Fullscreen HTML5Med Res (??MB)HTML4

1970s-style, lightweight Bass Guita tar Valve AM MP If there’s one place valves are still used extensively, it’s in musical instrument amplifiers, where their soft overload characteristics and non-linearities are not just desirable but critical to the sound. This bass amp has a pretty simple circuit, but it makes an excellent practice amp. Vintage Electronics by Fred Lever T he Fender Bassman article by Brandon Speedie (April 2024 issue; siliconchip.au/Article/16218) struck a chord with me (no pun intended). In the 1970s, I made dozens of “Woden” bass amplifiers, first with valves and then transistors. I used amplifier designs from the electronic magazines of the time, and 92 Silicon Chip recommendations from the valve manufacturers, modified to suit what I thought was the way to go. One thing I never did make was a classic all-in-one small valve bass amp like the Fender. Most of my amplifiers were head units that could sit in a rack or on a speaker box. The speaker box used may have Australia's electronics magazine been an existing unit or made using 12-inch (300mm) or 15-inch (380mm) drivers. An amplifier the size of the Fender Bassman was OK for small stage use or for practising, but would not fill a venue with driving bass. If that was required, a Fender could be monitored with a microphone and that signal fed to a bigger amplifier. The other aspect of a Bassman amplifier was the fabled Fender ‘tone stack’. I never fitted any more than the regular cut and boost or an equaliser to my amplifiers. Guitarists inserted voice boxes and equaliser preamps of their own into the signal path to shape the music, and my amplifiers really just provided the ‘guts’ to fill the audience area with sound. The ‘tone stack’, combined with the all-valve lineup, was a bit of a mystery to me. So it was finally the time to make the effort to build one and find out. This amplifier is a nice matching unit for a 1960s-style lightweight bass guitar I made, as you can see in the lead photo. I found a modern 9kg solid-­body bass guitar just too heavy to wear standing, so made a ‘stick’ bass. I grafted a ready-made stock fretboard, pickup and string fittings onto a length of Tasmanian Oak and completed the body with three-ply, giving a total weight of 5kg. My amplifier cabinet here is similar to the Fender Bassman in that it has a 15-inch single speaker with the amplifier head built into the top of the assembly. Unlike the Fender, I refuse to mount valves upside down; I hate valves falling into the cabinet during transport! So I decided to make a two-part assembly with the valves and heavy parts in a normal chassis that slides into place on cleats, and a horizontal upper panel with all the controls on it. I have some 6DQ6 TV sweep valves and plenty of 12xx7 twin triodes, so I used those and wired them up in a manner very close to the Bassman circuit shown on page 91 of the April 2024 issue. I did make some changes to suit the parts on hand, mainly the 6DQ6 valves. As the output valves determined what sort of transformers were needed, I decided to start there and wind my own, having the necessary wire, laminations and a winding machine. In the March 1962 Radiotronics application notes, Class-AB audio operation of 6DQ6 valves requires a siliconchip.com.au plate (HT) of 440V DC and a screen voltage of 200V DC. That results in a single output pair giving a theoretical 80W when biased with about -40V into a load of 3.6kW. The notes indicate a valve draw of 300mA, so I budgeted for a total of 350mA HT at 440V DC. Add the total heater consumption of about 5A at 6V and the required power transformer rating is 184VA. I confirmed this assessment by finding my chassis build notes of some bass amplifiers with 6DQ6s I built back in the 1970s and checking what transformers I wound then. Unlike the Fender circuit, with one valve rectifier and the full voltage applied to the plates and screen droppers, I used two silicon bridge rectifiers in a two-winding stacked supply to give 200V and 450V. Doing that reduces the voltage applied to each rectifier and the smoothing capacitors. Otherwise, I followed the Fender plan, with the 12AX7 phase splitter driving the output valves, the 12AX7 direct-coupled ‘Fender stack’ tone controls and the 12AY7 high- and lowlevel preamplifier. The output transformer had the primary winding for each valve wound bifilar to provide equal impedances for the valves. While home constructors can safely make their own inter-stage or output transformers, they should not attempt mains transformers since the safety requirements are critical. Given the relatively odd set of voltages required for this build (2 × 6V, 40V, 160V & 210V), to use off-theshelf transformers, you would likely need several with the primaries wired in parallel. A good source of possibly suitable transformers in Australia is: www.antekinc.com/power-­ transformers Failing that, there are local manufacturers like Harbuch Electronics (www. harbuch.com.au) and TorTech Pty Ltd (www.tortech.com.au) that can make custom transformers for you. Evatco also sell Hammond-branded transformers, which are well suited for valve amplifiers (www.evatco.com.au/ transformers). If you can’t get the exact voltages I used, you have the option of using slightly higher voltages (based on what’s available) and adding wirewound resistors to drop the voltages to be closer to mine. If doing this, make sure to over-rate the resistors for siliconchip.com.au dissipation and mount them so that the extra heat won’t cook anything critical! More on sourcing transformers later; first, let’s look at what I built. The chassis I made the steel bottom section with the transformers and output valve stages at one end and the input valve stages at the other end – see Photo 1. The chassis is made of some scrap 1mm sheet steel pieces I had, about 400 × 200mm. I bent and drilled one of those to my rough circuit layout and put plenty of holes in the surfaces to take all the parts my circuit indicated were needed. For the top control panel, I chalked in logical groups of parts with the signal flowing from the input sockets on the left, through the volume and tone controls in the centre. The power switches, fuse and a line output jack are on the right-hand side, above the output valves and transformers. A modern designer would make a 3D CAD drawing and place the parts. An old builder like me looks at the proposed circuit, shuffles the parts around on the chassis surface until they all fit, then draws centre lines, punches them and cuts holes. I sanded the metal until it was bright, washed it down with Prepsol and sprayed on a coat of plain grey epoxy anti-rust paint. Referring to Photo 2, the signal comes in through the phono jack sockets on the left and down to the lower chassis, then travels straight across through the preamp, tone control and phase splitter valves. The signal stream tees back out through holes next to the mini 9-pin valve sockets, up to the tone control groups. All that wiring is shielded to minimise hum pickup. The output valves are next, feeding down to the output transformer. The plate wires come straight out of the transformer, directly to the valve cap fittings. Photo 1: the drilled and painted chassis, ready for components to be installed. Photo 2: the control panel mounts right above the chassis. Australia's electronics magazine November 2026  93 Photo 3: I was able to keep the smaller components under the chassis fairly neat. Leakage in the old Bakelite tag strips caused me some problems, though! The two items next to the output valves are the bridge rectifiers, with the power transformer below. I used ceramic sockets for the hot high-voltage high-power items; I learned that lesson very early on. The bridge rectifiers are soldered into plug-in bases with an idea of possibly hiding them later in defunct valves for that vintage envelope look. The mains power controls are on the right-hand side of the top panel. Getting the output stage going I wired up the mains supply and output stages first to check that my old valves worked and the transformers were satisfactory. After some initial problems using 50-year-old parts (unformed electrolytics, weak valves, failing minor components), the output pair came to life and a load test showed my transformers were on the mark for voltages and the output impedance conversion. On a straight power output test, the HT supply drooped to 430V/190V for the 6DQ6 pair with a 60W continuous sinewave into 8W (resistive) at clipping. That meant I could call the amplifier a 60W unit. The HT power supply droops nicely, being 460V/210V with no load to 430V/190V on load. The 6DQ6 valves go gracefully into overload, with bags of low order harmonic distortion, which is exactly what you want for an electric guitar or bass (see Scopes 1 & 2). Hifi enthusiasts would faint at 30% THD, while 94 Silicon Chip guitarists just turn the volume knob to 11 to get more! In Scope 1, the yellow trace is the sine input and the cyan trace is the speaker output. The valves are starting to clip and distort on the flanks. In Scope 2, the yellow input level has just about doubled, but the output has just clipped off, with loads of flank distortion but no more sound level. All that changes is the tone; the string note sounds more like a raspy woodwind instrument. In Fig.1, the cyan curve is the maximum output with a 60Hz input sinewave with the load varied from 2W to 16W. The amplifier’s power delivery is pretty flat from 6W to 10W, ideal for an 8W speaker load. Fig.1: the cyan curve shows how the amplifier’s output power varies with the load impedance/resistance at 60Hz, while the red curve is its frequency response at around 40W average output. Maximum power is with a load of around 8-9W, so an 8W loudspeaker is the ideal choice. Australia's electronics magazine The red curve is the frequency response from 30Hz to 6000Hz at about 40W. Under 100Hz, the output sinks