Fullscreen HTML5Med Res (??MB)HTML4

AUDIO OUT AUDIO OUT L R By Jake Rothman The Transmanium fuzz box, part 1 Guitar electronics is a fringe activity, more of an art than a science. It’s all about musical feel and the creative neural feedback loop between the player and the speaker. Everything you have learned about in Hi-Fi audio design is almost irrelevant, if not despised, among the guitar fraternity. Present an Audio Precision curve at a guitar pedal convention, and you would likely to be booed off. Because of this attitude, fundamental electronic engineering errors occur and breakdowns are common. But there is a good side; a competent electronic engineer can diverge from traditional design and do something experimental. Impedance Impedance is one basic electronics parameter that does concern guitarists. Electric guitars came about in the age of valves, which are high-­ impedance devices; in fact, they act a lot like Mosfets. Consequently, guitars reflect this, using electromagnetic pickups typically wound with 5000 turns of 46 SWG (0.06mm diameter) wire, having an inductance of around 5H. You may recall this being discussed in the Electric & Bass Guitar Pickguards project article last month. To achieve a sufficient signal voltage 82kΩ 6.8µF Tant + Input –8.7V 15nF Negative feedback, no thanks! Another thing guitar electronics people have against modern audio design is its distortion curve, a result of high negative feedback and analog-to-digital conversion. This technology exhibits a consistently low distortion level as the level increases, ending with sudden hard clipping. For a good human interactive playing response, this is not acceptable. What is needed is soft clipping and a gradual increase in distortion, as is exhibited by acoustic instruments, 470Ω 68µF + 10kΩ Fig.2: the final clipping stage on the Silver Tone Bender. +9V 25kΩ Log 10pF 2x 1N4148 Volume OC75 –0.87V OC75 100nF 100kΩ 2.2kΩ Anti-log 470pF 470kΩ –776mV CW 15kΩ D1 100nF D2 –7.75V 220pF 100nF CW Input from drive circuit (same as yellow tone blender) Attack 6.8µF Tant + 0V 68 Fuzzy logic This high distortion characteristic loved by guitarists makes most Hi-Fi and studio engineers shudder, but since the 1960s, guitarists such as Jimi Hendrix took it to a whole new level. They made the distortion part of the musical instrument, with a distortion circuit called a fuzz box, taking the THD well over 300%. This is possible, where the sum of the harmonics is greater than the fundamental. An AP curve would be useless to characterise a fuzz box, just flat-lining at 100%. More useful are spectral analysis plots at different levels, but mathematics and the science of human perception mean little to guitarists, unless it’s a volume control that goes to 11. 1µF 8.2kΩ 18kΩ 33kΩ OC75 –9V 650µA such as a piano soundboard. This of course explains the preference of guitarists for valve amplifiers feeding stiff light-weight paper cone loudspeakers. It is said the first fuzz was created by a damaged (possibly deliberately ripped) speaker cone. Of course, every musically minded kid would stuff newspapers down the piano strings to get buzzy distortion or ‘fuzz’. I remember annoying my parents with it when around eight years old, possibly my first attempt at waveform manipulation! + Fig.1: the Solasound Pro Mk II Tone Bender circuit. The OC75 is a high-gain version of the OC71. 10kΩ The bias for the first transistor is a 10kΩ pull-down resistor, using the transistor’s leakage current. level, and to prevent the inductance from causing high-frequency signal loss, the guitar pickups need to be loaded with an impedance of 250kΩ or greater, easily provided by a valve input stage. This goes against modern audio design, with its low-noise op amps surrounded by low-value resistors for minimum noise. I’ve tried and failed to convince guitarists they should use low-­ impedance pickups with balanced XLR outputs. It’s just impossible to change 80 years of tradition. If you feed an electric guitar into a typical op amp circuit, with say 10kΩ input impedance, it sounds horrible: quiet, dull and hissy. So, 1MΩ input impedance it is, with unbalanced quarter-inch (6.35mm) jack plugs. 