My 8 Watt Single-ended El34 Guitar Amp, The Bluesman

The Bluesman

The perfect bluesman guitar amp?

by Bob Richards, all rights reserved.

There has been another birth in the family, The Bluesman. I built a twin amp style tweed cabinet somewhat modeled after the 1959 Fender Twinamp but scaled down slightly to have 10-inch drivers instead of 12's, and built up a new chassis with a newly brewed EL34 single-ended 8 watt circuit.

It was a winner at a recent jam session. The sound is addicting. Warm, creamy, bluesy, musical, and very soulful. It has the best-sounding distortion of any of my amps, and maybe the best I've ever heard out of any amp.

The "twin amp" speaker configuration somehow sounds better to me than a single driver. Less in your face somehow. More ambient. It would have a significantly different off-axis frequency response (and different to each ear). Plus I love the look of it.

Are 8 watts really enough in the real world?

Practicing in your living room at home goes plenty loud IMO. Used in a recording studio where you want the best tube-type distortion for effect, it's plenty loud. In the real world where you may have several guitarists running on the more typical 35 watts or higher amps, and a 300 watt P.A. system, and especially with outdoor gigs... you'll want at least 30 watts. Since this amp weighs about half of what my 60-watt Stereo "Deuce" amp weighs, I take this amp for practice sessions and smaller parties, and the bigger "Deuce" amp for the bigger parties or outdoor gigs. Since the Line Out circuit has no connection to Ground, this amp can be used as a preamp for a much more powerful amp, thereby giving you an excellent distortion sound, at much higher levels.

More Detail:

Any balanced (symmetrical) topology generates way more of the odd harmonic distortion products than ever. The first harmonic is the fundamental frequency by definition - the note you played. The second harmonic is the first "even" harmonic and is the same note one octave higher. Its presence will enhance anything. It's always a good thing. The third harmonic (which is the first "odd" harmonic) is another half octave higher in frequency than that. It forms a "major" chord with the fundamentals. That is not desirable when the song is in a minor key. The higher harmonics are hit and miss as to whether they are pleasantly musically related to the fundamental (the note you played). An accumulation of higher-order harmonics means an accumulation of "notes" that are not pleasantly musically related to the fundamental. They cause more of a fizzle-type sound. Also, the further away in frequency a distortion product is from the fundamental, the less it will be "masked" by the fundamental.

A single-ended topology will generate a second harmonic that is stronger than the third harmonic. Also, there won't be any kind of crossover distortion (where one tube turns off and another turns on, at the zero volts transition feeding the output transformer push-pull). It's almost impossible to get a symmetrical clipped waveform out of a single-ended topology, so the 2nd harmonic will almost certainly always dominate over the third harmonic distortion product. That's what sounds best to me.

I've been told by the tube gurus that paralleling tubes in the SE topology doesn't sound as "musical". I'm pretty suspicious of that statement. Most of what I've found on the web while researching all this is written by people who don't have all the facts and/or don't realize all the different interactions when they try something. There are many claims from experiments that were riddled with variables that weren't addressed. There's also the strategically unbalanced push-pull which would give you about 35 watts out of a pair of EL34's or 6L6's. That should be a close second. For this amp though, I wanted the highest potential of the musicality of distortion, with no compromise whatsoever that I know of. If I need more power, I can run the Line Out to another amp or P.A. system for a boost.

Design Goal:

The best possible vintage blues style tube distortion in a package that is fun and practical. Able to run relatively clean to usable levels as well. High on second harmonic, fast roll off of higher harmonic distortion products, a maximally gradual and complex overdrive transition, the projection of the twin amp style speaker configuration and open back cabinet, very few knobs or switches on the panel, at least ten watts, big enough to be warm but small and light enough to drag around to jam sessions without getting too annoyed, no hum, minimal noise, the line out circuit, good Rf filtering at the input and AC, very cool look.

The EL34 has a great reputation and will always be easily available. It's the most gradually distorting tube I know of - with one of the best track records in the guitar amp world. With two of the highly efficient vintage reissue blue-back Jensens with the alnico magnets, the P10-R, (from Antique Electronics Inc.), the acoustic power is quite impressive. The first four watts are relatively clean for a guitar amp unless you overdrive and then attenuate the 12AX7 stage using the Master volume control. At a recent jam with this amp in a living room of a medium-sized house, the volume knobs were nowhere near all the way up. It went way louder than I wanted, and sounded excellent at the level I did want. I set it so the 12AX7 "full-wave saturator" stage is just slightly distorting, then adjust the master volume for how loud I want that.

