Bass Guitar preamp design - baltika group press center
Contents
1. soo IILI 4 JP j T jy j jJ LI EL ELI Jc sede 452 2152525522525225 2545 255452222222555544444 II j 0 LI TL j 1 1 j j j Il eee l DX I 1 4 pr BIET 4 14 24 4 4 411 1 P THA 06 DUE Il2. 15 5 BOOSt CUT TOD OLOGY 16 pruden hb A uu un __ 6 17 2 OV OW u us u s 17 yu u uuu s 17 Audiodomain blogspot co uk 2 homewith dave daveisnot outlook com 501 RTT Tc 18 Po SOIC OWEN Fall 18 OC OOM 2 _____ 18 2 ______ _________ 4___ 19 2 5 1 Operational Ampliier 19 19 UEM MO o IE u ul 19 2 7 Design CoODSIOSFALIOLIS ___6 _6 pde 19 ERR 20 IE PECA EIE QR 20 Rec sini P 20 3 BIKO Sr ETT 20 u 22 Modular DE __4_ _6_ 6 ____ __ 22 2 ale oL ________ __ 22 4 3 Issues During Construction 22 24 25 5 1 L 0 S 25 5 2 Frequency 5 5 DUUM P _ __ ___ ____ _ 26 9 2 uuu uu u uuu u OO TT 28 5 3 doBaridpass Frequency RESPONS
3. 69 26 Impedance Calculations aba cose pepe bu st AE beb QUU v st doa Qua 69 _ __ O 69 PISA ME 70 27 Current Draw Measurement 70 28 TLO7X Electrical Characteristics 70 29 Maximum Gain Readings LLL LLL Ika Uh suse a mS TE 71 30 Maximum Gain Calculations 71 30 Tey Maximum leve 71 30 2 9v Maximum Line Level 22 0000000 0 01 em e eene nennen nenne 72 20 9 u 2 30 4 9v Maximum Instrument 72 2305s 2 95 u 73 31 Nyguis t WOON RERO T 73 32 Metric Halo Mobile 2882 Specifications 73 Audiodomain blogspot co uk 6 homewith dave daveisnot outlook com 1 1 1 HISTORY The first occurrence of an external bass guitar preamplifier came in the form of passive direct inject DI boxes
4. jJ II j MALE TTT TT LLL LE LE I I jj 1111 1 IILI I LL LL j 1 l T iii ii LLL ETE DD SE TE ES EE LE E tt ttt tt E Fa sa z B SR ES ES ER EE ES SE EZ EE EE sa ss EG SEE EE Sa Hz KE Es sa Es Es DEI erg PREECE tt te TR M9 I HH I FI HH A Iaa a as Ea sa Ea Ez Ea sz ss ss EE EEE zal sn sa sz zs Eg eee eee eee eee eee Re DER SE Mt FEE CE GA ES Si S Es Ha FS ER EE SE B EE Fa SE EEE EEE EEE
5. 1 2t f RC 2 2 Audiodomain blogspot co uk 13 homewith dave daveisnot outlook com At low frequencies the series capacitor has a high reactance creating and open circuit As the frequency increases towards its cut off frequency the capacitor s reactance decreases and shorts the circuit Figure 10 shows a magnitude response plot of a high pass filter v Va w 6dB octave Sch T pp Figure 10 High Pass Magnitude Response Image In Horowitz W amp Hill P 1989 Low PASS By reversing the order of R and C a low pass filter is formed Figure 11 The reactance within the capacitor is at a high resistance at lower frequencies causing the curve to roll off at the other side of the cut off frequency Figure 12 H T Von Figure 11 Left First order Low Pass Filter Figure 12 Right Low Pass Magnitude Response Image In Horowitz W amp Hill P 1989 The cut off frequency is standardised as the point where 3dB of attenuation occurs as described by Wayne Corr When this Frequency Cutoff occurs the output signal is attenuated to 70 796 of the input signal value or 3dB 20 log Vout Vin of the input Storr W n d online Passive filters are stable due to their simplicity but to create steeper roll offs more orders must be used The reactive components will interact with each other and circuits become undesirably large Their inability to produce gain also ca
6. The L type attenuator pad is simply a voltage divide In appendix 4 the component Impedance and attenuation calculations are listed 2 2 3 IMPEDANCE BRIDGING If the load impedance is 10 times more than the source impedance it is called a bridging impedance Bridging results in maximum transfer of voltage from source to load Barlett B amp Barlett J 2002 The input impedance of the preamp must be set high enough in order to transfer maximum signal from the bass guitar Typically instrument output impedances are around 6000 with typical input stages set at 1M This is known as impedance bridging and is common technique to ensure the signal is suitably transferred from one device to another The alternative approach is called impedance matching where both input and output impedance are set to be equal 2 3 OPERATIONAL AMPLIFIERS Figure 3 and 4 show the modern schematic layout physical layout of an Operational Amplifier Op amp respectively OFFSET N1 m OFFSET N1 NC OUT 7 Vcc IN 521 OUT Vec 5 OFFSET N2 OFFSET N2 Audiodomain blogspot co uk 9 homewith dave daveisnot outlook com Figure 3 Left Op Amp schematic Figure 4 Right Physical Amp layout schematic National Instruments n d online image Pins IN and IN refer to the non inverting and inverting input respectively and Vcc are the inputs for the power supply Offset N1 and N2 control the offset null
7. gp d pi pp gr fp p p pi prp p p p j jj j pj p ppp 8 16 31 63 125 250 500 1k 2k 4k Bk 16k Solo 5 Take Snapshot Overlay List Clear Data Figure 32 Bandpass Phase Response Looking at this graph alongside the phase shift vs frequency graph of the TLO7X Figure 33 shows the potential source of the problem At these low frequencies both signals are shifting in phase over the O 100Hz frequency band As these two phase shifts are occurring in the closed loop of the op amp this could lead to the instability within this frequency band A potential improvement could be the use of a Sallen and Key band pass topology instead of a multi feedback band pass Figure 34 shows the schematic layout of a Sallen amp Key band pass filter Figure 34 Sallen amp Key Bandpass Filter Zumbahlen H 2008 The Sallen and Key topology does not use filter circuitry in conjunction with negative feedback Keeping the filter and amplification circuitry separate entities means that the filter performance is not limited by the amplifier performance The 250 and 600Hz were identified in research as problem frequencies when refining a bass guitar sound They have a higher Q that was designed for cutting boxy and Nasal frequencies around the 250 600Hz regions The bass tone is made smoother and stronger when these frequencies are cut When boosted the 250Hz filter can offer a degree of g
8. LARGE SIGNAL DIFFERENTIAL VOLTAGE AMPLIFICATION AND PHASE SHIFT vs FREQUENCY Veca 5 V to 15 V Ry 2 Differential Voltage Amplification Phase Shift N ttt TA 1 10 100 ik 10k 100k 1M 10M f Frequency Hz Audiodomain blogspot co uk 27 homewith dave A yp Large Signal Differential Voltage Amplification daveisnot outlook com Figure 26 Vs Phase Shift of TLO7X Texas Instruments 1978 Bruce Carter describes the Avd as Large signal differential voltage amplification AVD is similar to the open loop gain of the amplifier except open loop is usually measured without any load Carter B n d At maximum drive the highest gain value of an amplifier is 2 6dB This correlates to a 3MHz cut off frequency This leaves the circuit susceptible to interference from radio and telecommunication signals During the testing period of the artefact this never posed a problem and the design was not updated to include a filter to suppress this issue 5 3 FILTER PERFORMANCE 5 3 1 BANDPASS FREQUENCY RESPONSE Each individual band pass filter circuit was tested Figure 27 shows the 4 results of this test N 3 14 2 Red 100Hz Blue 250Hz Brown 600Hz Magenta 1000Hz Figure 27 Bandpass filter responses The filters were all designed to be at unity gain at their centre frequency Fc The graphs
9. By using buffers it is possible to output at low impedances a very important characteristic if long cables are used Inverting and non inverting buffers can be utilised for creating a balanced output pins 2 and 3 of an XLR connection Op amp output stages are often seen as an alternative to the transformer output stage Transformer output stages offer good isolation between devices as well as internally setting output impedance Unfortunately they are expensive to be considered in the initial specification 2 6 2 GROUND LiFT If a ground loop is contained within a system a switch must be installed to break the loop A switch disconnecting the balanced ground connection is necessary and can be seen on the initial schematic in appendix 12 2 6 3 OUTPUT IMPEDANCE Typical line inputs and amplifier inputs have a minimum impedance of 1M Therefore an output impedance of 100K or lower is required Practical outputs have an even lower impedance of around 600 Ohms The output Impedance is set by attenuation networks tuned to attenuate as little as possible while still containing the required impedance 2 DESIGN CONSIDERATIONS As discussed in previous sections of this chapter various techniques are used to achieve gain filtering impedance balancing and power supplies Consideration needs to be paid to the way these sections interact A 20dB L Type pad and a Decoupling capacitor will form a High Pass Filter Component values need to be chosen
10. This is inefficient as power is dissipated as heat energy Using the more efficient class B or AB push pull system may improve the expected battery life 5 8 OUTPUT LEVEL The following subsections discuss specifics regarding the maximum output capabilities including 12dB boost at 1kHz of the circuit Everything below the maximum value is controllable with via the gain stage potentiometers Appendices 29 and 30 show all output readings and calculations 5 8 1 LINE LEVEL As the preamp can operate up to an 18v supply the output gain can be pushed as high as 14 2Vpp This maximum dBu reading of 16 2275 shows a safe signal level for professional audio equipment to accept Most professional audio systems have a maximum capacity of 24dBu at the input This puts the maximum output of the bass 12dBu above the 4dBu line level standard and 8dBu below a level that would be considered unworkable At a normal operating level the preamp outputs much closer to line level During testing the preamp was pushed to use the entirety of its headroom powered by 9v supply The maximum output of this test gives 8 3895dBu a closer reading to the intended 4dBu line level output Maximum gain before clipping of 13 6248dB is acceptable and only little over the average of many commercial products 5 8 2 INSTRUMENT LEVEL The instrument level is the normal line level signal through an increased pad The pad lowers the output to a lower level expected by bass guita
