For several years I've wanted a nice, very low power, all tube stereo guitar amp with "big amp" features like a full tone stack and an effects loop.  I stopped holding my breath waiting for Fender or Peavey to make one a long time ago.  More recently I've given up on the hobbyists, too.  I have finally decided that if I am ever to have such an amp I am not only going to have to build it myself, I am going to have to design it, too.

The BloozBaby is the project that will hopefully one day become my dream amp.  I am just now ready to begin ordering parts to build it, and it will probably be at least three months before the amp is built.

I decided to document the project here because there are probably at least one or two other people who are interested in such an amp.  If you are qualified, you're welcome to build one for yourself, but please keep in mind that right now the schematics have to be considered preliminary because I haven't yet prototyped the amp and I fully expect that some tweaking of coupling and cathode capacitors will be required to optimize the tone once the first amp is working.

Design goals.  It makes very little sense to go to all the trouble to design and build something you could probably buy more cheaply unless you know exactly what you want from the finished project.  Here's what I decided was important to me:

• A single-input preamp with a versatile tone stack.
• A stereo effects loop (1 ch out, 2 ch in) capable of driving a fairly low impedance (10k) at line levels without going distorting heavily.
• A drive or "master volume" control to permit independent control of the preamp and output stage.
• It must be possible to overdrive the power amp before overdriving the preamp.
• Approximately 3 to 5 watts per channel output.  The goal was really one watt per channel, but I didn't think I could get there while driving the power tubes hard enough to sound good.  I'm now designing another amp, the Tuppence Blues™, which may solve that problem by using a "new" type of power tube.  That project may also be a miserable failure because, as near as I have been able to determine, no one has ever tried to build an overdriven guitar amp with this tube.  If the tube refuses to break up sweetly that project will be nothing but a cautionary tale...
• It has to sound good (I've heard some homebrewed small amps that sounded like someone banging pots in the kitchen).
• It has to sound good when clean, lightly overdriven, and driven hard.  I don't want a "one trick pony."
• It must be reliable, I don't want to make a career of fixing it.
• It must use only readily available parts, nothing scrounged or salvaged that I might not be able to find if I want to build another amp later.
• I have to be able to afford to build it!  You might surmise from the fact that this item is last in the least that saving a buck isn't the primary objective, and you'd be right.  Obviously, I have to be able to afford to build it but if I have to cut corners so much that the amp doesn't meet the other objectives then it's a waste of time.

	CAUTION! Working on tube amplifiers can be extremely dangerous! Do not attempt this project unless you have been trained to work safely with high voltages.  This project works with AC line voltages and with DC voltages approaching 400 volts - either of which is easily lethal! Even if you do not get shocked while working on the project, if you do not build it correctly you or someone else can be electrocuted later when simply using the amp.  Even if no one is ever injured, an improperly built amplifier can cause a fire that burns down your home. If it sounds like I'm trying to scare you it's because I am.  This project is intended only for those who are qualified to accomplish it safely.  For the rest of you, the best amp in the world isn't worth killing yourself over! Use at your own risk!

Keeping in mind that these are preliminary schematics and that I haven't built the amp yet, here's an overview of each section of the amp, starting with the power supply.

The power supply is overbuilt, leaving plenty of voltage to spare as well as plenty of current in both the HV and filament sections.  This was both intentional and necessary.  Intentional because...

• I want enough spare voltage and current in the HV section that I can experiment with other output tubes.
• I want enough spare current in the filament section that I can add another tube or two as the design evolves without rebuilding the power supply.
• I want to be able to drive the output tube at a little higher than rated voltages.
• I want very heavy power supply filtering.
• I don't want the power transformer to run hot.  Hot running transformers tend to break down in time and, even if they don't fail completely, they sometimes begin to buzz or vibrate under the high load.

As it turns out, this "overbuilt" supply is kind of a homebrewer's dream.  This supply can power any class A or AB tube amp from 360V to 230V (or even less) at up to 130ma (using the FX transformer) merely by changing the value of R7 and R8 and inserting an appropriate zener diode between the HV center tap and ground (when necessary).  (When is the zener necessary?  It shouldn't be needed for class A single-ended amps using 300V or greater at the B+ output of the supply.  For Class AB push-pull amps you should not drop more than 50 volts across R7 and R8 combined, so a zener will be needed whenever the voltage to be used drops below about 335V.)

Note that an extra current reserve is a good thing, the more current reserve you have the cooler the transformer, rectifier, and filter components will be.  Note, however, that there is one disadvantage in using a PS with a large voltage reserve – voltage changes on the mains will be "magnified" on the secondary more than if a transformer that is "just big enough" is used.  This happens because of the greater turns ratio of the secondary and occurs even when a zener diode is used to reduce the voltage.  For that reason, I wouldn't recommend building a relatively low-voltage amp with a high-voltage transformer and zener diode except in those instances where you are uncertain whether the low voltage amp is going to perform well and you want to be able to re-use the power supply for a "bigger" amp if need be.

