THIS IS A LONG POST ZOMG -> deal with it.
This amp WON'T:
- Play loudly in live settings
Get fuzzy on its own / have a drive channel
Have any built in effects or reverb
Levitate or produce perpetual energy
The goal is basically to provide something better, within the space occupied by a sea of cheap solid state practice amps that are utterly terrible, and higher-end solid state amps like the ZT Lunchbox. I've taken a lot of the critiques & complaints about the ZT Lunchboxes to heart while working on this amp. I'm not sure on what the final parts cost will be (or labor for that matter); this thread is essentially documenting an R&D process in which I'll buy a lot of stuff and weed out the parts that won't work.
Most of my parts are sourced from Parts-Express and amplifiedparts.com so far.
Power Amp:
To start off, I'm basing the project on a class-D power amp. These are solid state chip amps that are very efficient and much better than the transistor amps of old. That said, like all solid state amps, it isn't designed to sound good distorted, so the point of this project will be to have enough power amp wattage on tap and not distort there, but still be able to play quietly. You may have seen these Class-D amps marketed as a "Class-T amps" and the hifi people really took to them, even though the first few "Class-T" amps were cheapie plastic things with spring clip terminals.
The board I've chosen to start with (and may end up getting a different amplifier, but this one seemed like a good start) is the Parts Express TA2024 Class-D amp, which costs about $17. It's small, has passive cooling, can easily be powered by a wall wart. BUT, it's only 15 watts per channel.. and that could be an issue. That said, remember that we're optimizing for practicing in a house, not playing live gigs. Most likely I will be building a combo form factor where one channel of output is going to the built-in speaker, but the second channel can power a separate speaker cab, doubling the output power. (Remember, solid state amps can handle running without a speaker load, but TUBE AMPS CANNOT.)
Here's the amp board:
The other interesting thing about the Parts-Express Class-D amp board line is that they quickly go up in wattage and number of channels. There's a 4x100 watt board and a 1x600watt board, and both make it possible to take my basic design for this hybrid amp (maybe with more preamp gain sections) and build much larger heads. the problem becomes sourcing power supplies for these larger amps, as they quickly go up in power requirements. Luckily, the little 2x15w board I'm starting with is easily driven off a common 12 VDC 2A Switching Power Supply AC Adapter (wall wart) I have some plans for building a class-D head in the future with one of these boards when I figure out the power supply. Imagine 600 watt SS slave heads
Note that it's also possible to build stereo combo amps with effects loops for stereo effects in the future, since almost all these boards are stereo.So far I've been driving my Epiphone Valve Jr's 8" Weber Special Ceramic speaker with good results. It doesn't get as loud as the cranked tube Epi Valve Jr, but a lot of people have complained about the relative loudness of the Epi Valve Jr for bedroom practicing. The 8" Weber in the Epi Valve Jr is a speaker that I like a great deal, and I've always appreciated its jazzy qualities. That said, the Parts Express amp sounds a little dead when I play a guitar through it. I was originally using my Alesis mixer as a guitar preamp, which was probably a mistake, because the guitar preamp in a mixer SUCKS. I've since tried my boost and overdrive pedals between the guitar and power amp to see how they stack up as preamps. So far, the MOSFET boost in my BYOC Triboost sounds the best for articulate, jazzy cleans, but it could use a little more warmth. The overdrives sound good but I have to be careful not to pump out too much output volume. These trials was basically just to test the output of the Parts-Express class-D amp board, and I'll have plenty of time to tweak gain and tone. I also have been driving the power amp with just an iPhone playing music, with a minijack-to-RCA cable. Overall output volume is suitable, and I was unable to completely crank the signal into it for fear of disturbing the neighbors. As you'd expect, recorded music played through the Weber 8" sounds a little weird, but the frequency response of a guitar speaker is different than studio headphones or computer speakers. Playing dubstep through it sounds *really* weird.
Preamp:
I'm going with a 12ax7 tube preamp, based on a lot of feedback I got from various forum threads. I'm quick to think that there's a bit of t00b worship in the fact that everyone wants at least a tube preamp, but then again, tubes do a great deal of good for a guitar signal, and it has a particular characteristic that I'm going to try to play off on this amp: a driven signal going into it should behave really well and interact with fuzzes, overdrives, etc.