11. gt Tg o gt 5 D N ll Audiodomain blogspot co uk homewith dave daveisnot outlook com i02 a13 B ael HJEL GEH 54 meaa 1937 Sore Er Teak us M m a ng 12 R17 C11 33k 38i z Lal R1 AGGP Presence BOL B SKT g g darz 4m 47Bp uf g Ln i ACE gi 94 L g ven Ben Fie LL 29 t8 1L 59 15 MODULE PHOTOS Power Supply Input stage and Pad Audiodomain blogspot co uk 60 homewith daveisnot outlook com First gain stage Summing Amplifiers Band pass filters and amplifier Second Gain stage 16 WiRE COLOUR CODING Carried Signal Vcc Ground AC Signal Filter Amp Input in 17 MOLEX PIN LAYOUT V V 2 Ov Signal F I 18 INITIAL SCHEMATIC Audiodomain blogspot co uk homewith dave daveisnot outlook com 64 Audiodomain blogspot co uk homewith dave 65 19 FINAL SCHEMATIC 66 Intrurnent Output 100k 1OuF 10k 33uF 12K 10K oost
12. 30 4 9v MAXIMUM INSTRUMENT dB 3 36 AB gain 2010410 12 8 9431 dBu2 6 72Vpp 3 36Vp Audiodomain blogspot co uk 72 homewith dave daveisnot outlook com Vrms 3 36Vp x 0 707 1 68 dBu 2010410 0 775 3 7085 30 5 1 2VPP REFERENCE SIGNAL dBu 1 2Vpp 0 6Vp Vrms 0 6Vp x 0 707 0 4242 20 Ne u 20810 5 7785 dBu 5 2346 31 NYQUIST Ismin fmax X 2 Where Fsmin Minimum sampling frequency Fmax Maximum frequency needed 32 METRIC HALO MOBILE IO 2882 SPECIFICATIONS 4 channels balanced XLR inputs Each input has e 24 bit A D converters 110dB SNR remote controllable pre amps with 40 dB of gain remote switchable input impedance characteristics remote switchable 20dB pad e remote switchable 48v Phantom power with current limit 4 channels balanced TRS inputs Each input has 24 bit A D converters 110dB SNR remote controllable pre amps with 40 dB of gain remote switchable input impedance characteristics remote switchable 20dB pad e remote switchable 48v Phantom power with 10mA current limit 8 channels balanced TRS outputs Each output has 24 bit D A converters 120dB SNR remote controllable output gain from 12dBV up Audiodomain blogspot co uk 73 homewith dave Mic Line Inputs Line 4 Gain Range Line 10 Gain Range Inst Gain
13. Audiodomain blogspot co uk homewith dave daveisnot outlook com 54 12 INITIAL SCHEMATIC Audiodomain blogspot co uk 55 homewith dave 13 SCHEMATIC AND COMPONENT CALCULATIONS 13 1 L TYPE ATTENUATION PADS Equations Instrument Input amp Input Pad Impedance Z 10000 Attenuation 20 Ideal Value R 9000 R 111 11 Balanced Output Z 600 Ideal Value 221 43 1025 83 13 2 GAIN STAGES Ry 2 13 3 MULTI FEEDBACK COMPONENTS Equations Audiodomain blogspot co uk 56 homewith dave daveisnot outlook com Bass Q 2 f 100 C 0 22uF 1 28937 262 R 14468 6311 R 2066 947 Mid a 0 4 f 250 C 0 18uF Ideal Values R4 28294 2121 R 14147 1060 R 456 3583 Actual Values R4 27K R 13K R 470 Ideal Values 21220 6590 10610 32954 342 2686984 Actual Values 22K R 11K 330 Treble 0 1 f 1500 C C 0 01uF Ideal Values 21220 6590 R 10610 32954 R 10610 32954 Actual Values 22K R 11K R 11K Audiodomain blogspot co uk homewith dave daveisnot outlook com Audiodomain blogspot co uk homewith dave 58 peor 9002 20 21 Aq 14 TRIANGULAR EAR 21 SCHEMATIC Lavel Tak D 2 Q gt p t D m
14. B n d Understanding Op Parameters online Available at http www ti com lit ml sloa083 sloa083 pdf Accessed 14 April 2012 Zumbahlen H 2008 online image Sallen Key filters Available at http www analog com static imported files tutorials MT 222 pdf Accessed 14 April 2012 Bohn J 1986 Constant Q Graphic Equalizers online Available lt http rane com pdf constang pdf gt Accessed 14 April 2012 Bohn J 2000 Audio Specifications online Available at http www rane com note145 html Accessed 14 April 2012 Lofft A Colquhoun amp Cumberland T n d How much distortion can we hear in music online Available at lt http www axiomaudio com distortion html gt Accessed 14 April 2012 Papin B n d Signal to Noise Ratio online Available at http www bcae1 com sig2nois htm Accessed 14 April 2012 Math Works n d online image Feedback Amplifier Design Available at lt http www mathworks co uk products control demos html file2 products demos shipping control o pampdemo html Accessed 4 May 2012 Audiodomain blogspot co uk 44 homewith dave daveisnot outlook com APPENDICES 1 ELECTRONIC CIRCUIT EQUATIONS 1 1 OHMS LAW Ohm s law states that voltage is proportional to the current through a load Figure X Ohms Law Where V Voltage Current R Resistance 1 2 S CURRENT LAW Kirchoff s current law states that
15. 27K 22uF 15K a AA EQ TLO 4 100K 10K 22K 2AF oost 0 1uF 22K 12K EQ amp X 100K 330 014 Artefact Final Schematic Size Document Number stbrDave Sampson 0860120 W 100K 3i dnesday 68uF 10K daveisnot outlook com Audiodomain blogspot co uk homewith dave 67 20 FREQUENCY RESPONSE MEASUREMENT TECHNIQUE Opectrafoo produces sine sweep Sine sweep parameters otart Frequency 20Hz End Frequency 20kHz 1 2Vpp Output Duration 2 Minutes owept Twice s 21 PHASE RESPONSE MEASUREMENT TECHNIQUE Spectrafoo sweeps the circuitry and compares the output to the input e This produces the transfer function for phase analysis Filter Output Spectrafoo Mobile IO lt Filter Input 22 IHD N MEASUREMENT TECHNIQUE opectrafoo produces a pure sine wave at the circuit input opectrafoo uses a tight notch filter to remove this signal from the output signal opectrafoo then measures the leftover signal for harmonic distortion and noise Circuit receives 1 2Vpp and outputs 1 2Vpp Ed s E 23 SIGNAL TO NOISE RATIO MEASUREMENT TECHNIQUE oignal generator produces 1 2Vpp Sine wave at the input 1 2Vpp measured at the output Bandpass filter limits the measurement to an audible frequencies Oscilloscope measures noise while circuit is in quiescent state Oscilloscope set to trigger highest voltage obtaining a worst case measur
16. 2L Sancamp Bass 46 MESI E E E EE nana E E 46 A EB MIC ODAS ANNE 47 OO IN S E __ ___ 47 Bass Ou ultpul MeasUu rIBBuu uu l u 48 Voltage Divider EQUA HONS a a 49 Slew Rate uuu 50 6 amp 50 50 51 Audiodomain blogspot co uk 4 homewith dave daveisnot outlook com 7 Filter Definitions 51 8 Impedance EQUSLIOLIS uuu uu L aR 52 ST j rl E 52 27 OMI L u u 52 S Series 52 S Min HOC CAINS 52 aas 52 9 1 Low Pass Transfer 52 9 2 Hiei Pass Ia Serr UN CHION uuu u l 52 9 3 Centre FIC OU CNC 52 ay j ree 53 10 MFBP Equations riii ______ _ 53 53 J Hill lena r __
17. an issue at the unity gain point below the Fc When the filter is boosted the unity gain point cuts and vice versa It s a property that is most evident on the 100Hz filter were the situation becomes serious Figure 32 is the phase response of the bandpass filters Appendix 21 shows how this graph was produced Red 100Hz Blue 250Hz Brown 600Hz Magenta 1kHz Audiodomain blogspot co uk 31 homewith dave eoo Transfer Function gt Power vs Frequency 7 Phase vs Frequency RR RR a ES J J P P Pp oP j Up IN II I ro iu ae OO L LLL L LT 150 LL L T L C LL T __ TT I I L T j i 41411 l 12 I LTT II TISIII I I II LU TT TTT TT TT TT Lac 11441 2
18. be at half the sampling frequency 48kHz If the circuit is producing high frequency oscillations or is inducting high frequency noise they would not be present in the results Spectrafoo uses Fast Fourier Transform FFT algorithms to convert the digital data from the time domain and into the frequency domain As stated in the Spectrafoo manual The spectrum analysis within Spectrafoo utilizes a multichannel high resolution real time FFT engine At its highest resolution setting Spectrafoo utilizes 64k point FFTs allowing you to see features with widths as small as 2 3 Hz Metric Halo n d Realistically the resolution of the final graphs look worse than suggested in the manual At lower frequencies it would seem that Spectrafoo has only been able to estimate the frequency response later smoothing it with an averaging algorithm As suggested in the mobile IO 2882 Users Guide a pseudo cable was made to utilise the differential line input stage of the unit Figure 22 shows how this cable was wired Figure 22 Pseudo jack wiring Image In Metric Halo 2012 Hugh Robjohns comments Impedance to ground of the two signals is very different The Hot wire will have an impedance of a few tens of Ohms while the cold wire has an impedance of zero Ohms So any interference that gets into the cable won t be a true common mode signal Robjohns H 2008 As there is electrical isolation between source and loa
19. into its enclosure this could offer increased signal to noise performance by reducing interference A significant use of molex terminals is another contributing factor to the noise in the circuit As will be discussed in section 3 1 a printed circuit board would eliminate the large volume of molex connections Although this measurement could be improved Berry Papin offers a more realistic and practical incite into how this reading could be perceived when he comments Many people don t realize how much a 30 or 40 dB ratio is A piece of electronics equipment with a S N ratio of 8088 may be good enough for all but the best systems Papin B n d This opinion puts a better perspective on the reading Although an extra 20dB would bring the signal to noise reading to professional standards a reading of 60dB of is acceptable 5 6 IMPEDANCE The testing procedure for input and output impedance can be found in Appendix 24 and 25 The measured 1 35M input impedance is higher than the designed value This has no effect on the operation of the input stage As long as the input impedance is ten times as large as the source impedance the circuits will be bridged maximum voltage transfer The output impedance of a bass guitar varies depending on pickups potentiometers and internal preamps It s a safe assumption that the output impedance of a bass guitar isn t higher than 10kQ With two outputs comes the need for two separate output impedances Instr