like a stone due to high-pass (AC-­coupling) filter roll-offs and transformer saturation. Above 1kHz, lowpass roll-offs do likewise; the amp is pretty much mute above 10kHz. Pre-amp stages The preamps worked as expected, apart from a hair-tearing time where leakage in the old Bakelite tag board caused havoc with the high-­impedance volume control circuits. The Bakelite had measurable resistance in the megohms range between the tags that the 12AY7 plate coupling capacitors and grid feeds were terminated onto. That put ‘ghost resistances’ into the potentiometer connections, and the volume controls interacted weirdly with each other. Turning one channel volume down would cause the other channel volume level to rise! Shifting those wires onto a new piece of tag board restored sanity. With that problem out of the way, I checked each stage for gain and measured the overall sensitivity. The full volume sensitivity with the tone controls at their mid positions was around 20mV for a 60W output on the number 1 low-level input. With a 60Hz sinewave input, the 12AY7 had a gain of 7 times from jack socket to plate; the 12AX7 had about 40 times from its grid to plate; the phase inverter 12AX7 about 17 siliconchip.com.au times; and the 6DQ6, about 12 times. Of course, there are insertion losses between each stage. An input of 20mV input yields 20V across the speaker coil. This gives an overall gain of about 1000 times. The ‘Fender tone stack’ I swept the input signal to the amplifier with a sinewave from 10Hz to 2kHz (remember, this is a bass amplifier) to get an idea of the effect of the controls on the frequency response. The reading CRO is connected across the speaker voice coil. There is an excellent article describing these tone controls in Silicon Chip, April 2024, on page 90. My controls may not do exactly what is described in the magazine, but they do something fairly logical. All the controls interact with each other. Scopes 3-10 were made with all the controls set to mid position as default and then one control moved up or down. The bass, mid and treble controls all work as you would expect, but the Presence control is the odd one. Being in the feedback loop, it affects the overall level as well as boosting the midrange and bass. So, you have four controls that interact with each other. A guitarist just turns the knobs until the sound is ‘right’. Scope 1: the input (yellow) and output (cyan) waveforms at the onset of clipping (60W in 8W). Scope 2: the same waveforms as in Scope 1 but with the level pushed up well into clipping. 3 4 5 6 7 8 9 10 Chassis assembly I mounted two long tag strips across the chassis, intending to place most of the parts onto them. On the right in Photo 3, the power transformer connections and bias rectifier parts use the majority of the strip. I used 0.75mm V90 mains-rated flexible cable, not radio hookup wire, for the power and HT wiring. For the signal connections, I used ordinary hookup or shielded wire. The HT filter capacitors are mounted in the centre, along with the various HT dropper resistors and smaller filter capacitors. Using the tag strip to mount the parts around the preamp valves was a disaster, as mentioned earlier. The Bakelite exhibited measurable leakage of a few megohms between tags. This phantom leakage played havoc with high-impedance parts of the circuit, so in the end, I connected most of the signal components point-to-point on the valve sockets. I also mounted a new, small strip on the back panel to provide points siliconchip.com.au Scopes 3-10: these sweeps show how the frequency response changes with each tone control adjustment. From top-to-bottom, left-to-right they are: bass down, bass up, treble down, treble up, midrange down, midrange up, presence down & presence up. Australia's electronics magazine November 2026  95 for high-impedance signal wires that couldn’t be run directly. I ran an Earth bar across the chassis from the mains Earth stud on the righthand side over to the left, securing the bars with lugs bolted into unused holes. Small components could then be mounted from the valve sockets to Earth. The heater wires were connected with two-pair twisted wire. Later on, during testing, each side of the heater supply was linked to Earth to test for the lowest hum level on full volume. On both supplies, Earthing one particular side of the heaters gave far quieter results than the other. To reduce the mains hum pickup, all the signal leads teeing off to go to the front panel were shielded, with the braid at one end only connected to Earth via the bus bar. Most of the small components are from Jaycar. The valves, valve sockets and main electrolytics were new-old stock from the 1960s. All resistors are large-body, 1W ±5% types. Under the top panel (Photo 4), the controls have the associated small components mounted directly onto them; the signal phono jacks are on the right, with the tone stack controls in the centre. All the mains-rated power wiring was spaced