Project Logs

Collapse

• Lucky Number Four

  Thomas Countz • 04/28/2021 at 00:27 • 0 comments

  ---

  Here's a GIF of Hexo! This project is complete!
  

• Meet Hexo.

  Thomas Countz • 04/21/2021 at 19:33 • 0 comments

• PCB Delivered!

  Thomas Countz • 04/18/2021 at 16:11 • 0 comments

  
  I just received my order from PCBWay in 6 days!! They all look great!
  
  Next steps: assembly (although I'm waiting on flux...)
  

• Parts Received!

  Thomas Countz • 04/16/2021 at 00:50 • 0 comments

  I'm still waiting on the PCBs (ETA: April 21), but for now, I tested out some of the LEDs. I hoping the orange really pops against the black solder mask. 0805 is tinier than I thought so I made sure to order a new flux pen!
  
  

• Gerbers Submitted

  Thomas Countz • 04/11/2021 at 01:48 • 0 comments

  Today, I've touched up the design a bit and sent off the gerbers to PCBWay for review!

  This has been an exhausting amount of work so far, but I've learned A TON already. Without a lot of experience, it's hard to tell whether or not the board will work when it shows up, but I'm excited to find out. I'm thinking that at the very least I'll need to experiment with the resistor and capacitor values to get the LED timing and battery usage right.

  My order from Futurlec still hasn't shipped. I contacted them a week ago and they said they're still working on it... my plan was to have those parts here by now so that I could run experiments on powering this circuit with only 3 volts, but it'll have to wait for the first run of the PCBs—I wouldn't be surprised if they show up first.

  If all else fails, I hope that I end up with some cool PCB pendants and a heck ton of knowledge!
  

• Oops.

  Thomas Countz • 04/04/2021 at 19:45 • 0 comments

  After finishing the layout, I started doing some electrical checks and the first thing I noticed was that I wired in all of the LEDs backwards.
  

• PCB Edge Cut Calculations

  Thomas Countz • 04/04/2021 at 15:51 • 0 comments

  
  I originally intended on using Inkscape to design and export a DXF file to specify the edge cuts on the PCB in KiCAD. However, I had a bit of difficulty understanding how to use Inkscape for precise layout. Instead, because my design was a rather simple geometry made up from a series of triangles, I could calculate the coordinates for each point and use KiCAD's coordinate system directly. The strange thing in PCBNew is that the origin is a random point in the upper left, outside of the border of the design space, but I just used that origin (instead of calculating offsets) and then moved the final polygon into the layout space.
  

• Initial Parts Selection and Layout

  Thomas Countz • 04/03/2021 at 20:36 • 0 comments

  KiCAD Hype.mp4 from Thomas Countz on Vimeo.

  Time to start building footprints and layout in KiCAD. I came up with a seven sided design that I'll upload in the next log. The intention is for the circuit board to be worn as a pendant.

  I'm still waiting on the THT components to arrive from Futurlec so that I can prototype the capacitor and resistor values. Until then, I don't think I'll be ready to order the SMD parts, but because everything will be 0805s, I should be able to order the PCB sooner, rather than later.
  

• Simulations

  Thomas Countz • 03/28/2021 at 18:38 • 0 comments

  Falstad

  Run the Falstad simulation here: https://tinyurl.com/yhqfvbpn

  LTspice

  To run the simulation in LTspice, you'll need to download the CD4000.zip file under the project documents and install the model libraries and symbols before downloading and opening hexo.asc
  