20. reinforcement at the pass band U3 controls the gain of the pass band As seen in Figure 15 U3 is a unity gain amplifier so when summed the output will sum to double or 2698 the input U4 is the MFBP topology as discussed in 2 4 4 R9 dictates how much signal is sent to U1 and U2 Many filter sections can be used in parallel under this topology as long as mixing resistors R8 are used before the summers Using the filter specifications discussed in Chapter 3 Figure 16 shows the predicted magnitude response of a 4 band equaliser Figure 16 Expected Response of initially designed filters 2 4 6 FILTER IMPLEMENTATION Commercial Preamps offer ideas regarding how flexible filters can be implemented without sacrificing simplicity of use The John East J Retro unit has a unique treble control When boosted the centre frequency is 3kHz When cut the centre frequency falls to 1 5kHz Using the boost cut topology this could be achieved by using a dual gang potentiometer One gang boosting 3kHz and grounding the cut the other cutting 1 5kHz and grounding the boost The initial schematic discussed in Chapter 3 and referenced in appendix 12 shows a switch selecting between 2 frequencies controlled by the mid control An in depth look at equalisation tendencies in bass guitarist could open up a wealth of opportunities for improving filter functionality in bass preamps 2 5 POWER 2 5 1 POWER SUPPLY Power supplies step down the mains v
21. signal At the filter stage bias voltage is present through the DPDT switch meaning a pull down resistor would lower the bias and make the filter amplifier un operational By moving the DPDT switch to after the mixing resistors the bias could be isolated and the switch would not suffer from this issue The initial response to this problem was to use capacitors to block DC from the switch This unfortunately led to some undesirable filtering of the audio signal The late consideration of this problem left no opportunity to implement a permanent solution Audiodomain blogspot co uk 23 homewith dave When building on strip board shorting connections occurred Continuity testing with multimeter helped pinpoint the problem but finding the exact point of contact proved difficult On several occasions complete boards were rebuilt to resolve a shorting issue This was a time consuming exercise lacking in efficiency A second issue with strip board came with the wearing down of the copper tracks By the later stages of the artefact strip board that had been used in the initial stages started to oxidise Tinning wash or varnish was not used to protect the copper tracks as all of the boards were under development right until the end of the build The solder joints were not defluxed for this same reason Decoupling capacitors throughout the circuit formed high pass filters in conjunction with resistive loads During
22. that will not interfere with the audio spectrum If two parallel resistive paths to ground or virtual ground are placed before the input buffer the overall input impedance will be limited by the lower resistance value The circuit load cannot draw more current than available from the power supply This shouldn t be an issue as most regulated power supplies at least 1 3A of current available If too much current is drawn from a battery the life span of said battery will be poor Audiodomain blogspot co uk 19 homewith dave daveisnot outlook com 3 3 1 SPECIFICATION A specification sheet for the preamp is included in Appendix 11 This gives rough design parameters that were used to create an initial schematic Many of the parameters are based upon existing models that have proved successful The specifications will be practically improved upon where possible 3 2 SCHEMATIC Appendix 12 is a schematic designed to the initial specification using the theory discussed in Chapter 2 The individual sections from the schematic will be mocked up on breadboard gradually adding different sections to the circuit will help diagnose any hardware problems Appendix 13 includes all the necessary component calculations 3 3 BILL OF MATERIALS Figure 20 is the Bill of materials required to build the initial schematic Over time as the design changes the list of materials will also change Resistors 9K1 Audiodomain blog
23. the testing of the circuit it was immediately noticeable that the cut off frequency of these high pass filters was too high around 400Hz The 0 68uF decoupling capacitor after the second gain stage was the main issue it was brought up to 6 8uF This helped bring the practical cut off back down to the sub frequencies Work continued with other decoupling capacitors until the values were raised enough to lower the cut off to an acceptable level Overall a lot of lessons were learnt during the production phase of this project Looking forward to building the next prototype more time will be allocated to production The issues have experienced will be taken into consideration and improved upon 4 4 FINAL SCHEMATIC The final schematic shows significant differences from the initial schematic proposed in the literature review Throughout the construction of the artefact every opportunity to improve the design was taken Appendix 18 shows the initial schematic and Appendix 19 shows the final schematic Instead of two non inverting gain stages two inverting gain stages were used This was implemented because the non inverting gain stages by nature reference ground at the inverting input of operational amplifier This posed difficulties as the circuit is biased at half the supply voltage The input pad was redesigned because the 20dB bass dropped bias voltage as well as AC signal voltage The bias drop became problematic at the first gain stage t
24. 2 The signal swings between and Ov with reference to Op Amps are unable to swing all the way between their Vcc and rails Therefore it is important that the split supply is as accurate as possible to get the most out of the op amp If a bias is set at 7596 of the power supply it is possible that little or no positive swing will occur Using smaller voltage divider resistors can help achieve a more accurate division however the amount of current drawn increases as resistance gets smaller If this current is needed elsewhere in the circuit more accurate voltage divider may have to be sacrificed by higher resistor values Integrated circuits such as the Texas Instruments TLA2426 are readily available to split single supply lines They simply contain refined voltage dividers and buffer stages The use of integrated circuits at this stage of the bass preamp circuit is more than likely unnecessary 2 5 4 DECOUPLING A final point to recognise is the use of capacitors to decouple the DC offset from the AC signal This is important from protecting the bass or line level application from DC voltage A capacitor in series is used to remove the DC from a signal Precision values aren t required a O 1uF capacitor will suit audio applications fine Smaller capacitors may be needed for higher frequency instruments Audiodomain blogspot co uk 18 homewith dave daveisnot outlook com 2 6 2 6 1 OPERATIONAL AMPLIFIER BASED
25. BASS GUITAR PREAMP DESIGN Dave Sampson l d like to dedicate this work to my long term partner Louise Without her love patience and encouragement this work would not have been possible Audiodomain blogspot co uk homewith dave daveisnot outlook com TABLE OF CONTENTS gorge n 7 LERTON T 7 1 2 Aims S ODIBSCLHVBS ERO 7 1 3 Summary of Report Structure sesta nennen nnn 7 AEN PPA HR 8 2 1 EXISTING OON uu 8 2 ij SAG 22222 8 E 8 La VC T 9 2 2 9 9 2 3 Operational Ampliflers UU UU Le 9 Tm 10 2 5 11 N 11 2 2 9 B lier DRUG 12 PAP I IU IE BI 13 vo estat 13 T 13 27 9 PF MID NECEM UMEN MEUM EE 14 2 4 4 Multiple Feedback Bandpass Filter Topology
26. E u u a usasapa apakuq aea 28 5 2 2 Filter aceti ____ 29 l 33 5 4 Total Harmonic Distortion Noise 37 5 5 Signal to Noise 38 S6 38 Audiodomain blogspot co uk 3 homewith dave daveisnot outlook com 5 7 Current dM en E 39 Dee OULU HOVE RR m 39 WERT H sees 39 9 8 2 Instrument LCV El 39 CHop er ___ ____6___ 40 GPU MOE Q us E OT OO asa 40 SZ COM CIUSI ONS RR EE m ET 41 BENNETT mm 42 POC GNC SS S E akak 45 1 Electronic Circuit EQUATIONS 45 45 geste c d Oo L ____ 45 _____________ __8_8 45 1 4 Therevin s eguivalent 45 icr cuc 46 li 46 2 Commercial Preamp Specifications 46 DAS Tech
27. EDANCE 9 PASSIVE FILTER EQUATIONS 9 1 Low PASS TRANSFER FUNCTION l 2nf RC Vout Vin 1 o2 R2 C2 2 9 2 HiGH PASS TRANSFER FUNCTION 2nf RC Vout Vin 1 xf RC 9 3 CENTRE FREQUENCY fe 1 2 Audiodomain blogspot co uk 52 homewith dave daveisnot outlook com 9 4 BAND 55 Q 23 10 MFBP EQUATIONS SR3C Ns Fo BA S S2R C 1 Ro 2 R3 R 1 2 oi 20 Ay p T f C A Kr Where A Gain at Resonant Frequency s Complex variable representing amplitude and phase Q Quality Factor 2Q Williams A amp Taylor F 1995 11 INITIAL SPECIFICATIONS TLO71 TLO72 TLO74 Acceptable Slew Rate Low Noise 100dB CMRR Cheap Low Power consumption Controls Controls Volume Bass Mid Audiodomain blogspot co uk 53 homewith dave daveisnot outlook com Mid Frequency Switch DPDT Treble Pad Switch DPDT Equaliser Bass 12dB O 100Hz Mid 12dB Freq 1 250Hz Freq 2 600Hz Treble 12dB 2 5 kHz Active Multi feedback networks Input Impedance 1M Switchable 20dB Pad L Pad Design Buffer Stage Unbalanced Jack Gain Stage IC Based Non Inverting 0 12dB Variable Gain Line Output Stage Impedance 600 Ohms Ground Lift XLR Instrument Output Impedance 10K Unbalanced Jack 20dB pad Power 9v Battery DC Jack DC Jack cuts battery Power Bias at 4 5V
