3Way Tone Control
High quality analog tone control. First class components make this tone control stand out from others. It uses 1 % polypropylene and 1 % silvered mica capacitors for filtering and potentiometers with only 10 % tolerance. Standard potentiometers have a tolerance of 20 %. This makes the design not low-cost of coarse. Added to bass and treble is a third potentiometer to adjust the mid range also. A PCB is designed with some customized footprints for the filter capacitors and potentiometers used.
| Bandwidth (P1/P2/P3 set to 0 dB) | 3.7 Hz…330 kHz (load = 10 kΩ) |
| THD+N | 0.0012 % (B = 22 kHz) |
| 0.0017 % (20 Hz…20 kHz, B = 80 kHz) | |
| Signal to noise ratio | > 98 dB (B = 22 kHz) |
| > 95 dB (B = 80 kHz) | |
| Channel separation | > 97 dB (1 kHz) |
| > 71 dB (20 kHz) | |
| Tone control (each individual) | ±14 dB (20 Hz) |
| ±13 dB (20 kHz) | |
| ±9 dB (1 kHz) | |
| Max. output level (P1/P2/P3 set to 0 dB) | 7.9 VRMS (THD = 0.1 %) |
| Supply current (no signal) | ±20 mA |
| Maximum Supply Voltage | ±17 V (LME49720) |
Schematic

Fig. 1. Schematic of the 3Way Tone control (260174-1, v1.0).
Many versions and variations of tone controls have been designed and published over the years since the original publication of P.J. Baxandall in Oktober 1952 in Wireless World (Negative-Feedback Tone Control). The added value of this design is: the use of high quality 1 % capacitors with dielectrics polypropylene and silvered mica, low tolerance potentiometers, an additional mid range control, excellent audio opamps (a dual-opamp per channel) and of coarse 1 % metal film resistors. Standard potentiometers have a tolerance of 20 %. The Long Life Cermet Potentiometer series P11L from Vishay Sfernice, lifetime 2 Million Cycles, is available with a 10 % tolerance. The datasheet even mentions 5 %, but I haven’t seen these anywhere.
The most important effect on analog signals in analog filters is the quality and properties of the capacitors used. Test circuits can reveal distortion caused by capacitors. Ceramic and electrolytic are the worst because of many negative properties. Generally standard plastic film capacitors almost all are polyester and perform well enough in many applications. But, for high end audio polypropylene, polystyrene, (silvered) mica and PTFE (a plastic largely known from trade names like Teflon and Tefal) are far better choices. The quality of plastic film capacitors goes hand in hand with the dielectric constant. The higher the quality the lower the dielectric constant of the plastic. Downside of this, the higher the quality the larger the capacitor. The larger size is of little consequence for audio frequencies. Polypropylene capacitors are often used in high voltage pulse applications. The KP1830 series from Vishay Roederstein is available in 100 VDC versions at various distributors. But curiously, the rated value of 100 VDC is missing from the datasheet, only 63, 250 and 630 V types are listed. However, all capacitors in this series have the same size, a lead spacing of 5 mm, and this appears to limit the maximum capacitance to 22 nF (a 63 VDC version only, according to the datasheet).
The design uses one 27 nF capacitor per channel (C1 and C12) from a different series, 715P from Cornell Dubilier. According to the datasheet the 100 V version with straight leads has a spacing of 10 mm, but the ones I got measure close to 12 mm, more than the 1.5 mm tolerance. Slotted holes for these capacitors to make sure they will fit smoothly and thanks to two extra normal pads 5 mm and 7.5 mm capacitors from different series will fit also, just in case.
Dual gang potentiometers with low tolerance are hard to find. Although datasheets list 5 % versions, I yet have to find a distributor that has these on stock. Sometimes they are listed as non stocked and they can be ordered. Generally this means a minimum (large) order quantity must be purchased. Looking for ones that are on stock is the next alternative. As a result, the types I found and used here have solder lugs (leads Y00, the datasheet lists no less than 31 types). To fit them on the PCB the footprint has slotted holes. But the solder lugs are not that different from the pins of the PCB mounted versions, just wider and a little shorter (see next photo). Except for the extra two pins to mechanically secure it to a PCB, otherwise that version would also fit.

