1. Aaron Gipp, Victor Salov, Udara Cabraal
Drum Volume Control
Aaron Gipp, Victor Salov, Udara Cabraal

2. Introduction Overpowering volume of drums in small venues
Methods of controlling volume already in place
Drum shields
Dampening Pads
Electronic Drums
Pitfalls of current methods

3. Objective Create more pleasant listening experience
Utilize Active Noise Cancellation for volume control
Attenuation of up to 6 dB in three directions
Minimal distortion between “dry” and “wet” signal

4. Scope Single 12” Tom instead of entire drum kit
Complexity of 3D spherical waves
Apply theory to one drum and extrapolate to full kit
Attenuation in single directions
Only one speaker for cancellation
“Nodal lines”
Cancellation at single point if too complex

5. Original Design Low-Pass and High-Pass Filters
Eliminate sounds not associated with drum
Bessel Filter design for exceptional linear phase
Eighth-order

6. Original Design
Images taken from:

7. Original Design Pre-Amplifier
Op-amp circuit for preliminary amplification
Vout = Vin(1+Rf/Rg)
Op-amp: TI LM741CN
Rf = kΩ
Rg = kΩ
Vout = Vin
Image taken from:

8. Original Design Inverter
Op-amp circuit used to invert polarity of signal
Resistor values nearly identical for unity gain
Theoretical 180° phase shift for all frequencies without time delay

9. Original Design Inverter Design Vout = Vin(-Rf/Rin) Op-amp: TI LM741CN
Rf = kΩ
Rin = kΩ
Vout = Vin(-0.99)
Image taken from:

10. Original Design Final Amplifier
Op-amp circuit used to control gain of inverted signal
Almost identical to pre-amplifier
Vout = Vin(1+Rf/Rg)
Op-amp: TI LM741CN
Rf = RV6NAYSD503A-P Clarostat 50kΩ single-turn ½ Watt potentiometer
Rg = 1.476kΩ
Vout ranges from 1.000Vin to Vin

11. Op Amp Golden Rules The op-amp has infinite open-loop gain.
The input impedance of the +/− inputs is infinite. (The inputs are ideal voltmeters).
The output impedance is zero. (The output is an ideal voltage source.)
No current flows into the +/− inputs of the op amp.
 In a circuit with negative feedback, the output of the op amp will try to adjust its output so that the voltage difference between the + and − inputs is zero (V+ = V−).

12. Original Design Additional Parts
Omnidirectional MIC48 Multimedia Computer (electret) Microphone
50Hz-16kHz frequency range
665-AS05008PR2R PUI Audio 2” 8ohm .4W speakers
550Hz-4.5kHz frequency range
Drum-Striking Apparatus
Consistency in measurements
Same height, same location on drum

13. Original Design Inherit drum signal with microphone Filter sound
Invert
Amplify
Project back at source for cancellation
“Dry” and “wet” signals propagate in opposite directions

14. Circuitry Testing Procedures
Take measurements of the input vs. output V.
Check for phase shifting on the inverter.
Vary input voltages and frequencies for inputs and check for gain.
Measure signal attenuation if it occurred.

15. Phase-Shift Testing
Used 39K resistors *2, 1.5K resistor *1

16. Tests for Original Design
Tested Pre-Amplifier for similar frequencies
Tested for unusual amplification or phase shift
1 kHz, 0.2Vpp
550 Hz, 0.2Vpp

17. Tests for Original Design
Tested Pre-Amplifier with Inverter for similar frequencies
Tested for unusual amplification or phase shift
1 kHz, 0.2Vpp
550 Hz, 0.2Vpp

18. Tests for Original Design
Tested final amplifier for similar frequencies
As potentiometer varies, we progress from unity gain to a gain of ~35
1 kHz, 0.28Vpp min gain
1 kHz, 0.28Vpp max gain

19. Tests for Original Design
Tested all three components for similar frequencies
Clipping for max gain from exceeding supply voltage
Unsmooth due possibly to high frequency
DC offset due to input offset voltage of op-amps
1 kHz, min gain
1 kHz, max gain

20. Drum Sound Experiment
Created using device for consistency.

21. Drum Sound Experiment Created using device for consistency.
Replicated multiple times to check consistency.

22. Drum Sound Experiment Created using device for consistency.
Replicated multiple times to check consistency.
Measured using microphone and oscilloscope.

