Epiphone Valve Junior Modification

The stock Valve Junior is a simple basic amp with a gain control. The fact that they could build and sell it for ~$100 is pretty impressive – a great value for the money. Of course, it lacks some features (tone controls, effects, etc) and some design corners were cut to keep the cost low. The amp is a heroic result for the price, but it has a dark tone to it, no tone control, and almost no bass response. The low end is more buzz than growl. For a bit of effort and money to cover the mods, the amp can be made to really sing beautifully, competing favorably with many higher-price amps out there. We get a good start by using the Epi platform, which is a good one for making mods, as indicated by the many sites posting them (see http://duhvoodooman.com/VJr/VJr_mods.htm)

The following outlines a number of changes I’ve made to the amp. The focus is on the two goals I outlined in the discussion of tube amps. First, I enhance the harmonic content by allowing the tubes to operate in large-signal mode, even at low speaker volumes. Second, I enhance the frequency response to reveal more of what the guitar is producing. The mods are described as a series of separate changes. You should feel free to pick and choose among them. Not all changes are important to everyone, including me. I’d probably only do a subset if I started over again. I admit, I got carried away, but when you start something like this, it is tempting to throw in all your ideas into a one big redesign/rebuild, since it’s difficult to add them later piecemeal.

The stock schematic is attached at the page bottom as a downloadable PDF file. My modified version is shown below. Click on it to see or download a full size version.

A note about my schematics… all the unmarked parts are my additions. Any parts with R or C numbers (e.g., R3) are in the stock design, although I have changed a few values. Let’s walk through the mods...

The stock power supply circuit is minimal and has insufficient filtering. Other sites have pointed this out. Since we will expand the bass response of the amp, this is important to fix. I added two new filter caps, each 470uf/450v, which is probably overkill. I suspect one would have been enough – between the diode bridge and R10. Both are large cans and need to be attached to the chassis (straps, clamps) and wired to the circuit board. The standby switch is optional, you can omit it unless you just like switches on a panel. Otherwise, the power supply remains unaltered.

The input resistors (R1, R2) are modified. I use the values common in Fender amps. Higher values enhance the high-end harmonics from pickups. Also, the changes don’t cut the signal level in half, as the stock values do.

R7 is superfluous – remove it. It only reduces gain, which is not something you can spare in a simple 2 tube amplifier.

Tone control is provided by two SPDT switches. The switches offer neutral (center off) or two boost settings. The two boost positions per switch move the 3-db point higher or lower, but the db/octave slopes remain fixed. This is a simplification over the more common use of potentiometers in tone stacks. Feel free to use any tone section you like. Good examples are shown at http://www.duncanamps.com/tsc/. I personally like the switches since they are fast and unambiguous, like PUP switching, and each RC network can be customized with a different knee frequency. If you’re adventurous, try different caps in both the bass and treble sections. My bass choices probably put the knee too low for most tastes.

There is a second fixed treble and bass boost section following the tone switches and the coupling cap C2. These are formed by the 680pf/1Mohm and 2500pf/R5 networks. Again, you can easily alter the cap values to suite your taste. I picked them based on the response curve I measured in the stock amp. The idea is to compensate for any high/low roll-off that occurs in tube circuits and to give the amp the high/low boost and mid-cut found in most amps.

The power tube (EL84) has a cathode bypass cap (C5) that is too small and cuts the bass significantly. Replace it with a 100uf cap of same or higher voltage rating.

Perhaps the most important (and expensive) change to the circuit is the replacement of the output transformer. There are only a few transformers easily available anymore, and you may want to put some effort into searching for a lower cost (and smaller) alternative. I found the Hammond 1628SEA as a great choice, in terms of sound, but it is a bit pricy and physically large and heavy. (See http://www.hammondmfg.com/1627.htm) If you ever wonder why tube amps are heavy, the transformers are a big part of the answer. Size and weight matter with transformers. If you want good bass response and clean signal passage, you need iron (literally) in the core. The Hammond is a beast and it does both beautifully. The transformer easily pushes 30Hz through and it won’t saturate at the power levels the EL84 creates. The result is smooth low tones as opposed to the raspy fuzzy bass the stock amp provides. Put some money in the iron – it will pay off.