100nF Output to tone control BC549C 100kΩ 100Ω 0V Practical Electronics | November | 2025 Photo 1: the Pro Mk II Tone Bender, the first successful British fuzz box (still in production today!). This rebadged version was made for Marshall. [Editor’s note: in audio reproduction, distortion is undesirable because the output signal is not as the artist intended, while in music creation, it is a tool to achieve a desired effect.] Fuzz circuits and transistors The first commercial fuzzbox was an over-driven three transistor amplifier, the Gibson Maestro FZ-1 Fuzz-Tone, using RCA germanium 2N270s running off 3V. This hit the market in 1962. A very similar circuit (Fig.1) using Mullard’s highgain OC75 was introduced in the UK by Gary Hurst, the Color/Solasound Tone Bender Mk II in 1966, shown in Photos 1 & 2. It ran on a 9V PP4 battery giving longer sustain. This three-transistor format still forms the basis for many designs, including the one in this article. There were cheaper two-transistor topologies, such as the Fuzz Face, using Newmarket NKT275s, but there were excessive gain variations between units. The sound of fuzz changed when silicon transistors replaced the original germanium devices. It became brighter, more consistent and less temperature-sensitive. The Silver Tone Bender, designed by Dick Denny, used three silicon BC169 transistors, a BC109 die in a thennew cheap epoxy package. The clipping stage used two IN4148 diodes AC-coupled in a feedback loop between the base and collector, giving a clean, consistent buzz. The pedal is shown in Photos 3 & 4, while the clipping part of its circuit is shown in Fig.2. By the way, Dick Denny was most famous in the guitar world for designing the Vox AC30 valve amplifier, many variants of which are still built today. Positive ground Nearly all germanium transistors are PNP types, meaning some people preferred to use a positive ground with a negative supply rail. If the fuzzbox has its own battery, this is Photo 3: the Tone Bender is the classic silicon transistor fuzz box. I have built many! not a problem, but if it’s powered from a common power supply with other negative-ground devices, the power rail is shorted out. When guitarists assemble pedalboards with multiple effects (Photo 5), this often catches them out. If one can obtain germanium NPN transistors, this problem can be avoided. My favourite device is the Philips OC140 used in the Yellow TB, but it Photo 4: the Silver Tone Bender PCB. Note the back-toback ... clipping diodes at lower right. Photo 2: the Pro Mk II internals. Note the OC75 germanium transistors. It is still built on Veroboard to this day. Practical Electronics | November | 2025 69 Photo 5: a guitarist’s pedal board. Usually, all the effects Photo 7: notice the use of a special germanium NPN are fed off a single power supply; this is a recipe for hum OC140 along with black anti-leakage diode CG92 in the loops and other problems. Source: Mike Grindle. Tone Bender. is rare and expensive. Colorsound’s transistors are available (Photo 6) it sounds disintegrated, like two source came from a Malaysian and they have low leakage. They separate guitars. Mix controls are RAF spares store for V bombers via carried on making them a lot longer useful in bass fuzz boxes to allow Cyclops Electronics of York. than the West, so maybe they were the fundamental through. developed further. I decided to use silicon transisJunk box junctions tors for the first two stages to get I suspect that most older audio Circuit structure high gain, with the final clipping constructors have at least one old Fuzz circuits have always used stage being germanium. I used germanium transistor in their junk common-emitter stages for maxi- this approach when designing the box, and it is most likely to be a mum voltage gain, usually with a Colorsound Yellow Tone Bender PNP type, such as the audio classic drive control in the middle, a pas- for Macari’s (see Photos 7 & 8 and OC71 or AC126. Old radio-frequency sive tone control on the output and https://pemag.au/link/ac8m). I types can also be used, such as the a master volume on the end. There found that this topology gives the alloy-junction OC44, but they sound is also a foot switch controlled hard best of both worlds. too bright. Alloy-diffused RF types, bypass, which is important since the The input and driver stages were such as the AF117, are no use be- musician won’t necessarily want the the same as in the Silver Tone Bendcause of their high noise level. fuzz effect all the time. er, with the BC169s replaced by the Interestingly, Soviet germanium This design follows the typical modern version, the BC549. The block diagram shown in Fig.3. Some circuit is shown in Fig.4. Photo 6: this Soviet germanium fuzz circuits have a mix control The circuit for the Transmanium transistor had zero leakage current between clean and clipped signals. is shown in Fig.5. Its input stage is according to the Peak analyser. I don’t like this approach because a JFET common-source amplifier 1st gain stage 2nd gain stage Clipping stage High-pass and low-pass filters Tone control Master volume Drive or ‘Attack’ High-impedance Input Output 0V Fig.3: the basic configuration of a fuzz box. 0V +9V 1.2kΩ 470pF 470pF 100nF 470kΩ 470kΩ 33kΩ 100nF BC549C CW 100kΩ Log 470nF 680Ω 100kΩ 100Ω 100Ω 470nF 100Ω 3.3kΩ 220kΩ OC140 * CG92 * 100kΩ 6.8µF 4.7V 1nF BC549C Fuzz Input + 1µF + 10kΩ 15kΩ Fig.4: the Yellow Tone Bender circuit uses the same two input stages as the Silver Tone Bender but has a germanium clipping stage (TR3). 