Fig. 2. Two potentiometers: conductive plastic with PCB pins and cermet with solder lugs.
It is best to mechanically fasten the three potentiometers to the front and the PCB to a metal enclosure to prevent stress on the pins and solder joints. Each potentiometer is supplied with a nut and spring washer.
Opamp
The PCB is designed for through hole components. There are many high quality audio opamps that would be a good choice for this circuit. I chose a LME49720 for its excellent properties and overall good appreciation in the audio community and its availability in a DIP-8 package (status is still active)
Some of its properties, Gain = 1, Vout= 3VRMS, fIN = 1kHz, RL = 600 Ω:
extreme low THD+N, 0.00003 %
extreme low IMD, 0.00005 %
high slew rate, 20 V/µs (typ.)
Gain bandwidth, 55 MHz
Input bias current, 10 nA typ., 72 nA max.
Of coarse the choice is a personal one and many other opamps will work as well, even an oldie but goody like the NE5532 (although bias current is a bit high with 200 nA typical), as long as they are unity gain stable. More modern ones often are produced in SO-8 packages only (largest version) or even smaller SSOP packages, like for example by the manufacturer suggested newer version OPA2891. Small adapter PCBs must be used then. The use of DIP-8 sockets is a choice, some prefer soldering the opamps directly onto the PCB.
PCB
The size of the PCB is 101.60 x 75.56 mm. The inputs and outputs are on the opposite side of the potentiometers. The large output capacitors (C9 and C20) and the in and output connectors (K1…K4) are inline and define the width of the PCB. The filter components are positioned in such way the crossing of tracks with different nets are avoided as much as possible. Of coarse this is not completely possible with the connections to the stereo potentiometers. The copper plane on the component side is connected to signal ground only and used as shielding. Power ground is routed separately on the bottom side to the electrolytic capacitors C23 and C24 and the screw terminal block K5 for the connection of the supply voltage and to the decoupling capacitors of the opamps. Close to the entry side of K5 is a non-plated hole in the PCB to route the power supply cables underneath the PCB is this preferred instead of having three wires run over the components. A small copper plane for the power ground is connected at the side of the in and outputs to the large signal ground plane.
Assembling the PCB
The potentiometers should be soldered last, after soldering the other components. Start with the lowest first and only then solder the potentiometers while making sure their support plates are exactly inline (front of the three potentiometers). One way to do this is clamping them all three upside down in a vice. This way when placing the tone control in an enclosure there’s no stress on any of the solder joints and pins (solder lugs) of the potentiometers, when they are fastened to a front panel. Place the PCB over the pins and flat against the potentiometers and solder all 18 leads. The distance between the center of the potentiometer shafts is 30.48 mm (1200 mil). This distance limits the maximum size of the knobs that can be used.
Measurements
To show the characteristics the following plots were measured with the potentiometers set to neutral, minimum and maximum positions and various combinations.

Fig. 3. Plots showing amplitude with bass and treble both set to max and min and only the midrange set to max and min.


Fig. 5. Bass and midrange set to max and min with treble to 0 dB. Midrange and treble set to max and min with bass set to 0 dB.

Fig.6. Bass set to max and min with midrange and treble set to 0 dB. Treble set to max and min with bass and midrange set to 0 dB.

