Now days most are trying to get rid of some power rather than finding an untapped source. A hundred watt Marshall and a 4x12? Where you going to play it at volume?
Last edited:
And how about 15 watts from a pair of little $1 valves that were supposed to put out only a couple of watts? How about 30 watts from a pair of $3 valves?
That's usable power, not 100 watt Marshall absurdity.
-Gnobuddy
Also depends on how it sounds to the person. Some people like tubes biased hotter, the colder ones make more power than those that burn up electrons as heat. I have been meaning to try Class AB2 for years, just was not that important. I have tubes to burn.
Gnobuddy: It is gratifying to see some discussion of my suggestion of a vertical line array of small diameter drivers, which I made in the very first reply to the original post.
The idea of four 6" drivers in a vertical array may not be acceptable due to the narrowness of the cabinet, but how about a four high by 2 wide matrix of 6" drivers? This arrangement would be equivalent to a 2 x 12" cab but with improved horizontal dispersion. In addition, it would be wide enough to place an amp head on!
As you say, the frequency response of the individual 6" drivers can be tailored to suit. The increase in voice coil inductance with frequency also acts to filter out the unwanted higher frequencies.
P.S. To help anyone unfamiliar with the mathematics, I'm going to take this opportunity to explain why four 6" drivers are equivalent to one 12" driver.
Cone area is proportional to the square of the radius of the driver.
A 6" driver has a radius of 3" so its cone area is proportional to 3 squared or 9.
A 12" driver has a radius of 6" so its cone area is proportional to 6 squared or 36.
Since 4 x 9 = 36, we need four 6" drivers to match the cone area of one 12" driver!
The idea of four 6" drivers in a vertical array may not be acceptable due to the narrowness of the cabinet, but how about a four high by 2 wide matrix of 6" drivers? This arrangement would be equivalent to a 2 x 12" cab but with improved horizontal dispersion. In addition, it would be wide enough to place an amp head on!
As you say, the frequency response of the individual 6" drivers can be tailored to suit. The increase in voice coil inductance with frequency also acts to filter out the unwanted higher frequencies.
P.S. To help anyone unfamiliar with the mathematics, I'm going to take this opportunity to explain why four 6" drivers are equivalent to one 12" driver.
Cone area is proportional to the square of the radius of the driver.
A 6" driver has a radius of 3" so its cone area is proportional to 3 squared or 9.
A 12" driver has a radius of 6" so its cone area is proportional to 6 squared or 36.
Since 4 x 9 = 36, we need four 6" drivers to match the cone area of one 12" driver!
Returning to the prospect of improving sound dispersion by incorporating a small midrange driver in a guitar speaker cab, may I offer the following information as an aid to anyone interested in experimenting with the idea?
In a simple first order, series crossover network the input is across a series combination of a capacitor and an inductor. The larger, main speaker must be connected in parallel with the capacitor and the smaller, midrange speaker must be connected in parallel with the inductor. N.B. Never operate the amplifier into the crossover network without the drivers being connected!
To achieve a crossover frequency of 1,000Hz with 8 ohm drivers, we need a 24uF (22uF paralled with 2.2uF) capacitor and a 1.25mH inductor.
The capacitors should be the non-polar type and have a suitably high voltage rating. The inductor should preferably be an air-cored type and should have a suitably high power rating.
The phase shift is 90 degrees so the phasing is unimportant - it makes little difference whether or not the midrange is connected with the same polarity as the main speaker.
If anyone has experimented, or decides to experiment, with this configuration then please share your experiences.
In a simple first order, series crossover network the input is across a series combination of a capacitor and an inductor. The larger, main speaker must be connected in parallel with the capacitor and the smaller, midrange speaker must be connected in parallel with the inductor. N.B. Never operate the amplifier into the crossover network without the drivers being connected!
To achieve a crossover frequency of 1,000Hz with 8 ohm drivers, we need a 24uF (22uF paralled with 2.2uF) capacitor and a 1.25mH inductor.
The capacitors should be the non-polar type and have a suitably high voltage rating. The inductor should preferably be an air-cored type and should have a suitably high power rating.
The phase shift is 90 degrees so the phasing is unimportant - it makes little difference whether or not the midrange is connected with the same polarity as the main speaker.
