This ‘Truvoice’ amp was produced around 1964 at Selmer’s WC2 factory (just around the corner from where I used to work – hard to imagine now anything being manufactured in Bloomsbury London). This was the famed ‘croc-skin’ period of Selmer amps (here is how Mod sharp one would have looked), though any trace of that covering was long gone when I got this one in its stripped and varnished plywood cab. It supposedly also had a failed output transformer, and some control knobs and one preamp shielding can missing.
Jim Marshall famously chose to copy the ‘tweed’ era Fender Bassman combo (5F6-A circuit) for his amp heads and the rest is history. With the Treble’n’Bass Fifty, Selmer went with a clone of the later Fender ‘blonde’ Bassman head (it most closely resembles the Fender 6G6 circuit). There are however some differences, notably the absence of a Presence control, and the positioning of the power supply choke ahead of the output valve’s anodes so that the entire current requirement of the amp goes through it. This results in much more significant voltage ‘sag’ under load than would in any case be caused by the amp’s GZ34 valve rectifier. The power supply was probably done like this to prevent the EL34 power valves (another major difference from the Fender) redplating at high signal levels under the excessive anode voltage put out by the amp’s power transformer – 480V or so is really too high for the output transformer load impedance of only 3.7kohm to suit a pair of EL34 s. I found that a well struck open E Maj chord could cause the amp to cut out entirely for what seemed like seconds at a time, although some of this was no doubt down to the tired original canister filter caps, which I replaced (mostly using 50uF sections instead of 32uF to stiffen it up a little). Perhaps it was this cutting out that accounted for the supposed failed output transformer ‘diagnosis’ – good job as this transformer is a now desirable Dagnall like later Marshall Plexis used.
Unfortunately, as stock these amps do not sound anything like as good as a ’60s Marshall. There are two channels, ‘Normal’ for guitar and one voiced for bass as was common back in the day. The Bass channel does not suit guitar at all, being way too dark. It’s really not very good for bass either; the amp is far too low powered at 50W and with that saggy power supply for any convincing stage level volume, and it just doesn’t have enough low end punch for a good foundation sound (bass tended to be recorded direct to the mixing desk in 60s studios and amps like this make it obvious why).
The Normal channel was better for guitar, with some lovely chiming cleans, but an overly rapid onset of breakup with increasing volume that was too muddy to be at all ‘classic’. Comparison with the Fender 6G6 schematic revealed that the Selmer has detail differences in this channel in terms of some tonestack component values, and also the positioning of the volume control in relation to the rest of the circuit. It’s not at all obvious why, but Selmer placed it prior to the tonestack, resulting in adversely heavy attenuation of the signal and tone suck, whereas the Fender circuit placed it after, also thereby doing away with the unnecessary extra capacitor required to keep high voltage dc off the Selmer’s volume pot. Faithfully copying the 6G6 arrangement (also including the addition of a Presence control in the power amp feedback loop for more top end sparkle) got the tone of this channel to be very good over a much wider range of the volume control, and when it does break up it’s fairly gradual, retains dynamic response to pick attack, and except at the very highest volumes and bass control settings is not too mushy to be unusable (the Beatles and The Who were blonde Bassman users circa mid sixties).
These changes also made it a better channel for bass than the stock Bass channel, which just goes to show that voicing specifically for a low response is not necessarily the best idea for a good primary bass sound (it tends to sap power).
These Selmer amps do seem to have a bit of a reputation as vintage pieces; there is the striking croc-skin cosmetics, and also an association with Syd Barrett (for guitar) during early Pink Floyd, and Jack Bruce (for bass – now I understand his driven tone) as notable among users.
With the cosmetics on this one being so compromised for any originality, and after only a little hesitation, I decided to strip out the Bass channel circuitry completely, and replace it with a Trainwreck Express type three stage high gain channel to nicely complement the clean Normal one. The Express circuit is well regarded as a fitting legacy of the late amp guru Ken Fischer, and I suspect a fair few actual Fender Bassman heads have had such circuit mods performed on their Bass channels in the time before they became collectible vintage amps in their own right.
Putting the Trainwreck circuit on to the preamp’s eyelet board section vacated by the Bass channel worked quite well (with the ‘Wreck’s tonestack components being hard wired on to the tone and volume pots). The required Middle control was added by moving the volume control to the hole for the channel’s second input jack (I’d already put the added Presence control in the Normal channel’s second input). I used some old Marshall knobs for this channel.
The Express circuit is notoriously fussy about its layout and lead dress in terms of keeping the noisefloor from becoming obtrusive, and oscillations at bay, but I didn’t end up with any problems with these aspects. Utilising the existing shielded grid leads and lowish R value gridstoppers on the first two stages has no doubt helped here. The grounding should ideally be a star schema, but I retained the bus type original already used by the Normal channel, and there isn’t too much backgound hum (although it is definitely a noisier hi-gain topology compared to a JCM or later Marshall). Decent valves being critical for stability with this circuit some old but still good RCA 12AX7A greyplates were used.
Inexplicably, Selmer left one power supply cap section unused even though they ran the bass channel’s four triodes off only a single section (two triodes usually being regarded as the max for reasons of proper decoupling), so this was brought in to use to provide the required extra supply node.
Most of the existing power amp was already to spec for an Express circuit, although some of the phase splitter tail component values were tweaked. It is important for the Trainwreck circuit that the preamp, phase splitter and power valves all begin to clip at about the same level of drive, in order to achieve the touch sensitivity and harmonic saturation with gain levels while cleaning up well at the guitar volume knob that these amps are so renowned for.
Tough 1970s RFT EL34 s were used for the output valves to guard against arcing, given how much stress their screen grids are likely to be under from the voltages. The low load impedance output transformer is rather different to the Express spec of 6.6kohm, and the supply voltages are all quite a bit higher, but much of the Trainwreck Express sonic character is there. When really digging in, much like a ’60s Marshall the power amp anode voltages do sag massively from 480V idle down to below 370V (the screen grids sagging to under 350V), so this probably results in pretty close to the same dynamics. (I suspect much of the reasoning for the original Trainwreck Express’ low voltages and high load impedance was for the allowable use of 6V6 valves interchangeably with the EL34s.) Anyway, as the cliche goes, so does this channel on the amp – from clean to mean!
So between its two channels, this amp now convincingly covers pretty much all the tonal bases found in amplified guitar since the blues had its baby and they named it rock and roll.
Looking at the pix now, I guess the wood finish probably was always saying “hey, make a Trainwreck”…
Oh, and the box on the right is an Airbrake-style attenuator (also a Ken Fischer design) – pretty necessary with the volume level arising from the gain of the Trainwreck channel – it’s a very, very loud ’50W’ when cranked.