  * /Documents/LTspice/Hexo.asc
  XU1 0 N003 N006 +V NC_01 N003 N003 +V NE555
  V1 +V 0 9
  R1 0 N003 4.7Meg
  C1 N003 0 100n IC=0
  R2 N003 N005 10K
  Q1 N005 N002 +V 0 BC557C
  R3 N002 +V 3.3Meg
  R4 N001 N002 10Meg
  S1 0 N001 N004 0 SW
  V2 N004 0 PULSE(0 1 0.5 0 0 0.1 0 1)
  XU2 0 0 N006 N007 N009 N010 N011 N012 N013 N014 N015 N016 N017 0 VDD 0 CD4017B VDD=3.7 SPEED=1.0 TRIPDT=5e-9
  D1 N007 N008 QTLP690C
  D2 N009 N008 QTLP690C
  D3 N010 N008 QTLP690C
  D4 N011 N008 QTLP690C
  D5 N012 N008 QTLP690C
  D6 N013 N008 QTLP690C
  D7 N014 N008 QTLP690C
  D8 N015 N008 QTLP690C
  D9 N016 N008 QTLP690C
  D10 N017 N008 QTLP690C
  R5 0 N008 4.7Meg
  C3 +V N001 1µ IC=0
  .model D D
  .lib /Library/Application Support/LTspice/lib/cmp/standard.dio
  .model NPN NPN
  .model PNP PNP
  .lib /Library/Application Support/LTspice/lib/cmp/standard.bjt
  .tran 0 15 0
  .model SW SW(Ron=1 Roff=1000G Vt=0.1 Vh=0)
  .include CD4000.lib
  .lib NE555.sub
  .backanno
  .end
  

• How the 555 Roulette Circuit Works

  Thomas Countz • 03/28/2021 at 17:54 • 0 comments

  You can find the PDF version of this schematic under the project's documents.
  
  When SW1 is closed, C1 quickly charges to VCC (I think there might need to be a current-limiting resistor between C1 and VCC, though maybe SW1 provides this). After C1 is charged, current begins to flow from VCC through both R3 and the Q1's emitter. Because Q1 base voltage is negative w.r.t. the emitter voltage (Q1 base-emitter voltage), the transistor is active forward and current flow through the collector. (It's a PNP).
  
  Let's continue as if SW1 is still closed. Now that current is flowing through Q1, the 555 begins to oscillate as if it were in astable mode. Charging and discharging C2 to 2/3 and 1/3 VCC respectively and setting pin 3 (output) high and low in kind.
  
   Pins 2 (trigger) and 6 (threshold) are tied together so that the 555 can continue to "retrigger" itself when C2 is discharged and the voltage reaches 1/3 VCC. In a typical astable configuration, a resistor is placed between pin 7 (discharge) and tied-together pins 6 (threshold) and 2 (trigger). This has the effect of adjusting the duty cycle by altering the charging speed of C2. In this circuit, the duty cycle is manipulated by altering the charging speed of C2 by way of a steadily reducing current through Q1 instead.
  
  (Note: I believe there should be a decoupling capacitor at pin 5 (control voltage)).
  
  Now that the 555 is oscillating, pin 3 (output) sends a square wave clock pulse to the 4017 pin 14 (clock). The 4017 is a decade timer and at each rising edge of the clock pulse it will consecutively set each of it's 10 output pins high. Only one pin is high at a time, the other are low. In the case of this circuit, we're only interested in using 6 of the 10  outputs, so we tie pin 5 (Q6), the seventh pin to be set high, to reset which will cause the 4017 to set it's first output (Q0) high instead.
  
  All of these outputs are used to drive LEDs which whose cathodes are connected to a common current limiting resistor before being tied to ground. These LEDs will continue to be turned on and off consecutively as long as a clock pulse continues to drive pin 14 (clock).
  
  Let's continue now as if SW1 is opened after being closed. Once SW1 is open, C1 begins to discharge and current continues to flow through VCC and GND by way of Q1. As C1 discharges, however, Q1 base-emitter voltage begins to climb, i.e. the base voltage starts to become less negative w.r.t. the emitter voltage. As this is happening, less current is flowing through Q1 and C2 is taking longer and longer to charge and therefore the clock pulses from pin 3 (output) of the 555 are beginning to slow (the time between clock pulses is increasing). This has the effect of diminishing the speed at which the 4017 is cycling through the LEDs.
  
  Eventually, Q1's base-emitter voltage reaches the cutoff threshold (~-0.6v) and current stops flowing through Q1. C1 continues to discharge, however, through R3. At this point, whichever LED the 4017 had set to high last, remains lit.
  
  I'll include a simulation in the next update so that you can see all of this in action with either LTspice or Falstad.
  
  
  

View all 11 project logs