3.9kΩ 100nF * Germanium 33nF 22kΩ Lin 3.9kΩ Tone CW 100kΩ Log Volume CW Output 0V 70 Practical Electronics | November | 2025 R5 1.4mA 150Ω –8.5V 0V R1 4.7kΩ C1 10nF –9V 0.9mA C2 22µF + 2.3mA **Optional ‘tuning’ capacitor R3 6.8kΩ –4.5V TR1 2N5460 R2 2.2MΩ Green G to get a high input impedance. The second stage is practically the same. Its high input impedance minimises input stage loading by allowing a high-value pot to be used for the drive control. The final stage is an overdriven germanium transistor, which gives the characteristic sound. Leakage current A major problem with germanium transistors is base leakage current, which can cause the transistor to switch on excessively, even when unbiased. It is especially bad with old ‘junk box’ devices, so I designed this circuit to cope with it. In the Pro Mk II first stage, there was no bias voltage at all. Leakage current alone was 22nF** C3 47nF VR1 1MΩ or 470kΩ Log R4 1kΩ Top view 2N5460 R6 4.7kΩ 5:1 Red Drive Black –8V CW TR2 2N5460 470pF* –5V 18Ω Black + C4 100µF to 330µF 6V *Optional stability capacitor TR3 –0.5V NKT214F C7 4.7nF R8 22kΩ VR2 100kΩ Lin 0.9mA –0.7V (bias) R7 560Ω D1 1N4148 B Red dot C Top view NKT214F + C5 220µF C8 47nF Gain x22 CW VR3 10kΩ Log Tone CW Volume Output 0V E sufficient at the low levels, and the resulting distortion was a good thing. With the Yellow Tone Bender, I used a reverse-biased germanium diode (D1) across the base-emitter junction to cancel out the increase in leakage current with temperature. A modern, more effective way would be to use an op amp in a servo loop built around the germanium device to hold its output at the right point. Then again, if I did that, conservative guitarists would be up in arms. Biased transformer coupling I decided to go back to a 1960s way of minimising leakage current effects: using a transformer to couple the input. In this way, the source resistance of the bias network can be reduced from, say, 10kΩ to around 50Ω, so the leakage current flowing through it generates little voltage to bias the transistor on. The bias voltage is fed in series with the bottom of the transformer winding, so there is minimal loading, feeding all the power from the preceding stage into the output stage, TR3. The bias voltage is created by a forward-­biased semiconductor, so it provides temperature compensation at the same time. So long as 0.7V bias is not too high, a silicon diode (D1) can be used rather than an expensive germanium diode. The voltage vs temperature Photo 8: the relationship is a 1970s Yellow Tone Bender has a lovely, grungy sound. semiconductor Practical Electronics | November | 2025 C6 3.3µF + Fig.5: the Transmanium circuit. To make it negative ground, change the JFETs to N-channel 2N5457s and output transistor TR3 to a Input germanium NPN type, like the OC139. The diode and all electrolytics must also be reversed. physics law, and is the same for both types. It’s just the forward voltages that are different between them. Bias shift There are some more advantages of transformer coupling, such as no input coupling capacitor being required. Such a capacitor can charge up on loud tone bursts, causing a bias shift, switching the transistor off briefly. This effect sounds horrendous, punching holes in the sound. This happens because the DC resistance path is unequal in the positive and negative directions. The transistor’s base-emitter junction is a diode after all. Another way around this problem is to place a reverse-biased diode across the base-emitter junction of the transistor, equalising the paths. This was an additional function of the germanium diode in the Yellow Tone Bender. Interestingly, a bit of a bias shift in the right direction can enhance the sound, causing a form of pulse-width modulation. Finally, transformer coupling provides a high-impedance AC load for the drain of JFET TR2, while having a low DC resistance. This doubles the effective supply voltage, the drain being at around 8.5V rather than 4.5V. This boosts the gain of the relatively low-gain JFET. The transformer has to be a stepdown device, typically with a 5:1 turns ratio, to avoid the low input impedance of TR3 loading TR2, reducing the gain. Obviously, we don’t need an expensive Hi-Fi unit here. Any interstage or driver transformer used in cheap transistor radios will do the job. Phasing All the stages in the fuzz box are inverting, which means the system is inverting overall. The