If you have cost constraints, have a look at the 125ESE or similar transformers at http://www.hammondmfg.com/pdf/5C_125SE.pdf. Use the 5000 ohm options. I have to stress that I did not try any of these lower cost transformers, so results may be different (likely). If anyone does try them, I’d like to hear from you on how it turned out.

The Hammond 1628SEA also has a nice feature used in hi-fi applications, the screen grid tap. The stock transformer doesn’t have one so the screen (pin 9) is simply tied to a B+ tap. That makes the EL34 work as a triode and adds non-linear distortion to the output. This is OK as the distortion is the large-signal variety that we actually want – it's why we use tube amps. However, since the tap is there, I added a switch to allow the choice of tying the tap to the screen grid or fixing the screen voltage. The stock fixed-screen option sounds a bit dirtier and more to my personal liking, but in some cases the ultraclean sound you get from using the tap may be desired. The tap provides negative feedback, so there is a gain drop when you use it, but there is enough in the rest of the circuit to make up for that. The feedback keeps the EL84 pretty linear until you hard-overdrive it, at which point the breakup sounds quite nice and smooth. Think of this switch as giving you two channels to pick from – regular and superclean.

The Hammond OT can drive the 4, 8, and 16 ohm speaker outputs. I use mostly an 8 ohm cab, so I modified that output to include the power control system. Recall, the non-linear effects of the circuit mostly kick in when the tubes are driven hard. This applies especially to the output tube, so you really control the tone quality of the amp by how hard you push that tube. In practice, you’re probably going to push that tube hard most of the time to get the richer more harmonic tones from the amp.

To keep your neighbors happy when practicing, or in small sessions, you’ll need to soak up some of that power and send only a fraction of it to your speaker. This is what passive loads do (see THD Hot Plate). External attenuators are common and recent amps are now adding them into their designs. The THD Univalve already include one, and this mod adds one to the Epiphone. The proper part to do this is called an L-Pad, and after some searching, I recommend the unit at http://www.parts-express.com/pe/pshowdetl.cfm?&DID=7&Partnumber=260-252&ctab=1#Tabs. It’s rated at 50watts (more than enough) and it fits nicely (but just barely) into the Epi chassis. There is also a cheaper unit rated at 15watts at http://www.parts-express.com/pe/showdetl.cfm?Partnumber=260-248, and it should fit easier, but I’ve not tried it myself.

There is a DPDT (center-off) switch shown in line with the L-Pad. The purpose of the switch is two-fold. First, if you ever do want to hook up a 4 or 16 ohm speaker to those jacks, you’ll want to disconnect the L-Pad, since it will otherwise be loading the 8 ohm transformer output. The center-off position completely disconnects the 8 ohm tap. One switch on-position connects the L-Pad, which feeds the speaker jack. The other switch on-position feeds the L-Pad through another divider network, which lowers the power fed into the L-Pad, giving you a bit more control-range at very low volumes. You could omit this option and simply using a SPST switch to connect the L-Pad or not.

Another interesting, but completely optional mod is the 915 lightbulb and the SPST switch that bypasses it. I’ve seen lightbulbs used in THD amps and others for various purposes. (The THD design purpose is backwards, IMO.) The point is to exploit the variable resistance of a light bulb filament to achieve some goal. Filaments have a low resistance when cold (dark) and their resistance increase when lit. So, my idea is to use the bulb as a signal compressor.

At low volumes, the power flows through the bulb almost unaffected. As you play louder, the bulb begins to light and its increased resistance soaks up some of the power, thereby reducing (compressing) the range of volumes you hear from playing soft and hard. This is essentially what expensive signal compressors (like the TL Audio 5150) do. It also mimics the effects of power supply sag, which doesn't normally occur in this small class-A amp. This was an experiment and it does work, so for the cost of the bulb and switch, I’d suggest trying it. You’ll have to play around with volume and L-Pad settings to get the bulb to light during louder playing and stay off or low for moderate/low passages.

That pretty much covers the electronic modifications. Now to the physical issues of mounting and wiring. Fortunately the Epi VJ comes in a big box. There is lots of room in both the wood box and the metal chassis. If you do all of these mods, you’ll pretty much fill the faceplate, chassis, and the wood case. I don’t think you can get much more into it!