well away to the left. Any exposed power connections were sleeved with heatshrink tubing. The chassis could be a conventional front panel layout, with much shorter wiring between the controls and the valves. The layout used here was simply something I had not done before, and I wanted to give it a go. As long as common sense is used, the layout is not critical. In Photo 4, you will note a spring bar arrangement between the two valves. The spring and bar exert a small pressure via the top caps. This is to ensure the top cap lead does not pop off from an impact during transit. A plate lead coming off means a glowing screen and a ruined valve. I prefer the extra expense of the spring assembly over a warranty call. The chrome handle at the back of the chassis is a grip point so you don’t put your fingers into the HV area when lifting the chassis out. Making the speaker box I dropped into Bunnings and bought a couple of sheets of ‘marine’ five-ply and eight lengths of 20 × 10mm rectangular bar. The design process started with a front speaker panel just wide enough to mount a 15-inch speaker inside a frame; everything else was sized from that. I placed a 15-inch old-stock speaker onto a five-ply sheet and figured I could fit the chassis and speaker into a 600mm-high, 460mm-wide and 225mm-deep box, making a simple open-back cabinet. I cut the front speaker panel to 600 × 400mm and then made a flat bar rectangular frame to fit that. This frame has butt joints screwed and glued together. For the wood, I would recommend using Titebond wood glue, which you can find online without much trouble. The next step was to cut the hole for the speaker in the front panel and glue the panel to the frame (see Photos 5 & 6). Then I plated the sides and bottom with sheets, leaving the top open for the control panel. To allow the chassis assembly to slide in, I cut the top rear bar out of the frame once the glue had set and the frame was solid. I added the necessary slide rails to carry the weight of the amplifier assembly and a removable rear panel to trap the whole chassis assembly into place. Photos 7 & 8 are views of the amplifier during construction. No thought was given to speaker box parameters or acoustics. This “Fender” is an open-back plywood box with all the back wave cancellation, resonances and foldback built in at no extra cost! I coloured the cabinet in the same style as the guitar, mainly in red with some black contrasts. I simply painted the timber cabinet with a Cherry Red enamel spray can from Bunnings (the horror!). The first step was to spray the inside with matte blackboard paint and continue the black to the outside in a few places. The surfaces that were going to get red were undercoated with two brushed coats of wood sealer. I cut that back with 120-grit sandpaper to smooth out some of the rough areas. I emptied a full can of Cherry Red spray over the box in two coats. First came a tack-dry coat, then a second layer as thick as I dared; the can ran out just as I finished. I left it for a day Photo 4: a better view of how the front panel controls and connectors are wired into the chassis. The spring contraption stops the valve caps from falling off if the unit gets rough handling – as such gear often does! 96 Silicon Chip Australia's electronics magazine siliconchip.com.au Photos 5 & 6: the cabinet frame after I cut out the loudspeaker hole but before I attached the side panels. And another photo with the 15-inch (380mm) speaker driver mounted. to cure. I did not touch the surfaces with abrasive paper after that; I wanted a slightly rough texture all over. All sanding back does is expose more imperfections in the surface, and there are plenty of those with scratches, nail heads, hammer and saw mishaps. This is definitely not a craft finish, but an ‘industrial’ treatment, as is the guitar. Someone building a similar cabinet could use a vinyl wrap as an alternative. Back in the old days, we glued black ‘leatherette’ straight over the bare wood. That covered up any stray holes and imperfections, and it can be stretched to pull around corners and held down with suitcase corners and tacks if necessary. A bit of added ‘bling’ is the name “Woden” in modest 3-inch (76mm) letters across the box front and a card on the control panel depicting the control positions and Mr Woden wielding a four-string axe (see Photos 9 & 10 overleaf). The loudspeaker I started by using a very old (1980s) Jaycar 15-inch loudspeaker (Cat CG2382) with a two-inch (50mm) voice coil as a trial for measurements, and left that in place to see how it would perform on actual guitar playing. The answer was: terrible. The speaker was a reject, having a loose magnet, with the voice coil rubbing slightly. The basket on those 1980s speakers was just too flimsy; even being dropped while in the shipping box could knock the magnet crooked. I straightened the frame and epoxied the magnet into place. That made the speaker usable, but on playing tests, it was pretty