Beginning at the left of the schematic we have a standard 3-prong plug and power cord.  The earth ground is attached to the chassis.  The hot lead passes through a 1 amp slow-blow fuse and both leads are switched.  L1 is optional, if used it should be any 120V pilot light.  On Feb. 14th I also added the 250V MOV across the primary of the transformer to provide some protection to it.

The Hammond 270EX transformer supplies up to 125ma on the 275-0-275 HV secondary and 4 amps on the 6.3VAC center- tapped filament secondary (we need 300VDC @95ma for a fairly conservative EL84 output section and we need 2.2 amps on the filament secondary for the four tubes in this design).  This transformer provides a lot higher voltage than we really need, so we can supply some very heavy filtering.  Note that we couldn't do that with a class AB or even many class A push-pull amps but in this class-A single-ended amp there is very little difference in average current between idle and a full signal so we don't have to worry about the voltage sagging because of the voltage drop across the series resistors R7 and R8.

All that extra voltage also means that we can experiment with larger power tubes and drive them at nearer their rated voltage than would be possible with the "ideal" (for an EL84) 190-0-190 secondary.  And, the extra current on the filament secondary means we can easily add a tube or two to the chassis without overheating the transformer.  Finally, note that this transformer also has a 5-volt secondary for those of you who might want to experiment with a tube rectifier.

The Hammond 270FX, at 150ma on the HV secondary and 5 amps on the 6.3V secondary would be an even better choice if you are certain that you are going to want to use a larger output tube.

On Feb. 14th I removed the doubled up rectifier diodes.  It didn't seem right that the PIV could be 1.2kv on this circuit so I went back and recalculated it and it is actually 770V so one set of rectifier diodes is all that is needed.  The .01uf caps across the diodes protect them from transient spikes when the amp is turned on and off.

Switch 2 removes the ground reference for the HV supply, effectively muting the output.  However, keep in mind that this does not render the interior of the amp "safe!"

The first filter stage is intentionally fairly wimpy.  Too much capacitance here can overwork the transformer even under no-load conditions.  We use a pair of capacitors in series (C5 & C6) because transient spikes when the power is turned on and off can go well over 450 volts at this point.  Resistors R5 and R6 form a simple voltage divider to ensure that each capacitor sees only half the voltage.  These resistors also serve the very useful purpose of bleeding off the large filter capacitors when the amp is turned off (but you should never rely on this - always check with a meter and/or short the caps to ground with a shorting stick).  Also, under the best of circumstances it will take quite a while for the charge to bleed off through such large resistors.  Because C5 & C6 are in series the capacitance of this input stage is actually only half the capacitance of C5.

Next come two stages of L filters.  Here we use large resistors to drop the B+ down to the desired 300V, which just happens to also improve the filtering over the smaller resistors used in most power supplies.  That, coupled with the very large 220uf capacitors, should give very low ripple even at fairly high loads.  Note that transients are not a problem here, so we use just one capacitor instead of two in series.

R9 through R11, VR1, and C9 form an adjustable voltage divider used to apply a DC bias to the AC filament heater voltage via the center tap of the secondary.  This reduces hum by moving the filament heater voltage away from the cathode voltage.  This circuit can put out a range of around 5 to 50 volts.  All of the cathodes except one are at 10 volts or less, the cathode of the cathode follower circuit that buffers the effects send is at about 80 volts.  This adjustment will let us put 30 to 40 volts of separation between the filament heaters and the cathodes of all stages.

My original intent was to build a DC supply for the filament heaters but as you may recall one of my goals was to design an amp that I could afford to build.  Building a 6.3VAC DC supply would have added significantly to the cost of the amp as well as to the weight and bulk of the amp.

The preamp, in contrast to the power supply, would be considered by some to be "underbuilt."  It is designed to push a clean signal into a 10k load at around 1.5 to 2 volts (i.e. line levels).  With this circuit it becomes very important not to clip in the cathode follower stage because clipping there is very ugly.  I found that I was able to build a clean preamp stage with a gain control and a full tone stack using just a single gain stage plus the cathode follower as a buffer.  At least, it looks good in Spice, we'll see how it sounds once I build it!

The goal for the preamp is that it be very clean and avoid clipping the cathode follower.  The idea being that if and when preamp distortion is desired one would use something like a TubeScreamer pedal to push the first gain stage into overdrive, then use the gain control to ensure that the strong signal doesn't drive the cathode follower into clipping.

As mentioned before, the power supply is overbuilt so it will be easy to throw in another 12AX7 if it turns out that another gain stage is really needed after all.  In fact, once the amp is prototyped and all the obvious kinks worked out I may decide to toss in another 12AX7 for a footswitched boost.

On the 10 February version I've added a 20dB attenuator.  Without the attenuator an input signal of about 1.5Vrms will begin to overdrive the first gain stage.  I added the attenuation switch so one can use a guitar with a hot preamp without overdriving the first stage, if desired.

