DIY Workshop: Marshall 1930 Popular
A rare Marshall combo from the amp giant’s early days is lucky find indeed – but it’s not much good if it doesn’t work.
How many golden era guitar amps were bought by keen guitarists who, when life got in the way, simply shoved them up the loft to sit unplayed for decades? Being the lucky person to come across such a vintage treasure is one of the ultimate guitar hunter dreams, and that’s basically the story of how this Marshall 1930 Popular combo came into the hands of its current custodian.
The Marshall had been bought new, along with an Olympic White Stratocaster, by its original owner, who used it extensively for a number of years, before circumstances caused both to be retired until a house move decades later. Upon discovering it, the Strat was understandably kept for sentimental reasons, but the amp was offered to a guitar playing family friend, who couldn’t believe his luck… at least until he turned it on.
Since the amp came out of the loft, the 2×10 combo has emitted a loud wooshing hiss that made it totally unusable, and lacking the know-how to correct the issue himself, this handsome-looking Marshall has remained unused by its owner for over a decade since.
The 1930 Popular is slightly obscure Marshall model – it was only in production from around 1967 to 1973 and it was only sold via mail order. Consequently they’re regarded as fairly rare, although that doesn’t necessarily mean that few were sold. One theory holds that supplies have dried up because unscrupulous types have been converting 1930s to sell as bona fide 18-watt 1974 models.
It would be a fairly easy to do because the 1930 and 1974 models share the same chassis along with the same complement of valve socket holes. Two of these holes are blanked off on the 1930 because it has one fewer preamp valve and solid-state rather than valve rectification.
These were regarded as low-end or ‘beginner’ amps at the time, and the build quality falls short of Marshall’s best standards. Maybe they got the trainees to build them for practice? Even so, it’s a favourite studio amp of Dan Baird from The Georgia Satellites and as a proper early Marshall, resale prices are very strong.
You’ve no doubt heard of ‘case candy’ so let’s say this amp came with ‘cabinet candy’ courtesy of its original owner. Inside, there was an envelope with a typed address post-marked with a 19/1/73 date containing an official Marshall schematic. This was both helpful and exciting, because I couldn’t find this schematic online and the only 1930 diagram that is available is a hand drawn effort containing a couple of component errors – if you’re restoring your own 1930 and would like to take a look at the schematic, drop us an email.
I also found the Marshall cabinet plate, along with a boxed Mullard ECC83 and a couple of AEG power valves. Unlike the 1974, which runs on EL84s, the 1930 has a pair of ECL86s. These are very similar to the ECL82s used in the WEM Westminster because each glass envelope contains a preamp triode along with a power pentode. The triodes are used as phase inverters/drivers for the power pentodes to generate 10 watts of power.
This 1930 is in great shape cosmetically, and is still fitted with its original Celestion 7442 ceramic speakers. The earliest 1930s had Elac alnicos and a plexiglass control panel, but this one has the brushed metal panel. Some maintenance work had clearly been done over the years, with a handful of replaced capacitors. However the big filter capacitors and most of the carbon film resistors look original.
Before firing up the amp, I give it a thorough inspection. Straight off I notice that only one of the speakers is connected to the output transformer. When I removed the rear panel yet another power valve falls out. I can’t even begin to speculate why it was left to rattle around inside. A couple of power supply resistors show signs of heat damage but they measure ‘close enough’ and the electrolytics are in similarly good shape.
I decide to go for it and power up the amp since the owner told me he had done so quite recently. However first I need to re-connect the second speaker. The schematic specifies an eight-ohm load and each of the Celestions measured 16 ohms, so they have to be wired in parallel.
Touching the wires of the second speaker with a nine-volt battery’s terminals confirms it’s working. It’s also a good way to check polarity because the cone will move forwards or backwards depending on which way around the wires are connected. If you can’t see it, simply place your finger on the back of the cone and you’ll feel it.
I test both speakers this way and mark the positive terminals on each. It soon becomes obvious that the last time these speakers had been used in tandem, they would have been out of phase – with the speakers properly hooked up the amp is ready and raring to go. Well, sort of…
The owner described the noise as being far from pleasant, and we can confirm this is most certainly the case. However, as nasty as that noise is, it’s also a reassuring sign of life, because it means that the mains and output transformers are probably fine. The control pots generate loud scratching sounds, and powering down is accompanied by a prolonged hissing squeal.
The first task is to find where the noise is coming from so first off, I cut the power, pull all the valves out and switch the amp back on. As expected, the only sound coming out of the speakers is a very low level hum. Next, I proceed by powering down, reinserting one of the power valves and turning the amp back on. This time there’s a bit more noise coming from the speakers, but well within the norms for amp hum.
Things remain fine with the second power valve installed, but returns with the ECC83 back in position, which means the problem has to be something in the preamp. Changing the valve is always the first option, so I install an ECC83 that I know is good and the noise remains – so no luck on this occasion.
Extreme caution is required for the next procedure, because it’s quite dangerous. Placing one hand into my back pocket I hold a long wooden chopstick in the other hand, with the amp switched on, I begin tapping the components with the end of the non-conductive stick [don’t try this at home! – Health & Safety Ed].
The 220k plate resistor and two of the tremolo capacitors are all extremely microphonic, so I decide to replace the lot. I also notice that the wiring looks like a rat’s nest and, in addition to being in amongst the other wires, the heater wires aren’t twisted.