ordinary. siliconchip.com.au Photo 7: the chassis installed in the frame with the side & back panels now attached. Photo 8: a better view of the internals with the back panel removed. I bought a cost-effective Celestion TF1525 unit with a 2.5-inch (63.5mm) voice coil and 200W rating. With a proper bass guitarist pushing the strings, the Celestion driver was way superior to the ancient driver. Overall, the tone balance was biased more to the bass end, but not so much as to skew the tone balance of the guitar fretboard. The big difference was the more direct translation of the strings and fingering and a lack of colouration by the speaker cone. That is the difference between Australia's electronics magazine November 2026  97 making a noise and making music. The Celestion speaker has a variable radius section, smooth cone that works much better compared to the basic flat-with-corrugations surface of the old unit. The Celestion cone moved freely, unsupported by any box back-loading, but did not sound like it was pushed into non-linearity on larger excursions. The valves are already doing that nicely, running near their maximum output. The amplifier simply cannot provide anything more than about 100W of power to the speaker cone. I am confident that the speaker can take that all night with minimal heating, so it is in no danger in this build. The speaker is fixed ‘back of panel’, not with wood screws, but with largeheaded through-bolts, to ensure the basket is clamped down tightly and cannot move or vibrate. A stock 15-inch mesh speaker protection grille is attached to the timber on the outside to give the cone some accident protection, as well as neatening the look of the speaker cut-out, and covering unused spare holes. The bass guitar The guitar is based on a solid-body bass kit from Pit Bull (similar to the kit reviewed by Keith Walters in the November 2017 issue; siliconchip. au/Article/10870). I did not use the heavier solid body that comes with the kit, but bonded the kit neck onto a beam of Tasmanian Oak (Photo 11) and made a hollow body. The hollow body was constructed from marine three-ply and simply serves to give some shape. The body has no other function, and could have been a triangle, star shape or none at all as it has no musical function (the Steinberger electric bass famously has a tiny body, just enough to hold on to). The fretboard neck is screwed and epoxied into a socket in the slab of oak (see Photo 12). That makes an almost continuous backbone for the strings to stretch from the anchor plate to the tuning pegs. The three-ply body is shaped and screwed into blocks glued around the body perimeter. The oak plank has recesses chiseled out for the pickups and anchor plate. I copied the overall dimensions from a finished kit so the string height and length were standard. The weight saving comes from losing the equivalent of about half the volume of the solid body. I also added some thumb rests. Keeping it simple, the only control is one volume control with the two pickups wired in series through the potentiometer to a standard phone jack socket. A builder may consider using a more complex tone and pickup combinations, such as those described by Brandon Speedie in September 2024 (siliconchip.au/Article/16581). I stained the oak and ply with clear varnish tinted with Cherry Red enamel. I applied enough coats to fill Photos 9 & 10: the finished amplifier looking spiffy with its Woden-themed artwork. And of course the knobs go all the way up to 11! 98 Silicon Chip siliconchip.com.au Fig.2: the circuit uses a tried and tested configuration, with twin triode V1 acting as the preamplifier and twin triode V2 providing further amplification. The output of V2 feeds into the tone control network, then to V3, which acts as a phase splitter to drive the output valves, V4 & V5. Warning: all the HT rails in this circuit (150V, 180V, 200V, 250V, 325V, 380V, 460V etc) carry lethal voltages when the circuit is powered and for some time after. All wiring must have a suitable high voltage rating and all joints need to be insulated properly. Do not make contact with any part of the circuit while power is applied. the wood grain until a smooth finish was achieved. The side section was filled with a band of alloy strip, adding to the ‘homemade’ 1970s feel, as seen in Photo 13. A normal shoulder strap with stud attachments sufficed for the weight. Making one of your own Making a valve amplifier nowadays will be dictated by which output valves you use and then sourcing output and power transformers to match the valve impedance and voltage requirements. I used the 6DQ6 valves here because I had them and had used them before (see my circuit diagram, shown in Fig.2). siliconchip.com.au Photo 11: this is essentially a complete electric bass – just a little hard to hold on to! Australia's electronics magazine Photo 12: gluing and screwing the fretboard neck onto the panel that will be the back of the guitar. November 2026  99 They