Fig. 7. FFT of 1 kHz and 1 V input signal. No harmonic distortion visible.
Figure 3 shows the amplitude with bass and treble both set to maximum and minimum (red and blue) while midrange is set to 0 dB and midrange set to maximum and minimum (green and magenta) while bass and treble both are set to 0 dB. The maximum and minimum level of the midrange control is intentionally less adjustable, otherwise the influence on the other two controls would be too large and 9 dB should be more than enough. Furthermore, the effect of the midrange control can be enhanced by adjusting bass and treble to the opposite setting (see next figure). The cyan plot shows the amplitude with all three controls set to 0 dB.
Figure 4 shows the amplitude with all controls set to maximum and minimum (red and blue) and bass and treble set to maximum and minimum while the midrange control is set to the opposite level (green and magenta). The latter setting of the controls makes the midrange control a little more selective (compare this with figure 3).
Figure 5 shows the amplitude with bass and midrange both set to maximum and minimum while treble is set to 0 dB (red and blue). Also shown is the amplitude with midrange and treble both set to maximum and minimum while basis is set to 0 dB (magenta and green).
Figure 6 shows the amplitude with bass set to maximum and minimum while midrange and treble both are set to 0 dB. It also shows the amplitude with treble set to maximum and minimum while bass and midrange are set to 0 dB.
Figure 7 shows an FFT of a 1 kHz 1 VRMS input signal with all controls set to 0 dB. With proper shielding of the PCB the mains frequency of 50 Hz and power supply ripple of 100 Hz are hardly visible in the noise floor. Also, no harmonics are visible. At an input level of 3 VRMS THD+N at a bandwidth of 22 kHz is less than 0.0006 %
Of course the amplitude levels shown in figures 3 to 6 depend on the tolerance of the potentiometers. Before fitting them on the PCB I measured the resistance of both channels of each potentiometer. All were within the specified tolerance of 10 %. The resistance of both channels of two potentiometers were pretty equal, around 46 to 47 kΩ. However, the third potentiometer showed a greater difference between the two channels, 47 and 51 kΩ. Butt this is still within its specification. This is a common problem with all analog tone controls that use potentiometers. The only way to avoid this is to replace the potentiometers by resistor dividers and use special/custom switches or circuitry, like analog multiplexers.
Bill of materials
Resistor
R1, R19 = 100 Ω, 600 mW, 1 %
R2, R3, R4, R20, R21, R22 = 10 kΩ, 600 mW, 1 %
R5, R6, R10, R11, R15, R16, R23, R24, R28, R29, R33, R34 = 270 kΩ, 600 mW, 1 %
R7, R25 = 68 kΩ, 600 mW, 1 %
R8, R9, R13, R14, R26, R27, R31, R32 = 3.9 kΩ, 600 mW, 1 %
R12, R30 = 47 kΩ, 600 mW, 1 %
R17, R35 = 47 Ω, 600 mW, 1 %
R18, R36 = 100 kΩ, 600 mW, 1 %
P1, P2, P3 = 50 kΩ, lin., stereo, 10 % (Vishay Sfernice P11L2V0FLSY00503KA)
Capacitor
C1, C12 = 27 n, 100 V, 1 %, polypropylene, LS 10 mm (Cornell Dubilier 715P27311NA3)
C2, C4, C13, C15 = 2.2 n, 100 V, 1 %, polypropylene, LS 5 mm (Vishay Roederstein KP1830222011)
C3, C14 = 3.3 n, 100V, 1 %, polypropylene, LS 5 mm (Vishay Roederstein KP1830233011)
C5, C6, C16, C17 = 15 n, 63 V, 1 %, polypropylene, LS 5 mm (Vishay Roederstein KP1830315061)
C7, C18 = 820 p, 500 V, 1 %, mica, LS 8.7 mm (Cornell Dubilier CD19FD821FO3F)
C8, C19 = 22 p, 500 V, +/-0.5 pF, 1 %, mica, LS 5.1 mm (Cornell Dubilier CDA15ED220DO3+XXB)
C9, C20 = 4u7, 450 V, 5 %, PP, LS 22.5 mm (Panasonic ECW-FD2W475JC)
C10, C11, C21, C22 = 100n, 50 V, ceramic X7R, LS 5 mm
C23, C24 = 220u, 35 V, Aluminum Polymer, 20 %, ESR 24 mΩ, D 8 mm, LS 3.5 mm (Yageo Kemet A759KR227M1VAAE024)
Semiconductor
D1, D2, D3. D4 = 1N4148, DO-35
IC1, IC2 = LME49720NA/NOPB, DIP-8
Other
K1, K2, K3, K4 = 2way pin header, straight, pitch 2.54 mm, contact gold
K5 = 3way screw terminal block, pitch 3.5 mm, 1.5 mm2
Optional, 2 PCB DIP-8 sockets
PCB

Fig. 8. Top overlay of the PCB of the 3Way Tone Control (260174-1, v1.0).

Fig. 9. Bottom overlay of the PCB of the 3Way Tone Control (260174-1, v1.0).

Fig. 10. Copper on top of the PCB of the 3Way Tone Control (260174-1, v1.0).

Fig. 11. Copper on bottom of the PCB of the 3Way Tone Control (260174-1, v1.0).
Fig. 12. Prototype of the 3Way Tone Control (260174-1, v1.0).
Fig. 13. Prototype of the 3Way Tone Control (260174-1, v1.0).

Discussion (0 comments)