JFETs are a possible replacement for the tube preamp, or in concert with it (for more than 1 "gain stage" -- using the term loosely here) -- see the OLC pedals for examples of JFETs behaving like tubes in nice little preamp circuits: http://www.olcircuits.com/index.html
To start with, and learn a little bit more about tube preamps, I ordered the relatively-inexpensive "The Persuader Tube Drive" kit from amplifiedparts.com
This is a little pedal kit that runs a 12ax7 in "starved plate mode".. which may end up being the eventual reason it's not going to work for me. See this review: http://guitarkitbuilder.com/content/kit ... rive-pedal and note that the pedal gets very distorted/fuzzy very quickly -- from what I can tell, running it on such low-voltage DC power means that it clips the signal while it's trying to get through the tube. Longer explanation that I clipped out of that review:
NSFW: show
So I'll build the Persuader kit and either sell it or keep it for myself if it sounds cool. I never bought it thinking that I'd plunk down their circuit in as a preamp, but rather to have a way to build something very similar, see how it sounded, and potentially experiment with values of components and circuit layouts in a point-to-point pedal format before committing to a preamp circuit.
Here's a demo video of the Persuader that highlights some of my concern with the pedal: rather than being a clean-ish tube preamp that starts to warm up and clip as its driven by pedals in front of it, it's definitely a distortion that's distorting on its own. Not quite what I'm looking for.
[youtube]http://www.youtube.com/watch?v=arsOqfXXgDA[/youtube]
In a similar vein to The Persuader circuit, there is the matsumin Valve Caster which has been floating around on forums and Beavis Audio, and I might also build one of these to experiment and hear the differences:
One of the neater aspects of this circuit is that you can build the majority of it right on the tube socket pins because it is so simple:
(note that this is someone else's build on the http://www.diystompboxes.com/ forums, not my build!)
I'll cover building The Persuader and testing it with the power amp in a later post (I'll probably be putting it together tonight out in the garage.)
Note that in both cases, I'll probably be running on 12VDC rather than 9VDC for the 12AX7, due to the power supply I'm already using. So I'll have to see how that behaves.
Thinking something more like the ART Tube Mic Preamp I have kicking around in a box somewhere might be more in line with what I'm thinking of for a preamp: it warms up signals nicely, but it doesn't really distort tons on its own. that said, the ART mic preamp has clipping protection built in, which is lame for guitar, and I wouldn't want something like that. Also, it requires an AC power source, which means that the power supply for the combo amp might get a lot more complicated.
Speaker selection:
Working on this front in parallel with figuring out the preamp section. I don't have a ton of money to throw at Neodymium speakers or other crazy things like $200 Celestions. But, there's also an in-built requirement that this amp is somewhat portable, so I've considered weight of speakers into other factors like sound quality and wattage.
An interesting factor that we're going to run into here is that guitar speaker designs factor in the simple fact that we'll be driving them at high volumes. As such, they sound good cranked, and there is some speaker distortion and color introduced when you hear, say, a cranked full stack. We CAN'T really reproduce that with a small practice amp in which the speaker will probably not be cranked anywhere near its power rating. So I've got a working theory that full-range speakers designed for hifi audio might have an edge here -- they're designed to reproduce the sound but not color it, where the guitar speakers may just behave poorly because they're not being driven hard enough. I'll be demoing/comparing the speakers against one another and picking the best, so I'm curious to see how the full-range speakers stack up.
Based on feedback I've read on forums about hte ZT Lunchbox, something like a 6.5" speaker isn't big enough, no matter how much power it handles or how good its frequency response is. It just can't reproduce the fuzzy highs and lows well enough. I've chosen from a range of 8" and 10" speakers to pick the correct speaker for this project. I'm going to judge them all against the Epi Valve Jr's 8" Weber ceramic special, as that speaker has been my favorite small speaker for awhile. (And it makes even crappy chip amps, like the LM386-based Orance Microcrush battery amp, sound good and jazzy at low wattages/volumes.)