28. ICATIONS 6 1 REAL Manufacturer oflMaximum Supply Voltage 50 Audiodomain blogspot co uk homewith dave daveisnot outlook com 2 gt O gt C 2 x gt 2 NO O lt WN lt rN NO O r5 G 613 ZHIN9T YYA 1 pueg 6 2 IDEAL e Infinite Bandwidth e Infinite Gain e Infinite Input Impedance Output Impedance ONoise e Infinite CMRR e Infinite Slew Rate 7 FILTER DEFINITIONS The band of frequencies passed by a filter network Stop Band The band of frequencies rejected by a filter network Cut off frequency The frequency at which the signal has been attenuated by 3dB Audiodomain blogspot co uk 51 homewith dave daveisnot outlook com Band Width The difference between 2 cut off frequencies in a band pass or stop filter Quality Factor A numerical representation of bandwidth in relation to Centre frequency Centre Frequency The centre frequency of a pass band Order The amount of orders in a filter is determined by the amount of reactive components Transfer Function The Difference between the output and input of a signal 6 IMPEDANCE EQUATIONS 8 1 IMPEDANCE Z R TX Jo L 1 joC 8 2 OHMS LAW 17 8 3 SERIES IMPEDANCE Z4 Z gt T Ly 8 4 PARALLEL IMP
29. In the 1960 s The Beatles were unable to be heard over their loud audiences the DI box was invented to give public address support to their backline Although bit strictly a preamplifier this was the first instance of interfacing the bass guitar with line level equipment In 1975 Sterling Ball tested a Leo Fender bass design the Stingray This bass is regarded as the first bass utilising Active electronics Containing a 9v battery and a 2 band Equaliser the bass outputted at line level removing the need for a DI box Passive DI boxes were still used to convert the unbalanced jack signal into a balanced XLR signal As technology has progressed over the past 30 years the opportunity for new preamps designs has constantly emerged Filters distortion and compression are common characteristics found within today s units 1 2 AIMS amp OBJECTIVES The aim of this project is to gain a thorough understanding of the electrical design elements of an audio application specifically bass guitar preamps The artefact will be a functional bass preamp designed to operate within the parameters of a technical specification discussed in this report The artefact will address the problem of interfacing the bass guitar with audio line level instruments Thorough testing of the unit will be undertaken and documented in a second report 1 3 SUMMARY OF REPORT STRUCTURE In chapter 2 considerations regarding the theoretical and physical design are discussed Wit
30. ND Lift 3 7dB 75Hz Channel 1 Contour Shaping 6 3 dB 350Hz Boost footswitch 11dB 5 6kHz Input selection footswitch 5 12dB 75Hz 2xinst IN Mid 10dB 470Hz Balanced DI Out High 12dB 5 6kHz Unbalanced inst Out TRS Insert Impedances Input 470 Output 10K DI Out 6000hms 3 BAss OUTPUT MEASURING Using the 10M probes on the oscilloscope the instrument was sensitive enough to get good readings from the bass guitars The following diagram shows the test setup Audiodomain blogspot co uk 48 homewith dave daveisnot outlook com Signal From Guitar w _ TTF di 10M Probe Oscilliscope The peak output voltage was held on the oscilloscope and noted To determine the absolute work peak possible the bass was played extremely aggressively 4 VOLTAGE DIVIDER EQUATIONS A full list of equations used to find the output voltage of a resistive or reactive voltage divider oni Out R A 1020 1 Ry 1020 1 R 1020 1 Where A Attenuation Z Series Impedance Vin I Ztotal Ztotal 4 22 Z2 Audiodomain blogspot co uk 49 homewith dave Vout sa Vin Vout 5 SLEW RATE EQUATIONS Slew Rate 21Vpy fc Baker B n d Where Peak to Peak Voltage fc Cuttoff Frequency 6 OP AMP SPECIF
31. Phase Lead Capacitor C Math Works n d online image Apart from the obvious improvements needed on the 100Hz filter a redesign could improve the filters dramatically For example moving the 100Hz filter down towards the sub regions of the audio spectrum would help give the preamp more weight and much more low end versatility The 250 and 600Hz filters perform very well when cut but when boosted they can sound unnatural To improve this a dual gang pot could be used That way different parameters could be used to sum between the boost and cut summers A lower Q could be achieved when boosted without affecting a high performance of its cutting function The preamp has a transparent effect on the overall tone of the bass signal With experimentation a more in depth analysis can be made of a characteristic bass tone These studies could become the basis of fixed filters within the signal path adding a characterizing tone before the filters are tweaked Alternatively they could be wired into the summing amplifiers and faded in at a given level This would give the user a master control for a hard wired tone made up of several filters Furthermore due to the filter topology used both sides of a linear pot could be used to control a different set of filter parameters Audiodomain blogspot co uk 40 homewith dave daveisnot outlook com For example a set of filters that would give a vintage tone could be wired to the cut half of a potentio
32. Range Mic Gain Range Mic Pad Gain Range Line Input Impedance Instrument Input Impedance Mic Input Impedance Mic Pad Input Impedance Maximums Max Gain Preamp Headroom Phantom Power Output Audiodomain blogspot co uk homewith dave daveisnot outlook com 2 dB 40 5 dB 13 8 dB 28 7 dB 0 dB 42 5 dB 0 dB 42 5 dB 20 dB 22 5 dB Q 200k Q 200k Q 12k Q with phantom Q 6k Q with phantom 42 5 dB 20 dB above Digital Clip 48v Regulated high current individually switch able P48 test compliant short circuit hot swap pro tected 26 dBu 74
33. The only alternative was to use a flat wood bit This made the hole untidy and sharp Efforts were made to remove all sharp edges with a file The conductive strip board tracks were covered with PVC LX tape in order to stop the circuit boards shorting The internal surface area of the enclosure was covered in duct tape to make sure the copper didn t short to the chassis The wires connecting the molex terminals were twisted and became rigid this was enough to hold the boards in place As the outer tin is conductive this offers protection from the coupling of interference and noise The tin makes conductive contact with the DC jack giving external noise a route to ground bypassing the signal path 4 3 ISSUES DURING CONSTRUCTION At the input of the preamp only half of the bias voltage was received at the non inverting input of the operational amplifier After debugging the system no solution was found and in the best interest of completing the artefact work continued beyond this problem Figure 21 shows a similar circuit to the input section of the final design Using a diode model of a transistor the schematic shows how the bias current flows through the circuit Audiodomain blogspot co uk 22 homewith dave daveisnot outlook com speaker M input Figure 21 Bias current draw Kuphaldt T n d online image As Tony R Kuphaldt states with relation to Figure 21 A voltage divider s output depends not only on the size of its consti
34. ______________ _ 55 13 Schematic and Component Calculations 56 13 1 L Type Attenuation 56 u 56 13 3 Multi feedback 5 2 0020 0 1130 200000000000 56 14 Triangular Ear 21 Schematic eu Eu 59 15 Module PNOTOS u uuu saat unuman S EE 60 Wire uuu u uu emt un EE 64 DF Molex Pin Rm 64 Io til sc hom i E TT 64 195 66 20 Frequency Response Measurement Technique 68 21 Phase Response Measurement Technique 68 22 THD N Measurement Technique 68 Audiodomain blogspot co uk 5 homewith dave daveisnot outlook com 23 Signal to Noise Ratio Measurement Technique 68 24 Input Impedance Measurement Technique 69 25 Output Impedance Measurement Technique
35. cally related content and other noise at the output that isn t present at the input As Dennis Bohn from Rane corporation states Distortion analysers make this measurement by removing the fundamental using a deep and narrow notch filter and measuring what s left using a bandwidth filter The remainder contains harmonics as well as random noise and other artefacts Bohn J 2000 Without the other artefacts a distortion reading may not give a true representation of artefacts inducted into the signal through the circuit Appendix 22 shows the test set up for measuring THD N The 0 005 THD N reading shows good performance in relation to standard THD specifications for professional audio equipment The preamp was tested at unity this significantly improves the THD N reading If the preamp were pushed towards saturation a lot more harmonically related content would be present at the output 5 5 SIGNAL TO NOISE RATIO The signal to noise ratio gives the ratio between the noise floor of the circuit and a reference level The reference level used during this test is 1 2Vpp Volts peak to peak A 20 20kHz band pass filter was added before an oscilloscope This was to remove any signal that would be measured unnecessarily Appendix 23 shows the signal to noise ratio testing procedure and calculations The noise worked out at 59 233dB This is a fair way from professional audio standards The reading was taken before the circuit was fitted
36. d and full shielding of the signal the pseudo jack cable ensured the signal coming into the 2882 was as noise free as possible The 2882 User s guide also Audiodomain blogspot co uk 25 homewith dave daveisnot outlook com states its input as having 11088 signal to noise Specifications like this made the 2882 as the most ideal hardware testing solution The Metric Halo IO 2882 specifications can be found in Appendix 32 5 2 FREQUENCY RESPONSE Appendix 20 shows the test procedure for the frequency response tests Figure 23 shows the frequency response of the entire circuit N 3 14 SN Set Black Connected Red EQ measured flat Figure 23 Frequency Response Without the filters connected we see a steeper roll off around 50Hz This could be the unintentional loading effect the filters have on the circuit If we use figure 24 as an example we see a high pass filter formed by a resistor and a capacitor Audiodomain blogspot co uk 26 homewith dave Figure 24 Passive High Pass Filter Kuphaldt T 2007 online image If Rload is a representation for the total impedance of the filters we can see how the filters would affect the frequency cut off By increasing R adding filters the frequency cut off is decreased Figure 25 proves this 1 fcutoff Figure 25 Fc calculation All high pass filters forme