After twisting, tidying and replacing the wiring and suspect components, noise levels drop considerably. The squealing during switch-off is also cured, but I still feel it has more hiss and hum than I would like. However, in this regard, the 1930 is simply living up to its reputation as a rather noisy amp, and the root cause is the circuit design itself.
The fundamental issue
Even with the wiring sorted and the components in tip top condition, this would never have been a ‘quiet’ amp. The issue is that the volume and tone controls are situated between the screen grid resistors and the grid of valve. What’s more, the two channels are joined by a couple of 470k mixer resistors, which attenuate the guitar signal before it even reaches the valve. The volume and tone controls of valve amps – including most Marshalls – should come after the first valve stage. Each valve stage generates noise, so by placing the volume control after the valve you can turn down the noise and your guitar signal simultaneously.
Most of us are prepared to accept noise when the amp is cranked to the maximum. Hiss and hum will be apparent, but the guitar signal will also be at its loudest, so the signal to noise ratio is acceptable. With the 1930’s circuit, all the noise being generated is passed onto the phase inverter and the power valves unchecked. You can turn down the guitar signal but so far as the noise is concerned, it’s as if the amp is running at full volume at all times. What’s more, placing all the extra components between the volume control and the grid ensures that a lot of unwanted extra noise is picked up.
The only way to sort the noise issue completely was to connect the screen grid resistors directly onto the grid and reconnect the control circuitry between V1 and the phase inverter, however there’s a slight complication.
The 1930 was an amp built to a price rather than a standard, and installing one fewer valve stages was a way for Marshall to lower costs and reduce build time. This really should have been a one-channel amplifier with tremolo or a two-channel amplifier without tremolo, but Marshall decided to have both.
Consequently they had to place all the controls and the passive mixer components straight after the inputs so both channels could share the same triode stage. This arrangement left the second triode stage available for use as an oscillator.
Rather than make one set of inputs redundant, I decide to make use of the blanked off preamp valve socket to install an extra valve, meaning that the 1930 could be a genuine two-channel amp with tremolo on one channel. There’s also the potential to voice the non-tremolo channel for a different tone with perhaps a bit more gain.
Shifting the stack
Turning this 1930 into a bona fide two-channel amp actually proves to be quite straightforward. I chat with Chris Fantana of Rift Amps and we decide that the mains transformer was probably rated for a 2A current draw for the valve heaters, so it will be able to handle one extra ECC83. To keep things simple, the new channel will henceforth be referred to as the ‘normal’ channel and the other will be known as the ‘tremolo’ channel.
I begin by disconnecting the all four 68k grid resistors from the volume pots then I disconnect the 10nF coupling capacitors from the 470k mixer resistors. The 470k mixer resistors and the 1M resistor tying the grid (pin 7) of V1 to ground are all removed from the circuit board and the 68ks of the tremolo channel are soldered together and attached to an extension wire to connect to the green wire leading to Pin 7.
The 500pF connecting the anode (pin 6) of V1 to the grid of the ECL86 triode stage (pin 1) and 470k resistor is then lifted at the ECL86 end and soldered directly onto the volume pot. With the 10nF capacitor leading from the output of the tone pot soldered onto the ECL86 Pin 6/470k junction, the relocation of the controls from pre-V1 to post-V1 is now complete.
Adding a channel
I now have a proper tremolo channel and the amp is running quieter than ever. However, I also have two inputs leading to nowhere, and a couple of redundant controls. The spare preamp socket hole is blanked off with a disk-shaped metal plate held onto the chassis by two bolts.
This is easy to remove and I install a new valve socket. Just in front of this socket there’s a preexisting hole in the chassis, so I use it to mount a short length of tag strip for the new valve connections. Since no extra hole was required, all these mods could easily be reversed if anybody wanted to return this amp to the circuit and layout that it left the factory with.
Only one half of the extra triode is required so I configure it like a 1974 18-watt Marshall with a 100k plate resistor connected to pin 1 and a 1K5 cathode resistor from pin 3 bypassed with a 25uF electrolytic capacitor. The 68k grid resistors are tied together and extended to fix onto the grid (pin 2) and pin 1 is connected to the volume control via a 0.0047uF coupling capacitor.
Crossing the channels
My initial plan had been to connect the new channel to the 470k resistor on the other side of the phase inverter. It works, but sounds terrible – with low gain and excessive treble roll off. The issue is no doubt related to the 22nF capacitor tying the grid to ground, and lifting the grounded end certainly restores the gain and treble. However it also introduces a load of noise and wild oscillation.
Talking it over with Chris Fantana, he suggests the capacitor was probably a ‘band-aid’ used as a quick fix for an oscillation issue with the circuit. After trying unsuccessfully to remedy the problem, I reconnect the 22nF capacitor and join the channels via a pair of 220k mixer resistors on the other side of the phase inverter.
Finally everything was running properly and, as I’d hoped, the new ‘18-watt’ channel sounded fatter, deeper and warmer than the brighter and more aggressive tremolo channel. Of course the tremolo itself still isn’t working and I’d like to repair the broken cabinet plate, but I’ll save that for later. All things considered, the Marshall 1930 in stock form isn’t a particularly good amp, but with a few minor tweaks it can be transformed into something pretty amazing!