transformer wires on either winding can be 71 flipped to change the overall phase. If the transformer is wired to be inverting, the whole system will become non-inverting. Note that in the circuit diagram, the phasing of the transformer is shown by dots. Current flowing into the dotted end on one side will flow out of the dotted end on the other side and vice versa. Magic? Here’s where some strange things can happen. I tried putting a phase polarity switch on the transformer, and the sound was better in one position. Mike Grindle, who can actually play guitar (unlike me) noticed the difference much more and preferred the transformer in non-inverting mode, as shown in the diagram. The sound could be described as “meatier”, with subjectively more low frequencies. The engineer in me was displeased, since I think any audio circuit should be non-­ inverting. Inverting equipment can cause cancellations when combined in various audio pathways in the studio. I’ve had this distortion phase effect before, and it is usually caused by curvature cancellation in cascaded inverting stages. (Yes, it is possible to add another stage and get less distortion). Another cause is an asymmetrical clipping action. I thought it best to stick with the configuration that gave the best sound, ie, inverting overall. I’ve found most guitar pedals to be inverting. All the Colorsound Tone Benders are, apart from the Supa Tone Bender, which has an extra stage, and no one complains. Power polarity flip Transformer coupling also allows a PNP transistor to be biased ‘upside down’, as shown in Fig.6, Rb – bias + Germanium PNP T1 transistor fuzz boxes, such as the Silver Tone Bender, the low-pass filtering is done by making the input impedance rather low, and wiring capacitors between the base and collector of each transistor. Its 70kΩ input impedance rolls off the HF from the pickup above about 3kHz. It also renders the guitar’s tone control ineffective, which is disconcerting. In the circuit here, we can wire a capacitor across the interstage Filtering Low-pass filtering of the input signal reduces harsh intermodulation products. High-pass filtering of the output removes the low-frequency beats resulting from intermodulation, and is usuVR2 ally accomplished Tone by using low-value coupling capacitors. CW On some bipolar C 2 Ce 220µF B TR3 C 3.3µF R 3 S D G TR1 T1 VR3 Volume R 5 R 6 C 7 CW C 8 TR2 S D G R 7 C 5 + VR1 Drive 220µF Output C 3 C 1 R 1 R 2 Rc 4.7kΩ Rc – collector load Ce – emitter bypass E + + D1 R 4 CW R 8 C 4 + C6 0V Re – emitter resistor Fig.6: a developmental circuit for using a PNP transistor ‘upside down’ to obtain a negative ground. 72 thus enabling a conventional positive power rail to be used. I tried this on a breadboard, and there was instability, possibly due to there being two ground reference points. I’m going to continue pursuing this, though, as I have 1000 PNP NKT214s in stock. Re 560Ω + 220µF Bias Silicon NPN Photo 10: inside the fuzz box, it’s ‘hard wired’, so very easy to modify. Source: Mike Grindle. V+ + 1N4148 Photo 9: the prototype Transmanium germanium fuzz box. Source: Mike Grindle. In Negative 0V power 0V 0V Out 0V is positive earth Fig.7: the PCB overlay for the Transmanium. There is provision for the optional stability capacitor just beneath TR2. Practical Electronics | November | 2025 Photo 11: the completed Transmanium prototype PCB. The capacitors have been chosen because they look pretty. Any old dielectric works fine for fuzz. transformer. It is possible to get a resonance roll-off effect in combination with the transformer’s leakage inductance. Since it depends on the particular transformer used, I’ll leave this for you to experiment with. With the 5:1 interstage transformer, a 22nF capacitor across the primary sounded good, getting rid of the finger clicks. Clipping The type of clipping is important; it should be gradual in onset, ie, soft. Symmetrical clipping is better for bass guitar, since it generates a strong third harmonic which enhances clarity, especially on loudspeaker systems with restricted bass response, where the fundamental is cutoff. Second harmonic in isolation confuses the brain by sounding like another guitar playing an octave higher. For lead guitar, asymmetrical clipping is better, as the second harmonic enhances the screaming quality. In most fuzz boxes, including this one, the clipping starts off asymmetric, enhanced by the square law of the JFETs. Eventually, it almost becomes a square wave when fully driven. Of course, the tone control on the output completely changes this. In Practical Electronics | November | 2025 the traditional Tone Bender tone control circuit, where each end of the pot track is fed by