.
  • Saturate with epoxy. In this case, castable epoxy is what I had. Better would be the marine epoxy used in the standard composite layup.
  • 

Line 3-5: Motor Control

Program Tiny45 board used in the output devices week to drive the 12V motor from the Fab inventory with programmed PWM at ~ 1kHz.

• PWM duty cycle depends on input signal from potentiometer, measured with Analog in. The on-time is just the analog reading, in microseconds, while the off time is fixed at 500 microseconds.
• The code for the control.
• Direction is controlled by a digital input pin. The pullup resistor is activated and an external switch connects the pin to ground to bring it LOW.
• The ON/OFF state is controlled by another input pin. In the next step, another board will send this signal based on input from the footswitch.
• The schematic will be generated to reflect the additions of switches, in order to document the design.
• The board, very modified: with much hot glue!

Line 6: Footswitch

• Sensor using wool and stainless steel fibers, mushed together and felted with soap and hot water. I think this idea came from the Sparkfun website
• 
• this material can be used for a foot switch, perhaps encased in a hinged acrylic wedge. May be more predictable if I spin a composite yarn, knit/crochet, and then felt the fabric.
• Progress on an acoustic comm. link.
  • Driving: nice 25 kHz square wave pulse drives piezo disc directly from the pin. Here is the code, using the AVR libc commands, after the Neil Gershenfeld examples. Much stabler, and faster than doing the PWM software with the Arduino commands.
  • Detecting. I have a microphone on a board, and have acquired a signal from ambient sound, but have not yet made the attempt to:
    • See a signal deriving from the piezo at 25 kHz.
    • synchronize the ADC aquisitions to detect this signal. I will start with Neil's reflective sensor program. I'll need to be precise in the period of measuring - use the counter/timer?

Another version of the table, reflecting the state of the project Dec 10 and planning for the last week.

• 0=done, or at least done enough for now.
• 1 = Get it done this week. Essential to success of project.
• 2 = Would make project much better. I want to get this done.
• 3 = Would be nice.
• 4 = not happening for now.