The cabinet is all solid pine, covered with tweed bought from Antique Electronics in AZ. I read somewhere that Eric prefers the sound of solid pine.

Since digital reverbs, compressors and etc. are so abundant and specific to each guitarist's tastes, I didn't see any reason to include any sound effect processing in the amp cabinet. Just pre and post volume, and bass and treble. Most of the built-in processors I've seen were not very impressive. I have a Boss SE-70 stereo digital guitar effects processor from the early-mid 1990s that sounds exceptional.

The front end circuit (6SJ7 pentode) is similar to the Octal Fatness circuit as well as many other amps from the early tweed era (Fender and Gibson). I ran a higher B+ in the frontend sections so there would be more clean gain when I want it. In the Octal Fatness version, you couldn't get a very usable amount of clean acoustic power out. Now it gets pretty loud clean and has a really nice palette of distortion coloration that comes into play when you manipulate the Input Level and the Master Level pots. A distortion that I actually like.

The above photo shows most of the lug strips (not all) that were needed. The bridge rectifier shown above is for the 6.3VDC filament voltage. It's bolted down to the chassis for heat sinking, directly under where the power transformer will be mounted.

3 Weeks Later:

The high current electromagnetic field producing power supply and output circuit is as far from the high impedance input stage as is practical (a must). All grounds return to the star center, which is 100 ohms off earth ground/third prong of the AC plug and chassis (100 ohms bypassed with a 2nF mica cap). The AC cord has an additional clamp inside the chassis, in case someone trips over it. The regular "strain relief" grommet thing is a bit meager by itself.

The 6SJ7 pentode tube (not in current production) comes in either black metal or glass. Both are excellent. It was the input tube of many guitar amps in the late 1940s Including the very first Fender tweed amp called the Dual-Professional, later renamed the Super. The 6SJ7 seems to contribute a sort of "woody" quality to the coloration. Heavy on the second harmonic, at the expense of intermodulation distortion (you need the right balance). This input tube also acts as a microphone (a three grid microphone) to some extent, picking up vibrations, both radiated and conducted, from the speakers. This is a form of feedback and contributes a slight but complex resonant coloration (not necessarily desirable but might be). The 12AX7 can be independently overdriven by manipulating the input level control and the master volume control. With all the research Fender has done over the years, they are still choosing to use the 12AX7 for their overdrive stage of any of their latest amps (many other brands as well). It seems to be a very good compliment to the distortion produced by the 6SJ7 and is a tube that will always be available. The EL34 contributes an early-Marshall soft crunch at high levels, but with the better harmonic distortion spectrum shape (more second harmonic and less higher-order harmonics) that you get with a single-ended topology with no feedback.

You can see the bias pot for the EL34 protruding south, just to the right of the EL34 tube socket (above). It's wise to locate it as close to the control grid of the EL34 as is practical since it's a high impedance node that could be susceptible to picking up stray fields. Ceramic sockets move the heat from the tube into the chassis the quickest (a good thing). There's a jack for a fan to be hooked up to the DC filament supply. Since the EL34 is biased ACTUAL class A, it dissipates almost 24 watts continuously, signal or not. Since it's mounted upside down, I felt that a fan would be a good idea (extend tube life substantially).

You can see the position of the fan above. By running a 12-volt computer fan on 6.3VDC, it's a lot quieter. The fan stalls at about 5V. It's shock-mounted to reduce noise also (rubber grommets).

On the right-hand end of the chassis (above) is the turbo unit (EL34 bias supply toroid). I didn't want to lose ANY of the B+ to the bias function (roughly 20 volts). It's a squared term in the power formula, so makes a big difference in final watts out. I had already bought the power transformer. It's the same one I used on my five-watt "Fat Mama" guitar amp.

A simple panel is minimally distracting when you want to stay rooted in the musically creative part of your brain.

I had the panel paint color custom mixed. I call it Caribbean blue.

The Line Out can be used to drive a second more powerful amp, a P.A. system, or whatever. The Line Out signal is taken right off either the speaker or the built-in load resistor if the speaker is unplugged. It has 5 volt Zener limiting and an Rf filter as well (to minimize its ability to damage an input stage by overdriving it). This way, you get the great sound of a single-ended topology, and if needed you can jack up the level with a bigger amp (for an outdoor festival). Since the bigger amp will almost certainly be a push-pull topology output stage (odd harmonic distortion generator), it's arguable that you'd be better off not overdriving it.