low-pass and high-pass filters, the pot has a midrange dip in the middle, there being no flat setting. If there were, it would sound horrible. The raw clipping has to be filtered to sound good. Flat batteries Some guitarists have noticed that their fuzz boxes sounded better when the battery was flat. I investigated this, and mostly this was down to earlier clipping, giving more distortion. I used this effect in the Yellow Tone Bender by running the germanium transistor at a lower voltage of 4.7V rather than the normal 9V, which is what zener diode D2 is for. I also noticed that the Tone Bender MkII sounded extra special when the battery was flat. This was due to positive feedback via the power rail, giving a peaky frequency response almost on the verge of oscillation. I did a special variant of the unit where I incorporated a 56Ω resistor in the ground line to simulate the internal resistance of a flat PP3 battery. I told you guitar pedal electronics was crazy! Construction Luckily, guitar pedal construction is much more sensible, often going back to the simple kitchen tabletop techniques of the 1970s. The prototype for this design was hard-wired using tag strips, shown in Photos 9 & 10. Appearance is a major consideration with guitar pedals. In my Colorsound work for Macari’s, I’m not allowed to use surface-mount devices, and all components have to pass the visual test. When I had a batch of polyester capacitors that were moulded in grey plastic rather than the previous blue, they were rejected. I suspect the saleslady at Rapid Electronics marked me on her computer system as nuts. I’m considering having my fibreglass PCBs coated in a specially made brown solder resist coating to emulate the Paxolin PCBs of the 1960s. In guitar pedal electronics, the desired aesthetic is a PCB that looks like a colourful flower bed. The standard green or blue is considered undesirable. I compromise and use the clear orange resist option provided by PCBWay and JLCPCB for my production units. I’m sorry, but the board given here will be standard green (shown in Photo 11). Fig.7 is the overlay diagram that indicates which components go where. Mounting components One of the trickiest parts of designing PCBs for pedals is that the controls have to be at the top of the box, rather than the side. This often necessitates pots with angled pins, as shown in Photo 12. These usually have to be specially ordered, with high prices and high minimum order quantities. Alpha Photo 12: a pot with properly angled pins. I got them specially made by Omeg. 73 Photo 13: a standard tag pot with shaped lengths of 22 SWG (0.7mm diameter) tinned copper wire. pots are normally the best value, but the MOQ for bent-pin specials is 3000. For the Colorsound work, I use Omeg pots, which are conductive plastic and have a long life. It’s still a big capital expenditure, so I have to get Macari’s to pay for them. At least they can claim the VAT back. It is often necessary to use standard tag pots, so wire links have to be used from the tag to the board, as shown in Photo 13. These are made from 35mm lengths of 0.7mm diameter enamelled copper wire, bent into a hook shape on a 3.5mm cylindrical mandrel like a 3.5mm diameter drill bit. The pots should be mounted on the board with nuts to ensure their cases are Earthed. This means extra nuts are needed to mount the board in the box. The pot bushes need to be at least 10mm long to account for the thickness of the PCB, front panel and washers. When the PCB is mounted up against a metal panel, it is essential to cut the component leads as short as possible to avoid short circuits. This is no problem where plated through-holes are used. On this board, it is possible to fit standard 16mm pots (Photo 14), but they won’t fully go through the holes, as shown in Photo 15. Like small LEDs, germanium transistors are sensitive to heat and movement while soldering. It’s not necessary to use a heat shunt on the leads, as has often been said in old magazine articles, but it is wise to do it quickly and use long sleeved leads (see Photo 16). With old components, the leads are often oxidised and need plenty of scraping to ensure the solder will stick to them. Axial capacitors The board has dual outlines for the capacitors, so if you are into the ‘retro’ look, you can install pretty 1980s Philips/Mullard-style parts. Black radial electrolytics cheapen the appearance of expensive audio electronics (although there is the consideration of reliability, where modern parts most likely have the edge). Photo 14: it is just possible to fit standard 16mm pots with a bit of judicious bending. Take care not to weaken the crimp. Transformer Oddly, the Greenweld interstage transformers I used