37. d by the resistor capacitor networks were designed to have a maximum cut off of 40Hz practically the cut off seems higher lying at around 100Hz As mentioned in Chapter 1 the initial cut off has been improved significantly since the first build The frequency response graph shows that with the EQ connected there are some slight contours in the shape of the response This is most notable in the region of 250Hz were there is a cut in frequency Despite the pot being tested to be in its middle position the pot would still appear to be in a counter clock wise cut position This could also be the cause of a slight dip in the 1kHz region Using pots with centre indents would improve the test procedure for a flat EQ as they are much more accurate in centre position It is also worth considering the 20 tolerance rating on the potentiometer this could bias the centre point towards the cut position Apart from this we see slight attenuation when the filters are connected The preamp shows good frequency performance up to 20kHz At no point in the direct signal path the circuit excluding filters is a low pass filter formed This makes the high end frequency response ofthe circuit extremely high The operation amplifiers op amps in the circuit are the only high frequency limiting components in the design Figure 26 shows the large signal differential voltage amplification Avd phase shift vs Frequency plot from the Texas Instruments TLO7X data sheet
38. ement O O O O O Signal 20 20kH one Circuit Oscilloscope Generator Bandpass Filter Reference Level 1 2Vpp Audiodomain blogspot co uk 68 homewith dave daveisnot outlook com Measured Noise Level 0 00131Vpp Calculation V 5 20109 V S N 201 Nay 20109 o 00131 Where Vr Reference Voltage Vn Noise Voltage 24 INPUT IMPEDANCE MEASUREMENT TECHNIQUE Circuit Under Test http www zen22142 zen co uk Theory images zin gif 25 OUTPUT IMPEDANCE MEASUREMENT TECHNIQUE Circuit Under Test htto www zen221 42 zen co uk Theory images zout gif 26 IMPEDANCE CALCULATIONS 26 1 INPUT Audiodomain blogspot co uk 69 homewith dave daveisnot outlook com _ 0 529 0 526 mc 10k 0 526 n 1 35M 0 0000389 26 2 OUTPUT Instrument Output V M _ 10k 1 0 42363 Z E15 d 0 42363 a Line Output _ 10k 1 0 824562 out 0 824562 27 CURRENT DRAW MEASUREMENT TECHNIQUE e Multimeter set to 20mA Power Supply 28 TLO X ELECTRICAL CHARACTERISTICS Audiodomain blogspot co uk 70 homewith dave electrical characteristics Vcc 15 V unless otherwise noted TLO071M PARAMETER TEST CONDITIONST TL072M s ef ao Input offset e R lt 50 0 a j Temperature coefficient of C O x input offset
39. h reference to existing solutions the individual sections of a bass preamp are looked at closely Chapter 3 outlines an overall plan for work to be carried out on the artefact Initial theoretical schematic design specification and bill of materials based on the design considerations discussed in chapter 2 will also be included Chapter 4 looks at the build process and the issues raised during production Chapter 5 documents the testing and test results of the preamp This report concludes with Chapter 6 Chapter 6 discusses future development plans and looks to summarize the entire project Audiodomain blogspot co uk 7 homewith dave daveisnot outlook com 2 Fundamental electronic theory referenced throughout this section can be found in Appendix 1 2 1 EXISTING SOLUTIONS Commercial products are readily available to perform the task of bass guitar pre amplification Designs address different problems but fundamentally are the same solution A full specification of the leading bass preamps is listed in appendices 2 These solutions will be used to create a specification for the artefact 2 2 INPUT STAGE 2 2 1 INPUT SIGNAL In order to design the input stage of any electronic circuit the output of the previous stage needs to be studied Datasheets and specifications for pickups often don t give peak voltage readings Vpmax Using a 10M probe of an oscilloscope peak voltages have been recorded from 3 bass guitars This
40. he ability to safely reconfigure boards With no connector present the terminals provided good testing pins A combined total of 95 molex terminals are used within the circuit Common practice is to use a crimp tool to secure a wire in a crimp terminal Without access to a crimp tool wires were soldered into the crimp terminals Appendix 16 shows the molex wire colour coding system used throughout the circuit At every input and output there is a 4 way molex terminal Appendix 17 shows the pin layout for this This was standard across all the boards ensuring they are all compatible with each other meaning that the power board can be connected to any other module for testing There is a test pin installed on the power board to input a signal onto 4 output pin All components were tested to be within their 5 tolerance before they were placed on the boards This ensured that all components were of the same value as the designs a precaution taken to maximise productivity during construction An exception to this is the potentiometers used these were rated at 20 During the discussion of test results in section 2 component tolerances are assumed to be a contributing factor to minor discrepancies in results 4 2 ENCLOSURE The circuit was enclosed in a tin box All holes were made in the tin with a step drill bit The only issue came when drilling out the XLR terminal A step drill bit big enough to cut the 25mm hole required was unavailable
41. he bias was too close to ground for the amplified to avoid The signal was decoupled before the switch and then reapplied after This made sure that the resistive network was only applied to the AC signal The resistive networks at the line output stage were reconfigured to decrease the loading effect of a long XLR cable Capacitance present in a long cable can alter phase shift of higher frequencies around the feedback loop This can cause oscillations due to positive feedback at the inverting input This consideration was made after the initial design of the instrument stage and so the larger resistive network was not adapted in the same way Audiodomain blogspot co uk 24 homewith dave daveisnot outlook com 5 5 1 5 NOTES Throughout the testing input sine wave at 1 2 was used this is the worst case input signal as stated in the literature review This level proved to represent an extreme worst case peak and should be taken into consideration in analysis of results The vast majority of testing was done using Metric Halo s Spectrafoo software and Mobile IO 2882 hardware In an ideal world tests would be carried out on specialist hardware such as a prism dScope The Mobile 10 2882 was running at 96kHz its maximum sampling frequency In order to avoid aliasing there will be a low pass filter before the analogue to digital converter ADC According to the Nyquist theory Appendix 31 the filter cut off will
42. ing Amplifier National Instruments n d online image The Signal is coupled with a feedback loop into the inverting input of the op amp The voltage divider set by R1 and R2 control the gain The gain is calculated by R2 gain Ga Inverting amplifiers are said to have negative gain This does not mean attenuation it refers to the 180 degrees change in polarity The output of the op amp will always try and make the inverting input equal to Vcc 2 Therefore if the input is positive the op amp will output negative and vice versa 2 3 4 BUFFER AMPLIFIER Buffer stages Shown in Figure 7 an integral part of preamplifier design Horowitz and Hill state the buffer amplifier is simply non inverting amplifier with R1 infinite and R2 zero gain 1 Horowitz amp Hill 1989 Figure 7 Non Inverting Buffer Amplifier National Instruments n d online image It isn t uncommon to see several buffer stages in a circuit examples of this can be seen on the Triangular Ear Bass 21 Schematic Appendix 14 They are very useful at the input stage of circuitry as they have a Audiodomain blogspot co uk 12 homewith dave daveisnot outlook com high input impedance and a low output impedance This draws maximum voltage from the preceding circuit and supplies maximum current to the proceeding circuit 2 4 FILTERING A filter is a device or network for separating waves on the basis of their frequency McMann
43. is required to unbalance the signal As a bass guitars output is unbalanced the CMRR characteristics of an operational amplifier are of less concern NUMBER OF CHANNELS Individual Op Amps can process multiple channels saving space and cost An example is the National Instruments TLO74 that has 4 channels If a circuit requires many Op Amps then it may be desirable to use integrated circuits that can offer more channels This will save on cost and PCB Strip board space SUPPLY VOLTAGE An Op amp has a maximum voltage that can be applied to its Vcc pins Too much applied voltage can damage the Op Amp If an operational amplifier can t accept a large enough supply voltage then there is a chance that an audio signal will clip the Op Amp For example an Op Amp with a supply of 5v will clip an audio signal that has an amplitude of 1 6v 2 3 1 SPECIFICATIONS Appendix 6 1 shows the study of the most common op amps found in commercial and DIY audio applications Appendix 6 2 includes the specification of the ideal op amp Audiodomain blogspot co uk 10 homewith dave daveisnot outlook com In reference to this Appendix 6 1 the appropriateness of different Op amps for audio applications can be reviewed The TL series of op amps proves value for money characterised by its lower slew rate and bandwidth The National Semiconductor LM617X shows much more superior specifications to the OPA627 at a fraction of the price The sonic differe