I'm working off the theory that a 10" speaker is going to provide the best combo of light weight, good tight low end, and power wattage handling. Later on, I may try more 8" or get a range of 12" speakers to try.
Currently what I have coming in the mail to compete with the Weber 8" are:
Jensen MOD10-35 - 10", Ceramic, 35 watts, 2.5 lbs - I've heard good things about these speakers, and a lot of people are replacing their factory Fender and other combo amp speakers with stuff from the Jensen MOD line.
More specs: http://www.amplifiedparts.com/products/P-A-MOD10-35
Kustom Turbo 10 - 10", unknown, 75 watts, 6.3 pounds - These are essentially a clearance item on Parts-Express, but they're cheap and perked my interest, because they're built by Eminence for Kustom for very high-end cabinets, and are said to have a darker tone that emulates 60's/70's paper cone speakers -- which sounds appealing to the pedal freaks that this practice amp is aimed at.
More info: http://www.parts-express.com/pe/pshowde ... er=299-406
GRS 8FR-8 Full-Range Speaker - 8", unknown, 60 watts, 4.15 lbs - This is the full-range speaker that's competing with the guitar speakers. I'll with hold commentary until I've actually demoed it. But some pluses are that it's cheap, it performs well in reviews, and I got a pair of them to actually build myself a set of hifi speakers after I've tried one as a guitar speaker.
More info: http://www.parts-express.com/pe/pshowde ... er=292-430
If none of these speakers make the cut, I'll be checking out more 8, 10, and 12" speakers from Weber, as well as another Jensen speaker, the 10" Falcon, that looks interesting: http://www.amplifiedparts.com/products/P-A-JC10-40FA
Combo cabinet:
I'll be building the combo cabinet out of birch ply rather than the Epiphone Valve Jr's MDF construction. I hope for better sounds and lighter overall weight with that idea. Most likely the cabinet will be open-back, but I plan to experiment with both by building both an open back and closed back configuration. The decision on open versus closed back also depends a great deal on what speaker I choose to go with -- the full-range speaker above basically requires a sealed and ported speaker cabinet rather than the most traditional somewhat-open-back guitar cab.
Depending on whether I can get to Milwaukee to use it, I have the ability to CNC the pieces of birch play out of a single sheet at the Milwaukee Makerspace on their CNC router. This would mean I could essentially build it to tight tolerances, with box joints, and have a very tight fit between pieces. This would result in a very solid cab and it should be very aesthetically pleasing.
I'm not sure about finishes yet, but let's just say that I've been inspired by the black-stained woodgrain on this Z Wreck that's been one of my computer desktop backgrounds for a couple weeks: http://www.drzamps.com/Images/Amps/HighRes/zwreck.jpg
Amp chassis:
Just a quick note: the prototype will be built in a Hammond 1441-12 chassis, which is a folded and welded box rather than the die cast boxes were used to for our pedals. I'm using this big box to start because I'll probably be mounting different things and drilling a lot of holes, and the space will be nice.
But, when I can source a chassis in the correct size (after knowing the overall size of power amp and preamp boards) my plan is to use a die cast hammond box, and learn to do a resin finish like McSpunckle uses, for the front where the controls will be. Goals for this amp include it looking good as a piece of furniture and art, so I've thought about the faceplate design quite a bit.
Current chassis:
More posts in this thread as I make progress.. Will probably record demos of each piece as well for audio comparison.
References:
- Guitar speaker cab design+construction: http://www.duncanamps.com/technical/speaker_cab.html
Tube amp "class": http://lenardaudio.com/education/13_guitar_amps.html
Another DIY solid state practice amp project: http://www.experimentalistsanonymous.co ... one_ss_amp
The matsumin Valve Caster circuit: http://www.diystompboxes.com/smfforum/i ... ic=63479.0
Other: There's a ton of other stuff I've been thinking about; stuff I've been sketching for cabinet design, knob layout, and visual branding for the little amp so that it looks cool; tons of thoughts that I'm not capturing here on voicing and how to drive a practice amp so it sounds HUGE without pushing bass frequencies (building rattle / mechanical transmission of sound) or too much ear-piercing treble (also tends to wake the neighbors), etc. I'll try to document more of this as the project progresses, but this initial post is already HUGE.