had flexible leadout wires rather than pins. There were also no mounting clamps. This made them difficult to mount. As usual, Grindle (my PCB designer friend) came up with a solution. With modern PCB fabrication, it’s easy to make any shaped hole, so the PCB has a rectangular slot to hold the transformer core, shown in Photo 17. The transformer is then clamped down with a cable tie as illustrated in Photo 18. drain is almost the same as the rail voltage, as is normally the case with inductor loading. That’s why the output voltage swing is double compared to using a resistor. High-frequency stability Fuzz boxes have a very high gain, typically 40-60dB; in this case, it’s 600 times or 55dB. Feedback at high frequencies from input to output, causing high-frequency oscillation, is a common problem. The input lead should be screened, and the unit enclosed in a metal box. The metal cases of the pots must also be Earthed. If HF oscillation occurs, an added 470pF capacitor (shown dotted in Fig.6) helps. Increasing the gate stopper resistors on the JFETs can also fix it. R1 can be increased to 10kΩ, and a similar resistor could be added to TR2. These mods are shown in Photo 19. Failure modes The simplicity of the circuits used in guitar pedals means failures are rare. I find damaged battery clips to be the number one cause. DC power connector switch contacts are second, and finally, broken pots and switches due to the guitarist’s feet. The germanium transistors can suffer internal detached emitter/ collector connections and tin whisker shorts to the metal cans. Many of these problems can be traced back to the days when the production yield for these devices was very low, and the market was flooded with rejects. I used to buy whole bags of these untested transistors from companies Photo 15: with the 16mm pots, the tags won’t quite go fully through the holes, but you can still solder them to the plated barrels. Testing As usual, it’s the DC conditions that are the first things to check if it doesn’t work: power first, then the voltages at the transistor outputs. The voltage on TR2’s 74 Practical Electronics | November | 2025 Parts List – Transmanium fuzz box 1 double-sided 85 × 80mm PCB coded AO-NOV25-01 1 Hammond or Tayda metal box 1 5:1 turns ratio interstage transformer or small transistor driver transformer, eg, 1 Tayda 3PDT latching stomp switch (S1) LT722/LT44 (T1) [AOShop] 2 stereo switched 6.35mm (¼-inch) chassis-mounting jack sockets (CON1, CON2) 1 PP3 battery clip (BAT1) Qty Value 4-band code 5-band code 1 2.2MW 1 125mm-long, 2.5mm wide cable tie 1 22kW 3 lengths of 1mm inner diameter sleeving (for TR3) 1 6.8kW 3 extra potentiometer nuts 2 4.7kW 1 1.0kW Semiconductors 1 560W 1 150W 1 2N5460 P-Channel JFETs (TR1, TR2) 1 NKT214F, OC71, OC81D, OC75, AC126 or similar PNP small-signal LF germanium 1 1N4148 75V 200mA signal diode (D1) transistor (TR3) Photo 16: leaving the leads of germanium devices long prevents heat damage when soldering. like Bi-Pak for a few quid. Many devices had no markings at all. I’ve used some of these defective devices to make unique musical noise and distortion generators. Sadly for today’s constructors, these dud devices are now reappearing on eBay at high prices. Boxing it up I couldn’t find a suitable box apart from the Colorsound ones, which are not mine. So we’ll leave it for now until I find a suitable die-cast enclosure. That will be described in a future column, in which wiring up the device and suitable switching will also be covered. If you want to learn more about fuzz boxes, check out the book “Level and Attack – the untold story of the Tone Bender fuzz” by Simon Keeping, Anthony Macari and Steve Macari (https://www.11publishing. co.uk/books/tonebender). PE Capacitors (all can be radial or axial) 2 220µF 6.3V electrolytic (C4, C5) 2 47nF MKT, polyester or similar (C3, C8) 1 22µF 10V electrolytic (C2) 1 10nF MKT, polyester or similar (C1) 1 3.3µF 10V electrolytic (C6) 1 4.7nF MKT, polyester or similar (C7) Potentiometers (26mm size preferred for ruggedness) 1 470kΩ-1MΩ single-gang logarithmic (VR1) 1 100kΩ single-gang linear (VR2) 1 10-22kΩ single-gang logarithmic (VR3) Resistors (all ¼W ±10% axial or better) 1 2.2MΩ (R2) 1 6.8kΩ (R3) 1 22kΩ (R8) 2 4.7kΩ (R1, R6) 1 150Ω (R5) Photo 19: an extra resistor and capacitor soldered between the potentiometer tags can eliminate any high-frequency instability that may occur. Photo 17: modern PCB processes allow the cheap production of odd-shaped holes. I’ve had a slot cut in it to support the transformer. Practical Electronics | November | 2025 1 1kΩ (R4) 1 560Ω (R7) Photo 18: a cable tie is used to hold the transformer in place. It needs to be longer than you would expect. 75