Bachelor Thesis Tex-files

Timeline

Date
7.6.	Start
14.6.	Table of Contents
9.8.	last possibility for "first draft"
30.8.	Deadline
6.9.	Kolloquium

Content

• I2C

  • v2
    • Umbau
      • packed union based data structures (removes the need for magic numbers
      • (safe types inside pi2cproto.h)
    • ballast
    • reused
      • remaining compatability to v1
    • new
      • UUID readout
      • r/w-ablility for both 1.8V potentiometers
  • raw memory mapping
    • why is mon48v = 0x11?
      • missing __attribute__((packed)) declarations
    • why are digipoti and anapoti only readable from 0x02 and 0x04 not partial?
      • same as above
  • instability on mass communication
    • timebased / overflow ? -> most likely race-condition on writing any value to (semi-)global objects
    • gdb-server feasable? -> NO: only 6 registers for breakpoints and no stacktrace

• Calibration

  • characterize adcbehaviour
    • plot resulting fitted incline of $\Delta$V vs. num. of readcycles
    • see if v_pin same as v_keith_pin
      • on dV of small enough -> adc circuit working correctly -> calib
    • reasoning for using a given sampleTicks value
      • smallest gain error -> inconsistent
      • smallest offset -> inconsistent
      • hightst samplerate -> sampleTicks = 1
      • pin plot -> sampleTicks = 0 is no go
      • pin plot -> sampleTicks in 1-7
  • polynomials
    • polynomials rw-able
  • 10I calc
  • differences between PIts?
    • if yes then we need a PITDB -> pitdb.yaml
    • standard only for default
  • Voltage
    • redo after selection of sampleTicks value
      • 48V
      • 10V
      • 1.8V ana
      • 1.8V digi
    • calibrate based on fit
      • 48V
      • 10V
      • 1.8V ana
      • 1.8V digi
  • Current
    • without calib
      • 48I
      • 1.8I ana
      • 1.8I digi
    • with calib
      • 48I
      • 1.8I ana
      • 1.8I digi
  • Temp
    • accuracy?
    • position in system
      • inside chip -> on underside of board, between frame and board
    • relevancy
      • peace of mind if readout correct and not too high (no experience)

• 1.8V Regulation

  • compare regulation concepts

    • there's overshoot: less oscillation is good, as long as we don't take forever to achieve those
    • stabilize

  • regulate the system (with missing information)

    • (possibility: unused reticles on each wafer can be used to get some feedback on initial calibration, would result into additional dynamic values for calculation and calibration o fthe 1.8V regulation)
    • write / formulate approach and resulting formula / algorithm

Table of Contents

• Introduction
  • What is this about?
  • What's the PowerIt Subsystem?
  • Goals
    • Firmware
    • Calibration
    • Regulation
• Theory
  • Hardware Component Behavior
  • ADC Calibration
  • 1.8V Output Regulation
• Experiments
  • Setup

    • Hardware

    • Softwarestack

      graph LR;
          fw[PIt Firmware] --- pot[Potentiometers];
          fw --- srv["PItStop Server (RPi)"];
          srv -.- agg[PItStop Aggregator];
      

  • Characterization
    • Voltages
    • Currents
  • after Calibration
    • Voltages
    • Currents
  • 1.8V Regulation
    • Characterization of Dropoff
    • after Correction
• Results
  • Firmware
    • I$^2$C mapped register- Table
  • Calibration
    • Calibration-Table
    • "How to calibrate a PowerIt"
  • Regulation
• Outlook

Notes

1V8

RS = 2* * 4.7k RP = 75k

9V6

R1 = 30k R2 = 10k