were also used in an Electronic Australia guitar amplifier in June and July 1967 by Tony Leo, called the Playmaster 116 & 117. Those articles are worth reading; you may prefer the tone controls or even the complete circuit they used. While the 6DQ6 is basically a TV sweep generator, the maker was happy for it to be used as a linear amplifier. The smaller 6CM5 was not a good valve for audio work, basically being a switchmode valve. I know they were used by AWA in commercial PA amplifiers, but I had no luck in getting consistent results. There is a raft of other big valves you can use, like the KT66, that sell for silly amounts. In later Woden builds, my go-to valve was the 6CA7. These yielded about 90W per pair with 700V on the plates. They would work well in this circuit; all the connections are via base pins, so you can keep the killer voltages inside the chassis. Once you have your proposed valves, consider the transformers. If you can wind your own, the problem is solved. There are some interesting websites that detail the use of stock power transformers as output transformers, using either a single unit with a centre-tapped primary or two transformers (try a web search for “cheap output transformers”). Suppliers such as Evatco list sensible units such as their 1750L for around $200-300. Keep in mind that this is a guitar amplifier, and very expensive high-­ fidelity transformers are not needed. Power transformers with 6.3V heater windings and useful HT windings abounded in old valve TV sets. What could feed a B/W TV containing a 6DQ6/6CM5 sweep valve and a dozen other 9-pin valves is going to have enough wattage for any small guitar amplifier. Alternatively, transformers always bob up on eBay and other auction sites. New transformers can be bought for large amounts of money. With the valves and transformers sorted, the rest of the parts can be straight from Jaycar, Altronics or other suppliers on the web. The thing with a guitar amplifier is that nothing is critical. The parts can be ±5% or ±10% tolerance; achieving flat responses and perfect linearity is not important. As for the fabled ‘Fender Stack’ tone controls, well, you can take your pick and use those or just put in a Baxandall bass and treble set like the PM117. A bass player will just set their pickups and any tone controls to get the ‘right’ sound. The chassis can be a conventional design with the controls on the front, just like the PM117. The actual dimensions of the chassis will be dictated mainly by the size of the transformers; the speaker box and amp box can be all-in-one. My theory for bass is that you want mass in the box, so the box and speaker frame stand still and only the cone vibrates. A light box with a weak front panel can have the cone tending to stand still and the panel and box vibrate! That isn’t the most efficient or accurate way to promote bass sound waves. My Woden FB60 weighs in at 20kg. It needs a trolley to move it, or a big roadie, but the box does not flap around. Building the system into separate head and speaker boxes makes it easier to transport. Since making the Bassman replica amplifier, I spent a significant amount of time investigating sources of power transformers, output transformers and output valves that may be used in a modern guitar amplifier. Power transformers I came to the conclusion that there are a lot of ex radiogram and television power transformers that can be pressed into service. In their former use, these provided a 250V HT rail and lit up a dozen or more 6/12V valves, similar to the wattage load of a 50W guitar amplifier. A doubler or tripler may be used to get a higher HT value. Of course, you can always buy a new transformer to suit your valves from a supplier like Evatco. Output transformers The field here is limited. You are pretty much stuck with buying a purpose-made unit from Evatco or finding one. I tested some alternatives, such as using stock power transformers. There are limitations and drawbacks to this idea but they can be made to work. Alternatively, you can use a transistor based amplifier as the final stage, giving a so-called ‘hybrid’ design. However, then you lose a lot of the character of the valve amplifier that’s so desirable to musicians. Photo 13: the completed bass guitar. It weighs about half of a solid-bodied equivalent. 100 Silicon Chip Australia's electronics magazine Output valves While any high-power tetrode or pentode will work, the best value valve to use is the 6CA7 tetrode, also known as the EL34 pentode. These were the workhorse valve in the olden days and nowadays are still available at a cost-effective price. For guitar work, odd pairs are fine – matched pairs are not needed. I have bought single valves for as little as $40 each from various sources, so I suggest these are the valves to use. This article is a condensation of two main projects lodged on the Vintage Radio forum website, see https://vintage-radio.com.au/default. SC asp?f=12&th=139 siliconchip.com.au