44. l Bypass Accepts 48V Ground Lift 2 2 EAST STMP 01 RRP 216 40 Technical Specifications http www east uk com Features 6dB gain for output jack Bass 12dB XLR line Mid Bass Boost only 50Hz 1848 Audiodomain blogspot co uk 46 homewith dave daveisnot outlook com Treble Footswitch for Mid Frequency Ground Lift Footswitch Full Bypass Bright Switch 8dB 7Hz 2 3 EBS MICROBASS 2 RRP 286 Features Mute Footswitch Boost Drive Bass Treble Mid w variable frequency Volume FX Mix Tube Sim GND Lift Parallel Output Headphone Output Enhancer Filter Bright Switch FX Send Return Speaker Sim 2 4 RADIAL BASSBONE Treble 12dB 12dB 1K Mid with Variable Frequency 1288 150Hz 3KHz http www bass se 2 009 microbassll htm Technical Specifications Frequency 0 3 4820 20k Hz Tone Controls Bass 15 dB 100 Hz Treble 15 dB 10 kHz shelving Bright 12 dB 10 kHz 15dB gain Midrange 12 dB 50 2 000 Hz Q 0 80 Impedances Input Impedance A 10M Input Impedance B 1M Output Impedance 10K RRP 245 Technical Specifications Audiodomain blogspot co uk homewith dave daveisnot outlook com 47 Features http www tonebon e com tb bassbone Boost Contour Shape 1 features htm Bass 45dB 110Hz Mid 5 1dB 650Hz Treble 2 7dB 4 3KHz Level 1 Level 2 Contour Shape 2 G
45. meter A modern set of filters could be wired to the boost half of a pot The preamp was originally designed to operate under guitar pedal parameters including designs for running off a 9v battery The process of splitting the supply and biasing the circuit puts a lot of limitation on redesigns Whether adopting virtual ground integrated circuits or transformers from mains voltage forming dual supply lines would remove any biasing and decoupling issues in the current design This would change the preamp from class A to class B but wouldn t necessarily have any detrimental effect on the sonic performance of the preamp A class B push pull system may compromise the good THD N performance due to cross over distortion but given the frequency content of the bass signal it is unlikely to be audible in the output The dual supply would allow more efficient running of the op amps using all of the available transistors inside the amplifier This more efficient use of power could improve battery life a specification that is poor in the current design There is also an option of combining the above solutions to create a class AB system this is often the preferred method in audio amplifier design Adding small bias current to the input of a transistor will move the bias point above O so crossover distortion is greatly attenuated The transistors still share the work so energy isn t dissipated as heat The power consumption of the circuit needs to be further anal
46. method is included in appendix 3 Figure 1 shows the peak output voltage of 3 bass guitars It is noted that the following readings are worst case scenario measurements They were recorded with significant attack on the bass Fender Precision Bass Split Coil Humbucker 0 5189 strings Fender Jazz Bass 2 x Single Coil 0 3894 Warwick P J Split Coil Humbucker Single Coil 0 3567 Figure 1 Peak Voltage of 3 bass guitars The amount of gain needed to bring instrument level to line level can be calculated from the following equation Vout ABgain 20 log T in Where Vin Instrument Level 0 6V Vout Line level 1 737Vp ABgain 9 232 have rounded the largest reading up to 0 6V as an absolute worst case scenario reading to cover the many different bass guitar designs on the market As active basses will produce significantly higher output swings than passive basses attenuation is needed at the input stage Audiodomain blogspot co uk 8 homewith dave daveisnot outlook com 2 2 2 ATTENUATION An attenuator or pad is an arrangement of non inductive resistors in an electrical circuit used to reduce the level of an audio or radio frequency signal without introducing appreciable distortion Ballou 2005 It is commonplace to add a switchable 20dB pad to account for active basses Figure 2 shows an unbalanced L type attenuator pad R1 series R2 shunt Figure 2 L Type Attenuator Chinn R n d online image
47. mf files static mcgrawhill images 9780071382014 f0319 01 jpg Accessed 26 December 2011 Baker B n d Select the Right Operational Amplifier for your Filtering Circuit online Available at lt http ww1 microchip com downloads en devicedoc adnOO3 pdf Accessed 09 December 2011 Kuphaldt T n d online image Available at http openbookproject net electricCircuits Semi 03135 png Accessed 14 April 2012 Kuphaldt T n d Lessons in Electric Circuits Volume Ill Semiconductors online Available at http openbookproject net electricCircuits Semi SEMI 4 html xtocid722313 gt Accessed 14 April 2012 Orman J 2008 Pull down Resistors online Available at http www muzique com news pulldown resistors Accessed 14 April 2012 Metric Halo n d Spectrafoo User s Manual P25 Image In Metric Halo 2012 Mobile IO User s Guide p165 Audiodomain blogspot co uk 43 homewith dave daveisnot outlook com Robjohns 2012 Non Standard Cables online Available at lt http www performing musician com pm febOS8 articles technotes htm Accessed 17 April 2012 Kuphaldt 2007 online image Lessons in Electric Circuits Volume AC Available at http www allaboutcircuits com vol 2 chpt 8 3 html Accessed 14 April 2012 Texas Instruments 1978 TLO7X Specification sheet online Available at http www ti com lit ds symlink tlO71 pdf gt Accessed 14 April 2012 Carter
48. mlink tlO71 pdf Accessed 24 December 2011 Ramsey F amp McCormick T 2009 Sound and Recording 6 ed p348 Focal Press Rutkowshi G 1993 Operational Amplifiers Integrated and Hybrid Circuits p26 Wiley Blackwell National Instruments n d online image Available at http www ti com lit an sloa058 sloa058 pdf Accessed 28 December 2011 National Instruments n d online image Available at http www ti com lit an sloa058 sloa058 pdf Accessed 28 December 2011 Horowirz W amp Hill P 1989 The Art of Electronics 2 ed p177 Cambridge University Press Horowirz W amp Hill 1989 The Art of Electronics 2 ed p179 Cambridge University Press McMannus S et al 2005 Handbook for Sound Engineers 34 ed P651 Focal Press Horowitz W amp Hill P 1989 The Art of Electronics 2 ed P29 Cambridge University Press Image In Horowitz W amp Hill P 1989 The Art of Electronics 29 ed P36 Cambridge University Press Image In Horowitz W amp Hill P 1989 The Art of Electronics 2 ed P36 Cambridge University Press Image In Horowitz W amp Hill P 1989 The Art of Electronics 29 ed P37 Cambridge University Press Image In Horowitz W amp Hill P 1989 The Art of Electronics 29 ed P38 Cambridge University Press Storr W n d Electronics Tutorial about Passive Low Filters online Available at http www electronics tutorials ws fil
49. n blogspot co uk 35 homewith dave Figure 40 600Hz Cut Figure 41 1kHz Boost 36 Audiodomain blogspot co uk homewith dave Sete Senn ICT Figure 42 1kHz Cut When connected in parallel the 100Hz and 250Hz filter show a lower Fc This suggests that the filters are loading each other lowering Fc The unconnected filters tend to perform better than the connected ones the known issues are still present but look less extreme and smoother This may offer another partial explanation for poor filter performance when summed back to the original signal John Bohn offers a potential solution to increase summing performance A technique used almost universally to improve the combined responses of graphic equalizers involves two series summing circuits This way two adjacent bands are not added together by the same summer Bohn J 1986 This creates physical separation of filter modules and reduces the chance of interaction between them Figure 43 is a graphical representation of this BOOST SUMMER 2 FT ep i 25Hz lt 1 Figure 43 Dual Summing Filters Bohn J 1986 5 4 TOTAL HARMONIC DISTORTION NOISE Audiodomain blogspot co uk 37 homewith dave daveisnot outlook com Total Harmonic Distortion Noise THD N measures harmoni
50. nce of these Op Amps is subjective The quicker slew rate of the LM617X and OPA627 will reduce high frequency distortion That said the TLO7X data sheet states that harmonic distortion is at 0 00396 human hearing is not sensitive enough to recognise this This is also common THD N reading for audio products working at 20Hz 20kHz 4dBu Therefor appropriate for an audio circuit 2 3 2 OP AMP RULES Op Amp functionality is based upon the following rules that must be recognised when designing with Op amps 1 Opamps will always attempt to make the difference between inverting and non inverting inputs zero 2 Theinput of an op amp draws no current Horowitz amp Hill 1989 2 3 3 AMPLIFICATION There are two main methods of amplification with op amps Inverting and non inverting Non Inverting Amplifier Figure 5 Figure 5 Non Inverting Amplifier National Instruments n d online image The signal is fed into the non inverting input A feedback loop is sent to the inverting input The voltage divider controlled by R1 and R2 controls the amount of gain Gain is calculated by ER ain g R The Op amp tries to make the inverting input the same as the non inverting As the signal first passes through a voltage divider the output is raised to compensate for the voltage loss Audiodomain blogspot co uk 11 homewith dave daveisnot outlook com INVERTING AMPLIFIER FIGURE 6 Min Figure 6 Invert