If anyone has experimented, or decides to experiment, with this configuration then please share your experiences.
Last edited:
The trouble is, when you have two side by side sound sources, reproducing the same signal, the pair have similar treble beaming problems as one single source that spans the entire width. Not just beaming, but comb filtering, where there are dead spots in the sound field, and the positions of the dead spots vary with the frequency of the sound from the speakers.
The exact same phenomenon was used in a brilliant physics experiment in 1801 which proved that light consisted of waves ( 27.3 Young’s Double Slit Experiment – College Physics: OpenStax ). Two side-by-side loudspeakers emitting identical sound waves behave exactly like Young's double slits emitting identical light waves - where Young saw bright and dark spots on his screen, we will have loud and quiet zones of sound.
This problem with multiple small sources emitting the same wave behaving like one big source is bad for us when it comes to speakers, but good for things like radio telescope arrays (you want them to beam very, very tightly, to pick up light from tiny objects far away in the sky), where a handful of radio telescope dishes spaced widely apart acts almost like one single enormous dish, as in the VLA: Very Large Array - Wikipedia
So horizontal arrays aren't good for speakers, even if individual drivers in the array have good dispersion. You can't cram two 6" drivers into a 12" width (there has to be some baffle width between them), so I suspect two 6" drivers side by side would actually be a bit worse in the treble beaming department than a single 12".
The little bass guitar rig I mentioned earlier was two 8" or maybe 10" speakers, vertically arrayed, in an enclosure with proportions similar to a home Hi-Fi speaker (tall and narrow). The amp was one of the new tiny class D jobbies, small enough to fit neatly on top of the speaker.
There are a fistful of "lunchbox guitar amps" out there now, which would probably sit comfortably on top of a tall, narrow, vertical stack of guitar speakers. But ye olde vintage Marshall or Fender head which is nearly two feet wide isn't going to work!
-Gnobuddy
I know you are always creative and experimental in your approach to DIY - I have seen a huge variety of ideas from you on this and other Internet forums.
I don't know if I really need a 30W amp (though all the magazines and books seem to think that is the right size for a gigging amp). But I don't think I can pass up the chance to see 30 W squeezed from a pair of $3 valves for myself - I have to try this, out of sheer curiosity!
I have just about everything I need to make a bipolar +/- 30V - 35V power supply now, so that will be the first step on the road to my first AB2 valve amp.
-Gnobuddy
Hi Gnobuddy!
I'm well aware of the interference effects to which you refer as I actually attended my university physics lectures!
To be pedantic, Young's slits experiment does not prove that light consists of waves but simply that light can behave like a wave. Light has a wave/particle duality - it sometimes behaves as a wave and sometimes as a particle.
A fascinating aspect of this is when single particles (photons) of light are fired, one after the other, at an equivalent of Young's double slits. At first the individual photons appear to arrive at the screen on the far side of the slits in a completely random manner. However, over time, the arrival of many photons builds up a pattern of light and dark areas on the screen.
How can photons passing individually, and at different times, through the slits possibly interfere with each other?
I, for one, haven't a clue!!!!
I'm well aware of the interference effects to which you refer as I actually attended my university physics lectures!
To be pedantic, Young's slits experiment does not prove that light consists of waves but simply that light can behave like a wave. Light has a wave/particle duality - it sometimes behaves as a wave and sometimes as a particle.
A fascinating aspect of this is when single particles (photons) of light are fired, one after the other, at an equivalent of Young's double slits. At first the individual photons appear to arrive at the screen on the far side of the slits in a completely random manner. However, over time, the arrival of many photons builds up a pattern of light and dark areas on the screen.
How can photons passing individually, and at different times, through the slits possibly interfere with each other?
I, for one, haven't a clue!!!!
Sorry! I certainly didn't mean to school you on something you already know well. Thing is, on this forum one never knows the background of the people one is communicating with, so I try to make sure the basic science behind it is clear when I post something. There are so many posts on Internet technical forums that have no basis in actual knowledge, but are just opinions - I try to avoid making those sorts of posts, and to explain why whatever I just said wasn't just an opinion plucked from thin air. ("Red Teflon-insulated wire sounds better! Paper-in-oil capacitors sound better! I have golden ears, and it's true because I say so!")