On the 14 February version I've added notes concerning a different value for filter resistor R114 and different voltages expected when this preamp is used in the Tuppence Blues™ amp.  I also added C112 and C113 to filter sharp transients that are not always well filtered by the electrolytic capacitors.  These HV poly. film capacitors should be mounted as close to the tube as possible.

There's not a whole lot to talk about in this preamp.  The first gain stage is a typical 12AX7 circuit that has probably been used in a hundred different amps.  I originally put in a standard Marshall style three-band tone stack but upon further reflection decided to replace it with a Baxandall stack.  In spite of having only two controls this tone stack actually offers much better tone shaping capability.  I used values recommended at Duncan's Amp Pages. I wanted to go this route to begin with but I didn't think it would work well with such high impedance on the input and relatively low impedance on the output of the stack.  However, I simulated it in Spice and it turns out that while the Baxandall in this circuit is not as responsive as it would be in a more ideal environment it still far outperforms a Marshall style tone stack.

V101B is the other half of the single 12AX7 tube we use and it is configured as a simple cathode follower.  Cathode follower stages provide no voltage gain but they have a lower output impedance and will therefore drive a lower input impedance without clipping or sagging.

One noteworthy thing about this preamp is the wiring of the effects loop jacks and the presence of the "Dual/Single" switch SW101.  The switching jacks are wired such that if no jacks are used the signal will be routed internally from the send jack to both receive jacks, unless the Dual/Single switch is in the "single" position.  Thus, if we are not using the stereo effects loop (i.e. not using the loop at all or only returning to the left channel) we can choose to drive both halves of the amp or only one half, depending on how much volume we want.  Also, by switching to "single" but then plugging something into the right jack (regardless of whether we use the left jack) we can drive the right channel from any line-level source, i.e. we could plug a keyboard or what have you into that channel regardless of whether we used any effects in the loop of the left channel.

Note that when we elect not to drive the second half of the amp, the input will be grounded but the amp output section will still be powered up.  We cannot put just one-half of the amp into standby without radically raising the voltage to the other half due to the large filter resistors in the power supply.

Each half of the power amp is a single-ended class-A amp using a 6BQ5 (EL84) power tube driven by one-half of a 12AU7.

The 12AU7 buffers the effects return input and provides voltage gain to drive the power tube.  A line-level of 1.5V peak is adequate to drive the output stage very hard into heavy distortion.  The buffer stage does not go into clipping even with a signal of 2V or better.  If for some reason you were to have inputs that were well below line level you could replace the 12AU7 with a 12AT7 or even a 12AX7 for more gain.

The power section is driving the 6BQ5 fairly hard though not as hard as some circuits I've seen.  Some experimentation with the cathode resistors R207/R307 will probably be required.  I designed this relying on Spice rather than the characteristics curves from the old books and I'm not sure if the Spice model I have for the 6BQ5 is accurate in this bias region.  This should be well in the ballpark to safely test the circuit, at least since I'll be using some tough-as-nails Sovtek EL84M tubes until all the bugs are worked out of the amp and I'm confident that the bias levels are reasonable.  Then, I'll put in the better sounding but less rugged JJ tubes EL84s.

The voltages are a bit high for a Svetlana ECC83 but the screen is fairly "cold" so I bet that tube could also be used with nothing more than perhaps a slightly larger cathode resistor.

The transformer is a Hammond 125ESE, a "universal" output transformer optimized for SE stages (meaning it will handle higher DC current without saturating).  Note that this is not the world's highest-quality output transformer, it "bottoms out" at about 100hz.  It is about equivalent to what you would find in most production small guitar tube amps.  It does, however, offer the advantages of low cost and a wide range of impedances that can be achieved using it.  I intend to use this transformer for the initial prototyping and then, if the low end response should prove inadequate, switch to a better quality transformer of approximately the same impedance.

On 14 Feb. 2004 I added C207 and C208 to the design.  These caps are intended to filter sharp transients that aren't always well filtered by electrolytics.  Also, mounting these components as close as possible to the transformer and tube (respectively) will help reduce any noise that might have been picked up in the circuit.

General notes:

Note that this amp is inverting because of the uneven number of inverting gain stages (the cathode follower in the preamp is non-inverting).  It may be necessary to reverse the primary leads on the transformer to get good feedback sustain from speaker to guitar strings.

Chassis layout is the next step and I will begin doing that once I've received all the major parts, namely the transformers and filter caps.  I'm looking to save a buck or sixty over purchasing a commercial chassis, plus I want to build the amp as a "head" with a built in rack for the Digitech S200 stereo studio effects unit that I'll be using in the effects loop.  I'm planning a field trip to Home Depot to see what I can find in the way of aluminum U-channel that might make a cost-effective long, skinny chassis.

Once I start laying out components and building the amp I'll update this site with plenty of photos.  Until then, if some impatient (and brave) soul should happen to build this before I do, send along any tips you think might be useful!  Note, I'm putting this project on hold for a little while to build a "no-frills" amp using the ECC99 tube.  If that tube breaks up sweetly then I'll redesign this amp around that tube.

John