51. of MFBP filter 5 Gl S R IR C S2R C 1 2 Where Transfer Function s Complex Variable representing Amplitude ad Phase Williams A amp Taylor F 1995 Appendices 10 show equations and design parameters essential to component values and filter stability for the MFBP topology Audiodomain blogspot co uk 15 homewith dave daveisnot outlook com 2 4 5 Boost CuT TOPOLOGY Adjustable gain is a common and expected parameter for filters This can be achieved by adopting a circuit patented by Kenneth James of Dolby Laboratories The band pass filter becomes isolated from the boost and cut summing amplifiers Figure 15 is a full schematic layout of this topology R1 R1 C1 Boost Summer R2 R3 Filter Gain vi C2 Figure 15 Boost Cut topology Brundy K 1975 This system is based on the phase relationships between the filters and the summing amplifiers Dennis A Bohn correctly states that at the Boost Summer U2 Boost Output 1 kBP And at the Cut Summer U1 Boost Output 1 kBP Where 1 Normalized Full Frequency Input Audiodomain blogspot co uk 16 homewith dave 2 Band Pass Signal Rane Coperations Bohn D 1986 All of the op amps are inverting so when the signals are summed their phase interacts The signal summed to U1 causes phase cancellation at the pass band The signal sent to U2 causes phase
52. of the op amp so the input transistors are perfectly balanced Operational amplifiers can vary significantly operation and quality The following sub sections look at the individual characteristics and their suitability for an audio application SLEW RATE Although not directly describing op amps Ramsey and McCormick define slew rate as The ability of an amplifier to respond to high level transients Ramsey amp McCormick 2009 Slew rate is represented in V us Volts per microsecond Insufficient slew rate may cause distortion at higher frequencies due to their inability to react quickly enough to a sudden change in voltage Using the calculations stated in Appendix 5 the theoretical minimum slew rate for an audio Op Amp 0 43V us Realistically a much higher slew rate would be implemented as the sharpest of transients cannot be accounted for Consideration must be paid to the slew rate and bandwidth in audio circuitry UNITY GAIN BANDWIDTH George B Rutkowski defines the perfect bandwidth of an opamp BW oo An infinite bandwidth is one that starts and dc and extends to infinite cycles second Hz Rutkowski 1993 COMMON MODE REJECTION RATIO CMRR The CMRR defines the rejection of common signals at 2 inputs of operational amplifiers For example a good op amp would reject DC offset by a ratio upwards of 100dB This would be of importance in a microphone preamp design where the input is balanced and a differential amplifier
53. oltage to give acceptable headroom for the given application The 9v DC regulated power supply is a very popular supply for supply guitar pedals Using a Parallel filter bank and series DC jack links it is possible to power 30 pedals from a single mains supply Audiodomain blogspot co uk 17 homewith dave daveisnot outlook com 2 5 2 DC REGULATION Referring to the linear DC regulator the basis of regulating power supplies Marty Brown states It relies on the variable conductivity of an active electronic component to drop voltage from an input voltage to a regulated output voltage Brown 2001 This can prove inefficient as power is lost across a DC regulator in the form of heat Despite this regulated supplies show good performance at producing a clean voltage source for sensitive circuits to run off As audio circuits can be susceptible to external noise the DC regulated power supply is seen as the most efficient way to supply power from mains 2 5 3 SPLIT POWER RAIL A split power rail is used to create a DC bias when using a single power rail As seen in Figure 18 a voltage divider is used to half the supply voltage to properly bias the Op amp R1 100 100 kt Figure 18 Split supply coupling National Instruments n d online image Op amps need to swing between a positive and negative rail with reference to ground In the above example a virtual ground is created at Vcc
54. r amplifiers The line level output can be used if the input stage of the amplifier has a pad With an 18v supply the output can be as high as 11 5004dBu and at 9v as high as 3 70dBu Both levels are high an expected instrument signal 5 2346dBu Audiodomain blogspot co uk 39 homewith dave daveisnot outlook com 6 6 1 FURTHER WORK In order to improve noise performance a review of the use of molex connectors is necessary Once all improvements to the circuit were made having a PCB etched would make the entire circuit more stable The circuit wouldn t necessarily need to be placed on one board a motherboard with all processing could be connected to a separate board for controls This is often seen in audio design and an idea that could easily be adopted in the preamplifier design A simple phase lead capacitor in parallel with the gain stages could be installed Paying more attention to high frequencies outside of the range of human hearing would bring the circuit more in line with standard practice within professional audio design These capacitors are unlikely to have any impact on the tone of the preamp as they will produce cut off points above 20kHz It would however give the op amp more protection against noise interference and op amp instability Figure 44 shows C and R2 forming a low pass filter around a negative feedback network Vp Feedhack Lead Compensation Feedback network bts Figure 44
55. rowl 1kHz filter is the strongest and most versatile Audiodomain blogspot co uk 32 homewith dave filter This is clearly evident in the frequency response graphs and the sonic output of the circuit When boosted the sound is opened up with a considerable amount of string and fret noise amplified a good characteristic of a rock bass sound When cut a vintage tone becomes available With significant attenuation in the high frequencies the sound becomes well rounded deeper with a tone that would be found in soul funk and blues music 5 3 3 INTERACTION The frequency response of the summed signal was tested with all four filters connected in parallel and with unused filters disconnected This was done to observe the interaction between filters Figures 35 42 show these graphs Black No EQ Connected Red Connected in Parallel Blue Other filters Disconnected N 3 Figure 35 100Hz Boost Audiodomain blogspot co uk 33 homewith dave 42 N O O ka 2 E LL T 4 1 2 Figure 37 250Hz Boost 34 Audiodomain blogspot co uk gt ol O gt 2 5 Qi c e O gt BI O Figure 38 250Hz Cut Figure 39 600Hz Boost Audiodomai
56. show that there is variation above and below unity between the four filters In the literature review the multi feedback band pass filter component values were calculated It is seen here that significant rounding of ideal Audiodomain blogspot co uk 28 homewith dave 2 values were needed to meet standard resistor and capacitor values Furthermore some of these values were not available at the time of construction so further compromising was needed Combined with component tolerances this offers reasonable explanation to this As the passive band pass networks were built on a separate board it was possible to swap out different filters for testing After several tests with various different filters the original design was used as it provided the most stability and performance 5 3 2 SUMMED FILTER RESPONSE Figure 28 13 show the frequency response graphs of the filters summed to the original signal Black No EQ Connected Blue 12dB Boost Red 12dB Cut N 3 14 JAN See Figure 28 100Hz Audiodomain blogspot co uk 29 homewith dave 22 Solo Figure 29 250Hz Figure 30 600Hz 30 Audiodomain blogspot co uk homewith dave daveisnot outlook com Sete SHAN See Figure 31 1kHz A common issue with all the filters is
57. spot co uk 20 homewith dave daveisnot outlook com 41 kumu ERN gt gt F TLO7X Various combinations of TLO71 72 and 74 nai MEN we rr 50 00000000000 Figure 20 Initial Schematic Bill of Materials Audiodomain blogspot co uk 21 homewith dave daveisnot outlook com 4 4 1 MODULAR DESIGN The circuit was initially to be built on one piece of strip board this would make troubleshooting the circuit difficult In order to maximise the efficiency of testing and troubleshooting the circuit was designed in a modular fashion The circuit was split into nine separate boards power input input pad input gain filter summers filter amps passive filter networks output gain and output Appendix 15 shows photos of the individual boards Removing modules from the design became crucial when testing the preamp it made examining the interaction between boards more practical and offered the opportunity of experimentation The modules were built on 2 54mm spaced strip board proving useful for prototyping as well as containing permanent components The filter modules and gain stages due to their increased complexity were first built and tested on breadboard Tracks were broken using a drill bit this helped maximise the space available on the board The modules were connected through molex connectors and terminals This solution offers a stable electrical connection with t
58. ter filter 2 html Accessed 16 December 2011 Lancaster D 1996 Active Filter Cookbook 274 ed p3 Newnes Audiodomain blogspot co uk 42 homewith dave daveisnot outlook com Elliot R 2000 online image Available at lt http sound westhost com project64 htm gt Accessed 11 December 2011 Williams A amp Taylor F 1995 Electronic filter Design An Introduction p5 43 Mcgraw Hill Brundy K 1975 Adjustable Equalizers Usable in Audio spectrum United States Patent 32201104 Coperations Bohn D Constant Q Graphic Equalizers 1986 Vol 34 1st April 1979 4 5 Elliot R 2000 online Available at http sound westhost com project64 htm Accessed 11 December 2011 Elliot R 2000 online image Available at lt http sound westhost com project64 htm gt Accessed 11 December 2011 Brown M 2001 Power Supply Cookbook 274 ed p11 Newnes National Instruments n d online image Available at lt http www ti com lit an sloa058 sloa058 pdf gt Accessed 28 December 2011 Whitlock B 2005 Sound Engineer s Handbook 3 ed p233 Focal Press Image In Talbot Smith M 2001 Audio Engineering Explained 274 ed p25t Focal Press Kuphaldt T 2007 Lessons In Electronics Vol online Available lt http www allaboutcircuits com vol_ 2 chpt 9 2 html Accessed 09 December 2011 Education com n d online image Available at http 01 edu cdn co