You just remove the upper rear panel to access the innards. I bought the Fender tilt back arms for the sides, so I can tilt the amp back for better sound. After using them at a jam, I like them a lot. Available at Antique Electronics Inc.

This amp also makes a nice seat for the guitarist at a jam when things are really cramped. Or a place to set things. I didn't realize how dark the amber shellac would be on the standard tweed covering. I then tried to remove/reduce it with rubbing alcohol on one of those red mechanics rags that was handy. The red dye in the rag dissolved and tinted the amber more reddish, which I like after all. Next time I will dilute the amber shellac with clear shellac (one-part amber to about three parts clear) before applying. I might even add some red dye.

Notice that the secondary of the output transformer has no connection to the Ground. This is so the line out can drive other bigger amps without there being a ground loop issue (potential hum problem).

The output transformer is wired such that the 16-ohm tap is driving the 8-ohm speaker load. I wanted max voltage swing and actually preferred to have some dynamic range compression, which will make the amp sound mellower, and give it more sustain.

In the upper-right area of the schematic, you can see the 100-ohm resistor on the B+ feed. That gives me the amount of "sag" that I might want, which most people seem to think requires the use of a tube-type rectifier. Only a fool would actually use a tube-type rectifier. It's a sales gimmick. "Sag" is strictly a function of the on-resistance of the rectifier tube working with the source current capability (limited by the DCR of the high voltage winding of the transformer) and the filter cap just downstream from it (22uF). That's what the circuit sees. The circuit has no way of knowing whether the resistance is a resistor or a tube since everything above about 70HZ is filtered out by the 22uF cap. The lowest note on a guitar is 80 HZ.

There's a belief that using all very expensive Teflon or styrene caps ($35 each) and very expensive custom transformers (>$100), and even Teflon insulated silver wire will sound better. I disagree. One boutique amp manufacturer even offers a gold plated chassis. These are in my opinion all sales gimmicks for the rich fool. Fancy transformers could actually sound worse depending on impedances and saturation rates. Regular polypropylene caps ("orange drop") are plenty good for this and most hi-fi applications in my opinion. They are only a buck or two each. Metal film resistors are the lowest noise types, they're very cheap, and should be used in the front end and 12AX7 stage wherever the value is over about 5Kohms. The rest of the resistors should be metal oxide for the sake of reliability (they handle overloads the best). Wirewound is fine in the power supply. I bought all the parts at Antique Electronics Inc., or Digikey, or Mouser. For Rf filtering, ceramic and mica go the highest in frequency before fizzling out. Mica is the better one when the value is available and they are also very cheap. I chose to use higher than needed voltage ratings on many of the caps just in case elements in a tube were to short out and cause the much higher voltage to occur temporarily. That actually happened in one of my amps. Tubes are funky, especially the old ones, and will occasionally fuck up. It's arguable that spare tubes should be stored in the back of the cabinet, in a shock-absorbing containment device.

The original tone circuit, a fender "stack" seemed a bit meager in actual use so I decided to go with a Duncan version of the Fender e-type tone topology which looked good in the Duncan analysis program below (it's a freebie that plots the curves - avail. on the web).

This e-type circuit is very similar to the circuit in a 1957 Fender e-type Twinamp; the one Eric Clapton has been using exclusively since about 2000. I like it more. It's also less lossy than most topologies, which was a critical issue in this circuit. The signal coming out of the tone stack was barely large enough to drive the EL34 into clipping and get the full 10 watts.

I never hooked up the footswitch/relay circuit. I can add it later if I decide I would actually use it. I might use it as an on/off for the line out. At a festival, it could turn on the connection to the stereo reverb and P.A. system for a lead solo (or a Leslie amp if you're playing the song Badge by The Cream). The power supply circuit in the lower right of the above power supply schematic is the preliminary bias supply for the EL34. It's nothing special. The additional diode after the bridge is redundant and not used. It's such a simple and basic circuit I was too lazy to draw it out beyond that. I just built it.

Results:

Click Here For Spectrum Analyzer Test Results

It turns out to only put out about 8 watts RMS into my 10-ohm load resistor, but it's got the tone control circuit Eric Clapton has in his amp and a real nice overdrive distortion sound. The Jensen P10R speakers might be my favorite guitar speaker out there. I hope this inspires other hobbyists to make their dreams come true. It was a fun project and a grand success.

Happy Jamming!