The biggest component is by far the output transformer and mounting it is something of a challenge. After removing the stock output transformer, I chose the right side of the chassis, near the power switch. Fortunately, there is a large open space on the chassis, however this location is near the power transformer and some of its magnetic field actually couples into the output transformer so there is a very low-level hum coming from the speaker, even with the volume set to zero and no input. Fortunately, it’s barely audible and it doesn’t change as the volume control varies or input cables are added (it's literally coupled outside of the circuitry). The hum levels you normally get from guitar pickups and cables are far greater.

Most of the resistor and capacitor components we added will need some support on the circuit board. I mounted some header strips in open spaces on the circuit board to hold parts and others are soldered directly to stock parts they connect to or the switches added for tone control. I left the volume control, power switch, and input jack where they were. The L-Pad is large, so make sure you leave room on the panel and behind it for clearance to the circuit board. The other switches and lights are distributed on the front panel to give a balanced look. The screen grid tap switch was added as an afterthought, so it’s mounted on the back panel of the amp in my case.

If you are interested, below is an attachment that is a PDF layout (1:1 scale) of a new front panel with labeling (VJr2 panel.pdf). I had one made at a sign shop, where they make thin brushed-steel signs for doors. A word of warning if you want to try this… while the cost of the panel is low (~$45) and it looks great, machining thin steel is brutal, especially for the rectangle power switch cut out. The steel is hard to drill and cut without bending and creating wrinkles. You may want to consider other options or just repaint the VJ chassis panel. Here is a picture of the final modified amp with the preamp and effects unit described in the gear section.

So, what can you expect once you’re done? I'm pretty happy with the sound. Played clean, the lows are so smooth and solid, and they don’t resonate and “boom” like with some amps. If I boost the treble, I get a nice chime out of high notes, while the lows stay clean. This is really hard for some amps to do. Strum a chord and the high notes shimmer while the bass strings are solid and clean and strong. With max treble boost and some breakup, I get nice harmonics. The range of clear tones and harmonics is many times greater than the stock Epi. That Hammond transformer pays dividends in so many ways.

When you crank up the volume control a bit and back off the L-Pad, the audio level stays the same but you start to hear the EL84 break up. The lows start growling and there is nice ringing and shimmer to high notes. Push the amp hard if you want to – it keeps adding harmonics, but the notes stay audible and distinctive. It doesn’t turn into noise and fuzz, which is the case for the stock Epi. Compare this amp tone to many amps under $1000, even the Univalve described on the other page. The Epi mod low-end response in fact exceeds that of my Univalve and Vox amps, and although each amp’s tone is a different, the Epi holds its own. It has a unique tonal quality. No two amps are exactly the same, especially when the designs are so different, and it’s great to have options. The Epi mod adds something unique to my collection.

Update 4/18/2011

I finally got around to creating some audio clips for this amp. It was a bit of a challenge since I could not get a decent recording that really reflects the sound I hear from the amp. Finally, I got a JDI direct box and that works pretty well. So, here are three clips of me just noodling around -- ignore the playing, but you can hear the range of sounds.

The setup is me playing a Deluxe Telecaster (volume and tone set to max). I'm just rambling, trying to demo the tone and range of the amp. The tele feeds the TLA5051 which adds just a touch of compression and flat tone settings. The TLA feeds the GSP-1101 which is set to clean (no modeling), flat tone controls, and a bit of reverb and tremolo. The GSP output runs to the right channel of these recordings and to the Epi input.

If you listen to just that right channel, you'll essentially hear the amp input. It's always clean regardless of the volume and breakup of the Epi. The Epi 8 ohm output feeds a 10" cab (Raezer's Edge). The 16 ohm Epi output feeds into the JDI box with the 15db Pad and the 30db Speaker switches on. The JDI output feeds the left channel of the recording where you'll hear a pretty good version of what I hear coming from the cab while I'm playing.

The clips are below. This was all done as one long recording that I broke into clean, breakup, and hi-gain clips. Recording software is Audacity 1.3 via an M-audio Fast Track Pro USB interface.