23. Drum Sound Experiment Created using device for consistency.
Replicated multiple times to check consistency.
Measured using microphone and oscilloscope.
On average we saw this in Voltage vs. Time plots:

24. Fast Fourier Transform of Drum Signal

25. Successes
Inverter worked (phase shift of 180°) for all frequencies with minor amplification
No noticeable phase shift for pre-amplifer
Incorrect amplification in inverter/pre-amplifier compensated by final amplifier
Drum-striking apparatus able to produce consistent measurements

26. Challenges Unsmooth signals for three-component system
Phase shift for final amplifier at max gain and high frequencies
Ex: 1 kHz => 2°
5 kHz => 13°
10 kHz => 30°
Op-amps unable to source sufficient current to power speaker
Max voltage without distortion = 0.2V => 0.005W
Filters’ original purpose deemed unnecessary
Original microphone not reliable

27. New Design/Replacements
Shure SM57 and Shure SM48 dynamic microphones replace electret microphone
Increased reliability and sensitivity
Mixer to replace pre-amplifier
Less distortion (smoother signals) at high frequencies
Behringer Eurolive B215A (400W), Marshall MG30DFX (30W) powered amplifier/speaker combinations to replace old speakers
More “head room”
Broader frequency band
Lower distortion (1% vs 5%)

28. Tests for New Design
Verified that old speakers would not be able to support voltage above ~0.2V
Ch1 = input to final amplifier, Ch2 = input to speaker

29. Tests for New Design
Re-tested inverter circuit for new prominent frequencies of drum (up to 500Hz)
Later learned frequency content up to about 6kHz
162Hz, 4Vpp
233Hz, 15Vpp

30. Tests for New Design
Verified microphone could inherit signal and interpret frequency, and inverter could invert signal
100Hz
1kHz

31. Tests for New System Tested for phase response of system and air
Distance between microphone and speaker = 0.0m
Setup: FG => Marshall => SM57 => mixer
Ch1 of scope measured output of FG
Ch2 of scope measured output of mixer
(Phase difference)/(frequency difference) constant in linear phase system

32. Tests for New System D=0.0m t=time delay (approximately) in ms
frequency 1 2 3 4 5 6 7 8 9 10
Avg
Slp t
100Hz 78 81 77 80 79
.86
2.4
200Hz
163
161
167
164
166
165
.21
.57
300Hz
184
188
185
186
.31
400Hz
215
217
218
216
214
.17
.47
500Hz
234
232
235
233
231
.19
.54

33. Tests For New System Similar test for D=10cm
Phase delay due to air = Phase (D=10cm) – Phase (D=0cm)
Expected phase calculated for vsound in air = m/s
Frequency
Average
Phase delay due to air
Expected phase
100Hz
2.3
283.4
10.49
200Hz
150.3
345.5
20.98
300Hz
170.3
345
31.47
400Hz
232.9
16.8
41.96
500Hz
264.8
31.9
52.45

34. Tests for New System Similar test for D=20cm
Phase delay due to air = Phase (D=20cm) – Phase (D=0cm)
Frequency
Average
Phase delay due to air
Expected phase
100Hz
348.5
269.6
20.98
200Hz
165.1
0.3
41.96
300Hz
182.6
357.3
62.94
400Hz
262.5
46.4
83.92
500Hz
317.7
84.8
104.90

35. Tests for New System
Tested for cancellation of a single frequency by pointing two speakers at each other
Setup: FG => Marshall => SM57 => mixer => inverter => Behringer
Measured SPL with Sound Level Meter
Cancellation at certain points called “nodes”
Discovered this orientation would create standing waves
Pointed both signals in same direction from then on

36. Tests for New System
Took measurements of dry drum signal from ten feet away with Sound Level Meter
30 samples, average = 76.24, standard deviation = 4.23
Deemed SPL meter somewhat unreliable, took measurements with scope

37. Tests for New System
Based on system/air non-linear phase response, we can only cancel narrow band of frequencies
Re-sampled drum to select peak frequencies above 0dB (193.5, 196, 212, 230, 236 Hz)
Frequency
Average
Distance (m)
193.5Hz
328
1.615
196Hz
327.1
1.602
212Hz
348.8
1.556
230Hz
207.9
0.836
236Hz
283.1
1.14

38. Tests for New System
Varied distance between microphone and speaker between aforementioned values (1-2m separation) and tested with varying volumes
Voltage Signal and FFT plot of drum by itself; 8.83Vpp for 1s, 2V/div; 500Hz span, -25dBoffset, 5dB/div

39. D=1.4m with arbitrary amplitude gain
Tests for New Design
Original Signal
D=1.4m with arbitrary amplitude gain

40. New Design Successes
Most of frequency content above 500Hz far below 0dB, so cancelling narrow band might have worked
System can inherit sound, interpret frequency, and invert signal
Learned both “dry” and “wet” signals must travel in same direction

41. New Design Challenges
Non-linear phase response associated with electronics (different frequencies delayed by different amounts)
Air being a dispersive medium (different frequencies will travel faster than others)
No notable attenuation; peak-to-peak voltage always rose, average must be computed “by sight”

42. Recommendations/Future Work
Using DSP, create filter bank for all prominent frequencies
Measure system phase response for all prominent frequencies
Implement linear-phase filter for each frequency, delaying each to match frequency with longest phase delay
Compensate for system non-linearity and air dispersion