As this thread has made clear, basic acoustics, interference, and diffraction effects have never made it into the guitar amp and guitar cab world, even though it's been required knowledge in Hi-Fi and P.A. system design for many decades now.
Incidentally, the first vertical line array I ever encountered was an ancient pair of Phillips "column speakers" that my elementary and middle school used as part of their P.A. system. Those speakers were designed and manufactured in the 1960's or early 1970's, I believe. They used an array of elliptical Phillips drivers (maybe 5x7 or 6x9) oriented long axis vertically, and also stacked vertically one above the other, in an open-back box with a front panel that was curved to widen the vertical dispersion pattern.
That design showed the Phillips engineers knew exactly what they were doing: they were getting good horizontal dispersion by keeping the width of each driver narrow, and stacking them in a vertical array. They were getting good vertical dispersion by bending the entire array into a banana-shape, center forward of the ends. They used drivers large enough to reproduce reasonable bass, and small enough to reproduce reasonable treble, for basic speech use (about A.M. radio quality and bandwidth.)
Reading through the previous two paragraphs, I was certainly a very odd duck as a child - I am 100% sure I was the only child in that school that ever noticed those speakers in this much detail, and the only one who actually understood the technical reasons for the design, and probably the only one who still remembers them, decades later!
Oh please, let's not get started on quantum entanglement! 😀
-Gnobuddy
Hi Gnobuddy!
No problem! You obviously have the educator gene! Got one myself, as my compulsion to explain the mathematics of speaker cone areas should testify!
I'll be happy to get 'entangled' with your discourses on physics!
P.S. Those interference effects you mentioned will occur whenever multiple drivers are used in a cabinet, no matter how they are configured. However, I think we are now in agreement that a guitar speaker cabinet utilising a single vertical array of small drivers would maximise horizontal dispersion - even if it would be impractical and unsaleable!
Even adding a small midrange driver via a simple series crossover network, as you have suggested, will lead to interference between the main driver and midrange driver over an appreciable range of frequencies centered around the crossover frequency.
Would this be an acceptable price to pay for an increase in dispersion?
No problem! You obviously have the educator gene! Got one myself, as my compulsion to explain the mathematics of speaker cone areas should testify!
I'll be happy to get 'entangled' with your discourses on physics!
P.S. Those interference effects you mentioned will occur whenever multiple drivers are used in a cabinet, no matter how they are configured. However, I think we are now in agreement that a guitar speaker cabinet utilising a single vertical array of small drivers would maximise horizontal dispersion - even if it would be impractical and unsaleable!
Even adding a small midrange driver via a simple series crossover network, as you have suggested, will lead to interference between the main driver and midrange driver over an appreciable range of frequencies centered around the crossover frequency.
Would this be an acceptable price to pay for an increase in dispersion?
If they are emitting the same frequencies, I agree. All we can do is pick the layout to provide best dispersion in the direction where it matters most (usually, that's horizontal.)
Agreed. I built a 2x8 open-back cab some months ago. It can stand on either its short or long end, so you can choose whether you want both speakers side by side, or one on top of the other.
Since I usually this at home with myself as the only audience, orientation isn't critical, but the sweet spot is certainly wider when the cab is standing on end.
As an additional heresy, neither speaker is a "proper" guitar speaker. Both are of the type loosely called "full range", i.e. each has a small whizzer cone mounted in the centre. Even worse, they were inexpensive, since the cheap plastic sticker on each one doesn't say Celestion, Jensen, or "Eminence. At this point I have been altogether ejected from the Guild of Propah Electric Guitar Players.
(Good thing I wired the speakers with red Teflon-insulated wire. We all know red Teflon-coated wire makes everything sound good, even naughty cheap speakers masquerading as guitar speakers. 🙄 )
Oh, it's perfectly practical, as long as you don't insist on balancing a huge guitar amp head on top of it! 😀
Unsaleable, probably yes, I agree!
My opinion: we can look to Hi-Fi and studio monitor speakers for the answer. The vast majority use a separate tweeter, and plenty of good ones place the tweeter to the side of the woofer, including monitor speakers made for the critical-listening task of mixing music.
So evidently, in practice, interference between the woofer and tweeter signals isn't a noticeable problem in a good speaker design.