59. the amount of current flowing into a node physical connection has to flow out of that same node Using Figure X we can state h tL l 1 5 Figure X Kirchoff s Current Law Education com n d online image 1 3 5 VOLTAGE LAW Kirchoff s Voltage Law stages that the sum of voltages within a closed loop circuit must equal the voltage Source Vource sss P 1 4 TH REVIN S EQUIVALENT CIRCUIT Audiodomain blogspot co uk 45 homewith dave daveisnot outlook com Any network of resistors and voltage sources can be made equivalent to 1 voltage source and 1 resistor This theorem is used heavy to simplify complicated resistive networks It is applied heavily in modular designs for distinguishing the ability of different sections to work together in a circuit 1 5 SERIES RESISTANCE Rtotal TS 1 6 PARALLEL RESISTANCE 1 1 1 1 1 R2 2 COMMERCIAL PREAMP SPECIFICATIONS 2 1 TECH 21 SANSAMP BASS DRIVER RRP 165 Technical Specifications http www tech21ny c com products sans Features Presence Brings out upper harmonic amp bassdriverdi ht content Level Drive Controls Overdrive and Gain Blend Treb bass 12dB Treble Blends between unaffected and effected Dass signal Drive Impedances Presence Input 1M Line Instrument Outputs Both XLR and Jack Output Low Z Parallel out Footswitch Ful
60. tuent resistors but also on how much current is being divided away from it through a load The base emitter PN junction of the transistor is a load that decreases the DC voltage dropped across Kuphaldt T n d The DC voltage dropped across R3 would upset the balance of the voltage divider giving a lower than expected bias voltage This could be the cause of the low input bias at the input of the buffer This issue would only become a problem if a lower value supply rail were used At the input the bias voltage is around half of its intended value A 5v supply rail would only produce 1 25v bias A strong bass signal would clip the supply rail When working with a 9v supply the bias voltage is 2 25v which is more than enough to pass a clean signal through the buffer stage If it wasn t for the worst case scenario planning this may have been an unavoidable problem The pad power and filter modules utilise a double pole double throw DPDT switch to change between circuits When operated there is a voltage spike at the output which creates a popping sound Jack Orman offers and explanation for this problem Capacitors will leak a tiny voltage which will build up on the input or output as a small voltage potential When the pedal is switched on the voltage is discharged into the signal path Orman J 2008 By adding pull down resistors at various positions around the DPDT switches this rogue DC voltage could be dropped from the
61. ument and line inputs commonly have input impedances of around 1M Because the instrument output requires more attenuation a larger output impedance had to be factored into the design As long as the output impedance does not exceed around 100K it should have no effect on output bridging 13 605k and 2127 Q output impedances are ideal and found amongst the industry standards Audiodomain blogspot co uk 38 homewith dave daveisnot outlook com 5 7 CURRENT DRAW Appendix 27 shows the technique for measuring the current draw With the preamp in it s quiescent state the circuit draws 32 1mA from the power supply With the preamp fully amplifying a 1 2Vpp signal the circuit draws 32 9mA from the power supply Appendix 28 is an excerpt of the TLO7X data sheet it states the maximum current draw from the amplifier is 2 5 13 amplifiers are used this gives a theoretical quiescent maximum state current draw of 32 5mA Based on a 550 mAh 9v PP3 battery the following lifespans have been calculated With the preamp in its quiescent state the battery will last 17 1339 hours and running at full capacity at 16 71713 hours These are acceptable values but by no means are they exceptional For example the Sansamp Bass Driver consumes only This is incredibly low and would result in 91 6 hours battery life All 13 op amps in the circuit are biased at half the supply rail this keeps the transistors inside the op amps turned on all the time Class A
62. us 2005 Appendix 7 includes filter terminology referenced throughout this section 2 4 1 REACTANCE In order to change the frequency response of a signal linear components sensitive to a change in frequency are needed Linear components don t change frequency but they do dictate frequencies contained within a signal in higher terms tonality Horowitz and Hill state The output of a linear circuit driven with a sine wave at some frequency f is itself a sine wave at the same frequency with at most changed amplitude and phase Horowitz amp Hill 1989 Linear components such as capacitors and inductors have frequency dependent resistance known as reactance Reactance is treat similarly to resistance but as it is frequency dependent calculations become more complex Calculations are found in Appendix 8 In order to manage the complexity of reactance the term impedance is used to represent the overall resistance of a network Ohms and Kirchoff s laws still apply but all terms need to be referenced as impedance and not resistance 2 4 2 PASSIVE FILTERS HIGH PASS Figure 9 shows a typical High pass Circuit a voltage divider formed by a capacitor and a resistor Figure 9 Passive first order High Pass Filter Image In Horowitz W amp Hill P 1989 Using the following equation the voltage output at a specific frequency can be determined See appendix 9 for complete equation list 2nf RC Vout Vin
63. uses problems for creative equalisation In order to address these problems we can apply passive components to an active network 2 4 3 ACTIVE FILTERS Active filters have long been the standard in audio instrument design Lancaster states Audiodomain blogspot co uk 14 homewith dave daveisnot outlook com An Active filter is some combination of integrated circuit operational amplifiers resistors and capacitors that does things that normally could be done only with expensive inductor capacitor passive filter combinations Lancaster 1996 Using capacitors in tandem with negative feedback filters is used to mimic the operation of inductors These gyrator networks bring the best of active and passive filters together 2 4 4 MULTIPLE FEEDBACK BANDPASS FILTER TOPOLOGY Figure 14 shows Multiple Feedback Bandpass MFBP filter configuration credited to L Huesman Figure 14 uses an inverting amplifier with two feedback loops one with high pass circuitry and the other with low pass MFBP filters are popular in filter design for their flexibility tolerance to sensitivity errors in components and can be tuned to almost any shape with a quality factor under 20 CS 100K C GHD GHD 4 Figure 14 Multi feedback implementation Elliot R 2000 online image The gain is controlled through the R1 R2 voltage divider C1 C2 and control the pass band Williams and Taylor state the transfer function of this design
64. voltage E S LI INN ENS Common mode input voltage range dis VICR 15 15 10 KQ 12 135 12 135 Maximum peak output 10 voltage swi in gt 2 sey ase LSS ae votage ampicatin 0 40 Mi Lp rejection ratio supply current each amplifier Crosstalk attenuation 29 MAXIMUM GAIN READINGS Without EQ With EQ With EQ 18v di 24 6272 392 36 9432 Kal 70 24 734 11 K Sa 30 MAXIMUM GAIN CALCULATIONS The input signal was a 1 2Vpp 1kHz sine wave This reading includes the 12dB of gain offered by the 1kHz boost 30 1 18v MAXIMUM LINE LEVEL OUTPUT dB AB 2010410 V 1 14 AB gain 2010410 12 AB gain 21 4621 Audiodomain blogspot co uk 71 homewith dave daveisnot outlook com dBu 14 2Vpp 7 1Vp Vrms 7 1Vp x 0 707 5 0197 dBu 201 5 0197 u 2 0010 70 775 16 2275 30 2 9v MAXIMUM LINE LEVEL OUTPUT dB 5 76 AB gain 2010410 12 dBgain 13 6248 dBu 5 76Vpp 2 88Vp Vrms 7 1Vp X 0 707 2 036 20155 2036 u 9910 0 775 dBu 8 3895 30 3 18v MAXIMUM INSTRUMENT dB 8 24 AB gain 2010410 12 dB gain 16 7349 dBu 8 24Vpp 4 12Vp Vrms 4 41Vp x 0 707 2 20584 abu 201 2 20584 u 25 0010 70 775 11 5002
65. yzed The circuit performs fairly poorly in the respect Given the timescale of the project it was always likely that the circuit wouldn t be as efficient as desirable It of my personal desire to include reamp functionality into the input of the preamp This would lead to supreme versatility within the studio environment The inclusion of insert sends and returns is also a likely inclusion in further designs This would be beneficial for using third party compression units or other processing 6 2 CONCLUSIONS Overall the preamp is versatile and stable across all available parameters offering acceptable sonic output The circuit is capable of accepting any instrument signal and outputting to any line or instrument level device The results of electrical testing are representative to the sonic output of the preamp The project has been a success on an academic and personal level Further practical work and redesigns have continued during the writing of this report and will continue into the future Audiodomain blogspot co uk 41 homewith dave daveisnot outlook com REFERENCES Barlett B amp Barlett J 2002 Practical Recording Techniques 3 ed p536 Focal Press Ballou 6 et al 2005 Sound Engineer s Handbook 3 ed p631 Focal Press Chinn R n d online image Available at http www uneeda audio com pads Accessed 27 December 2011 National Instruments n d online image Available at www ti com lit ds sy
Download Pdf Manuals
Related Search