PERCEPTUM TECHNOLOGIES - ensc.sfu.cawhitmore/courses/ensc305/projects/2008/7dem… · ENSC 440 - Project Presentation 1 of 30 ¾Music is a very subjective field ¾No single sound

PERCEPTUM TECHNOLOGIES

ENSC 440 - Project Presentation

~Author unknown

1



Project Presentation and Demo by

April 14th, 2008

Kyle Huffman - Director of QABen Shewan – Director of Hardware

Paul Carriere - CEODaniel Galeano – CTO

2 Of XX


ENSC 440 - Project Presentation3



Incorporate New TechnologyInspire New Music

Effects Pedal circa 1960s

Effects Pedal Today

4


ENSC 440 - Project Presentation1 of 30

Music is a very subjective fieldNo single sound effect can satisfy everyone

Amplifier

Guitar

Effects Implementation

Images used from: (Line6 Inc., 2008) (Sonicftp, 2005) (Dudes 411, 2004)

Controlling

Processing

RockIt addresses the control of effects, NOT the processing



Design an Intuitive control mechanism

Strive for maximum mobility and stage presence

Support existing systems and effects

Maintain sound quality



Develop our own analog effects

Develop our own DSP unit

Control existing analog effects

Control existing DSP units

Sensors

Sound effects



Two subsystems:•Sensor processor•Midi Controller

Two types of sensors:•Accelerometer •FSRs

User interfaceMIDI protocol

Images used from: (Active Robots, 2006) (Global B2B Network, 2009) (Secure Eserver, 2008) (Images Co., 2007) (Dimension Engineering, 2008)



Transparent to userTwo unique triggering methods

-Tilt Control-Bump trigger

Analog output with excellent accuracy

Analog and digital processingMatlab-based emulator platformRobust against sensor driftIn-situ tilt sensitivity control

9



Toggle Accelerometer State

Toggle Guitar Effect

FSR Housing in SolidWorks

10



PortableNon-intrusiveModular natureC based PIC microcontrollerSignal processing capabilities



Maps Sensor Processorsignals MIDI protocolModular designUser interface

12



http://shop.sinky.net/photos/200573111214835.jpg

13



R&D Project: Can we honestly compete against foot pedals?

Open Source Project-Great for product development-Judge market interest

How we profit: sell DIY development kits!

MIDIBox Portal: http://www.midibox.org/dokuwiki/

14



Cost Breakdown of System componentsTwo 18F PICs $10

Accelerometer $10

FSR: $6/sensor

Enclosures Optional - $30

Auxiliary Items: Cat-5 cable, connectors, resistors, etc.

$30

Single Sided PCB’s $20

Total Kit Costs $125

Assumptions: No rushed orders and purchases in bulk

15



No commercially comparable productsOther interesting research projects

B-Tar 3000Input Device and Music Interaction Lab: McGill Universityhttp://www.music.mcgill.ca/~benjamin/btar3000/

Mouse Pedal

16



Risky Venture: Community adoption

Great for electronic hobbyist who also play guitar

Requires more development before considering commercialization

Future??

17



Project costsGroup dynamicsTime managementDifficultiesWhat was learnt

18



Accelerometer development kits: x3 $120

FSR: Samples and final set: x6 $70

Rush order PICs: x2 $25

Enclosures, Connectors, Break-out board, Surface mount capacitors, sockets, etc.

$155

MIDI-to-USB Adaptor $45

Custom FSR Enclosure $30

Total Project Costs $445

Entirely member funded

19



Parallel system development-Two design groups-Excess components: PICs, breadboards, etc.

Every member became single component expertCentral Desktop: Project management wikiMatching PICs: reduce PIC problemsMilestones with tangible results as outcome:

-MIDI Monitor-ICD2 Debugger-Matlab emulator

20



FSR

-Finish testing & characterization: -Design PIC interface & circuitry:-Final assembly:-System test of FSR/PIC:

Accelerometer

-Finish testing & characterization:-ADC interface with PIC:-Finalize design:-System test of Accelerometer/PIC:

MIDI Controller

-Arrival of MIDIBox unit:-PC to POD interface:-Generate MIDI messages:-System test of MIDIBox to PODcommunications:

Sensor Processor

-Blinking LED:-UART interface:-System test of PIC & Sensors:-Test of Sensor Processor to MIDIController communications:

Feb 22Feb 29Mar 14Mar 25

Feb 27Feb 29Mar 14Mar 25

Feb 22Feb 29Mar 14

Mar 25

Mar 03Mar 10Mar 17

Mar 22

21



FSR

-Finish testing & characterization: -Design PIC interface & circuitry:-Final assembly:-System test of FSR/PIC:

Accelerometer

-Received accelerometers:-PIC ADC functional:-Finalize tilt and bump algorithms:-Finish MATLAB test bed:-Implement tilt algorithm in PIC:-System test of tilt with POD:-Finalize z-axis sensor:

MIDI Controller

-Arrival of PIC 18F2420:-PC to POD interface:-Generate MIDI messages:-System test of MIDIBox to PODcommunications:-User interface integration:

Sensor Processor

-Blinking LED:-UART interface:-System test of PIC & Sensors:-Test of Sensor Processor to MIDIController communications:

Feb 20Mar 12Apr 08

Mar 21

Feb 15Feb 28Mar 10Mar 26Mar 17Mar 28Apr 04

Mar 05Mar 16Mar 20

Apr 07

Mar 03Feb 25Mar 12

Mar 20Apr 04

22



Human interfaces are trickyPhysical packagingChoice of microcontrollerDebugging tools are essentialGroup organization should be flexible

23



Only 12 weeks to conceptualize, outline, and complete project

Narrowing design optionsMIDI options

24



RockIt is an Expansive ProductUser Interfaces Require Iterative TestingCapabilities to expand

Advanced components (PIC/Accelerometer)Wireless ComponentsEmbedded SensorsOpen-Source for Customized Use

25

26



Accelerometer ImplementationFSR ImplementationCommunication ProtocolMIDI Monitor

27



Digital & Analog filtering: fs = 120Hz-Analog RC low-Pass filter: fc=12-23Hz-Digital filter: Simple averaging technique

1.4Hz Low pass, 2-22Hz bandpass→8-9 bits of readout accuracy

Analog Devices AD XL330 accelerometer:Analog output, 3-Axis, ±3g of range

Robust against sensor drift: -Relative comparisons for tilt and shock: Common mode rejection

28



Not practical to monitor real-time ADC valuesSolution: Matlab EmulatorCreate scripts that emulate system components:

1-Gravity projection onto sensor axes

2-Acceleration-to-voltage transducer

3-Analog RC low-pass filter

4-Analog-to-Digital Converter: 10 bit accuracy with variable sampling frequency

5-Digital Filter

Algorithm implementation, analysis and debuggingall within Matlab

29



Tune Mixed Signal Filter Response in software

10-1

100

101

-20

-15

-10

-5

0

Z-Axis response with: fs = 120Hz, fc = 23Hz

A<d

B>

Frequency<Hz>

30 sample average20 sample average10 Sample Average5 Sample Average

Roll-off due toDigital Hi-pass Filter

Roll-off due to analoglow-pass filter

30



De-bug angle detectionEnsure correct values and mappingsAnimation

2 3 4 5 6 7 81

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2

Time: <secs>

Acc

eler

omet

er V

olta

ge<V

>

Analog Signal after LPF

x-Analogy-AnalogADC->DAC XADC->DAC Y

2 3 4 5 6 7 8350

400

450

500

550

600

Time: <secs>

Dig

ital v

alue

s

Digital Values

x-digitaly-digitalx-averagey-average

2 3 4 5 6 7 8-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

Time: <secs>

θ of

g w

rt X

<ra

ds>

Tilt Acc

theta Analogtheta Digital

31



Combination of experiment and Matlab to identify thresholds, sampling frequency and MIDI mappings

0 0.5 1 1.5 2 2.5 3 3.5 4 4.50.5

1

1.5

2

2.5

3Time Domain Signal

Time<secs>

Vol

tage

Oscilloscope Data

0 5 10 15 20 25 30 35 40 45 500

0.05

0.1

0.15

0.2

|Y(f)

|

Frequency(Hz)

Filter versus Experimental Data

0 5 10 15 20 25 30 35 40 45 50-9

-6

-3

0

3

Filte

r: A

<dB

>

FFT with 220 points

Mixed Signal Filter10 Samples, fc = 23Hz

Time and Frequency Analysis

Random Movement

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.91

1.5

2

2.5Z-Axis Data: Data versus Average:

Time: <s>

Vol

tage

<V>

Scope DataAveraged Data: 20 samplesSampled Data

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9-1

-0.5

0

0.5

1

Vol

ts

Time <s>

MIDI versus Acceleration

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

152535455565758595105115

MID

I Sta

te

Processed data MIDI Output

Intended Triggering

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1.5

2

2.5Z-Axis Data: Data versus Average:

Time: <s>

Vol

tage

<V>

Scope DataAveraged Data: 20 samplesSampled Data

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9-1

-0.5

0

0.5

1

Vol

ts

Time <s>

MIDI versus Acceleration

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

152535455565758595105115

MID

I Sta

te

Processed data MIDI Output

32



Large initial investment of timeEasy to port code from Matlab to PICDifferentiate between software and hardware issuesQuick turn-around on algorithm improvementsCreates a template for further work

33



Expected BehaviourImage Used: Interlink Electronics FSR Datasheet

Resistance Range: 10KΩ to 100KΩStandoff Resistance: >1MΩTolerance: 15-25%Active Area: 12.7mmNominal Thickness: 0.46mm

34


ENSC 440 - Project Presentation35

Top View

Profile



Variable Pressure MeasuringAnalog DSP controlsImplement Current-to-

Voltage Converter

Image Used: Interlink Electronics FSR Datasheet

36



One way communication between PIC to PIC

Four types of data:

•Accelerometer

•FSRs

•Bumping

•Status messages

User-defined tags map the data received from the sensors processor into corresponding MIDI commands.

FSR triggered

Sensor processor sends a FSR status message to the MIDI controller

Message is mapped into the corresponding MIDI command

Users define tags to map sensors

with effects

Effects unit receives Midi command and triggers

effect

Sens

or P

roce

ssor

MID

I Con

trol

ler



Logic Express software PC to PODxt communicationsSanity check of processing chain

http://www.fruity-loops.com/images/store/small/EM_MA-9900-40717-00.gif

38



Connection time-out: soft-shutdownActivity LED: transmission, FSR active, connectionTilt sensitivity controlPreview mode for MIDI controller

39

40



TechnicalPatrick LeungSteve WhitmoreJason LeeBrad OldhamAlbert LeungFred HeepCarlo MenonGary Houghton

MusiciansYannick ChampagneMatt PelkeSol DominelliTim Parent

OtherEveryone in ENSC440 lending us tools

41



“boudist”, http://www.boudist.com/archive/2006/06/14/come_together_preview.php

“Anwers.com”, http://www.answers.com/topic/vox?cat=entertainment

“Vox”, http://www.voxamps.co.uk/pedals/v847.asp

“Dunlop/MXR”, http://www.myhrbraaten.no/index.php?cPath=142&osCsid=mrd5qnramt4hkq0s293u8uqfk0

“JamBase”, http://www.jambase.com/Articles/Story.aspx?StoryID=11098

“Heil Sound” http://www.heilsound.com/pro/gallery/

“Virtual DJ Software”, http://www.virtualdj.com/blog/xeon/?category=0

http://140.115.95.15/wenchi/freshman/isbell/isbell1.htm

“Alexplorer’s Axe Hacks”, http://alexplorer.net/guitar/text/guitarists.html

“SenSyr LLC”, http://www.sensyr.com/store/index.php?act=viewProd&productId=10

42

http://www.boudist.com/archive/2006/06/14/come_together_preview.php

http://www.answers.com/topic/vox?cat=entertainment

http://www.voxamps.co.uk/pedals/v847.asp

http://www.myhrbraaten.no/index.php?cPath=142&osCsid=mrd5qnramt4hkq0s293u8uqfk0

http://www.jambase.com/Articles/Story.aspx?StoryID=11098

http://www.heilsound.com/pro/gallery/

http://www.virtualdj.com/blog/xeon/?category=0

http://140.115.95.15/wenchi/freshman/isbell/isbell1.htm

http://alexplorer.net/guitar/text/guitarists.html

http://www.sensyr.com/store/index.php?act=viewProd&productId=10



Documents

PERCEPTUM TECHNOLOGIES - ensc.sfu.cawhitmore/courses/ensc305/projects/2008/7dem… · ENSC 440 - Project Presentation 1 of 30 ¾Music is a very subjective field ¾No single sound