Crossover design certainly has something to do with this. After all, interference effects would only occur during the limited range of frequencies where woofer and tweeter crossover networks allow overlap, both drivers working at the same time. So with a reasonably tight (as in high-order) crossover network, interference effects would be minimal.
The passive studio monitors I know about seem to use 3rd order crossover networks. Active ones often use 4th order crossovers.
IMO guitar is an extremely uncritical sound source. This is why we get away with the crudest of enclosures, and speakers made to a recipe from circa 1940 (except for updated Kapton voice coil formers, modern glues, and maybe a magnet material - ferrite, aka ceramic - from the new-fangled 1960s rather than the hallowed 1940s.)
That being the case, I don't think we need to worry about 4th order crossover networks, but I would guess 2nd order would be a good starting point. First-order crossover networks are pretty much useless, and I have never seen one in a well engineered speaker design.
Given that we need to have at least a half-way decent crossover network, and also re-create the erratic and peaky frequency response of a traditional guitar speaker, I really think a good experimental starting point would be to use a solid-state, two-channel, class D power amp board to drive woofer and tweeter, with a simple active crossover in front.
Drive the solid-state power amp with a valve preamp, or a signal derived from the output of a low powered valve guitar amp. Add electronic EQ ahead of the active crossover network to emulate the peaky frequency response and steep treble rolloff of a Celestion or Jensen or whatever your favourite is.
It won't be a 1950's 5E3 Fender or a 1960's Marshall JTM45, but a hybrid amp like this could be a good test-bed for testing how two speakers and a crossover work with guitar.
I have already tinkered with several parts of this recipe, but have not tried the crossover yet. I have to find a suitable (and affordable) candidate for the "big tweeter" first.
My existing unconventional 2x8 cab works well enough for my tastes that I haven't really pursued the crossover idea. Like most of us, I have multiple other projects lined up in my head, several of them more interesting to me.
-Gnobuddy
From the theory I've read, a 4X12 tends to focus the sound into a tight, narrow beam. A 2X12, with the drivers mounted vertically above each other, spreads the sound into a flat wide arc.
To clarify: stacking two 12" drivers vertically doesn't make the horizontal dispersion any wider / better than a single 12". So you still have all the unpleasant beaming effects we've come to expect from guitar speakers.
However, the vertical stacking layout also doesn't make horizontal dispersion any worse than it already is from a single driver. In that respect, it's better than a 4x12, or 2x12 with the speakers mounted side-by-side - those latter two arrangements provide even worse dispersion than a single 12" speaker by itself.
-Gnobuddy
as a sound guy I prefer you get one of those things that aims your amp up and then put the amp in front of you aimed at your face.
That way you can be as loud as you want and I won't have to deal with it by making the rest of the band as loud as you are or running a reverse polarity and eliminating you from the mix.
That way you can be as loud as you want and I won't have to deal with it by making the rest of the band as loud as you are or running a reverse polarity and eliminating you from the mix.
Nice one sound guy!
I can feel your pain!
But we guitarists would prefer you disperse our guitar sound rather than dispense with it!
I can feel your pain!
But we guitarists would prefer you disperse our guitar sound rather than dispense with it!
Back when I played live regularly (mostly bass) my main 4x12 was in a conventional position (perhaps a little closer to wings), but angled in at the drummer and rest of band.
The second speaker was a fold back wedge, which was nestled in amongst the the PA wedges. It really didn't reproduce much under 100Hz but it was very useful in hearing myself clearly once at the front of the stage, particularly if playing "clean". (If not playing clean it provided a pile of different feedback tones 🙂 ).
From the sound guy's perspective (and that of my back) the big advantage was that I could run this off a single 150W/4ohm amp, rather than the 1/2kW and 4x12 + 2x12 + 1x15 stacks I'd been using previously. It probably damaged my ears slightly less too.
Ironically this was all to hear myself over the guitarist and his single 10" combo sitting on a milk crate making a nominal dozen or so watts. Which was cranked.
Last edited:
Isn't it strange that the instrument that was once too quiet to hear in a band (acoustic guitar) morphed into the most obnoxiously loud instrument in the room / club / stadium?
It seems that too-loud acoustic drumsets (evolved by armies from much less-loud folk drums), and too-loud electric guitars (evolved from too-quiet acoustic guitars), have forced tens of thousands of musicians to play too loud, and eventually damage their hearing.
-Gnobuddy