Let's Talk Speakers Q & A
Have a question about speakers, but haven't been able to get an answer?
Maybe we can shed some light on it, maybe not, but we'd be happy to give it a shot. Also, if you see a question to which you have a definite answer, by all means, Answer it and we'll put it here for everyone to see.Go ahead, Ask away. Please note that we know very little about current and popular car audio, home HIFI, surround sound, and subwoofer speaker/driver applications. Please refer questions concerning those speakers and systems to the various Internet newsgroups applicable to them.Thanks!
OK, let's start with basics. What are the main differences in sound between comparable speakers, one with a ceramic magnet, one with AlNiCo? Also, how does the size of the voice coil affect the sound.Jeff Austin
Jeff, good questions. The whole 'AlNiCo mojo' is about smooth compression at high average levels, such as what you would have running the amp flat out. AlNiCo (Aluminum-Nickel-Cobalt) is an alloy magnet and all alloy magnets are easier to demagnetize than comparable Ceramic (Strontium Ferrite) magnets. What this means is that as the voice coil starts moving in response to the input signal, it generates a magnetic field of its own that tries to demagnetize the magnet. As its effect lowers the available magnetic field of the AlNiCo magnet, the speaker becomes less efficient, the voice coil moves less, etc. The physics of it is that the small magnets near the surface of the magnet poles (called 'domains') begin to change state, or flip directions. The result is smooth compression, the same kind of operating curve compression that occurs in a tube amplifier. The ceramic magnet, on the other hand, doesn't compress or demagnetize as easily, so the voice coil moves to its mechanical limit and won't go any farther. This is why some players say ceramics sound a little edgey at high average levels as opposed to AlNiCo. However, by properly designing the entire magnetic circuit, Ceramics can be made to behave quite well for desirable guitar amp tone and dynamics. You might compare the two magnetic circuits to solid state amps versus tube amps, where the solid state amp gives it all its got then clips hard, while a tube amp compresses nice and smooth. The extension of this idea, then, is that with the AlNiCo, like the tube amp, you can seem to have a louder average volume since it gets compressed smoothly. By the way, the compressing or demagnetization that occurs with the AlNiCo is not permanent. It springs right back to its design operating point.
A voice coil is like an electric motor. The bigger the voice coil, the more wire used, the more torque or pulling power you have to move the cone. With the proper match of components, you can get more sensitivity, wider frequency response, and more power handling ability.
What's the difference in sound between paper and kapton/synthetic voice coil formers? Is the choice of material substantial to the sonic character?Larry Barras
Larry, despite the fact that using a paper former is a great marketing tool for vintage style speakers, the fact is it has no effect on the sound itself. It, like Kapton, is a diamagnetic material, and it's effect on the magnetic field and the effect of the magnet field on the former itself is negligible. The big difference is in the mass, or weight of the formers. In the early 70's when the race was on to have the highest powered solid state amp, speaker manufactures had to keep up by making their speakers take a lot of power. Therefore, Kapton was put into use. Because the Kapton is thicker, heavier, etc. (a typical one looks like an orange plastic pill bottle), it takes a lot of power to get it moving. So, the movement was to heavier Kapton voice coils, heavy, damped cones, and damped surrounds for a high powered, relatively low sensitivity speaker. Now we're going the other way. Low powered amps, lightweight speaker components, and high sensitivity to make the lower powered amps sound huge. The one possible exception to the coil former having an effect is the use of an aluminum former. Some believe that since it is a closed loop of metal, there are 'eddy' currents circulating in the former. These eddy currents could affect the sound by causing a braking action to voice coil movement, thereby increasing the damping of the speaker.
Does having an aluminum voice coil cover affect frequency response? I've always been told that it adds high end. Is this true?Rich Santucci
Rich, let's look at the history of the voice coil cover, or 'dustcap'. At first, its only purpose was to keep dust and debris out of the voice coil gap. If you look at any speaker made in the early years, like a Jensen P12R, it had a simple flat felt dustcap about the size of a quarter. Then, as research on the perfect speaker proceeded, it was discovered that if you use a domed dustcap made out of the same material as the cone, you could change or smooth out some of the peaks and dips in the frequency response of the speaker. Next came a combination marketing/engineering idea. A big aluminum dustcap would certainly look cool, and since aluminum has high thermal capacitance, it would suck heat out of the area around the voice coil and radiate it into the air. It's a win-win-win. Cool look, adjust the frequency response, and get heat out of the voice coil. So, to answer your original question, yes, by sizing the dustcap appropriately, within reason, you can affect the frequency response dramatically, including extending the highs.
I recently purchased a '62 Brown Princeton which has a noticeable speaker buzz. I determined the problem is probably a warped voice coil or something in the gap, because I can push in on the cone and hear a rubbing noise as the cone moves. The speaker isDavid Knechtges
David, congratulations on the purchase. Nice amp. The noise is definitely a rub, either from, as you suggested, a warped (from overheating) voice coil, or flakes of paper or other material stuck in the gap between the voice coil and pole or front plate hole. There is a way to correct that if it isn't too severe. I'll detail it here, then you can make the decision whether to try it or not. The result is that you correct the problem without reconing the speaker, thus preserving the value of the original speaker. First, since you will be performing this operation without demagnetizing the magnet, make sure your work area is very clean and you have plenty of light. Lay the speaker on its back with the cone facing up and with a scalpel, carefully cut out the dustcap, leaving about 1/16" of dustcap where it is glued to the cone. This is important because the voice coil wires pass through this point and you want to make sure you don't cut them. Next, use a vacuum cleaner or clean, dry pressurized air to suck or blow the dust and other debris out of the gap. If you hold the speaker upside down with the cone facing downward it will probably help getting the dust and debris out. Next, take a 3x5 index card and cut it into a strip that is the correct length so that you can form it into a circle and stick it down into the gap between the inside of the voice coil and the outside of the pole. This will help form the voice coil back into a circle. Next, lay the speaker back down on its back. Take a Q-tip or small paint brush and dip it into a bottle of acetone (finger nail polish remover). Spread a small amount of this acetone on a couple of the rings of the spider, which is the brownish/yellow corrugated disk attached to the backside of the cone at the base of the basket. Next, place a jar lid or other disk on the cone where the dustcap was and let the speaker set overnight. The lid or disk will prevent dust from getting into the gap overnight, and the acetone causes the spider to relax and reposition slightly, thus repositioning the voice coil. The next day, remove the lid and the index card strip and see if you still have a rub. If you do, try the acetone again, same procedure. If, after a couple of tries, it seems hopeless, then professional reconing is the only resolve. I think it's worth trying though, to preserve the value of the original speaker. If it works, contact me with the size of the dustcap and I'll send you one to replace the one you cut out and instructions on how to replace it. As far as using the speaker, if you plan to use it regularly, at high volumes, I would suggest packing the original away and install a replacement speaker. Many speakers would work well in that amp, such as a Mojo MP10R, a Naylor 10, a Kendrick 10, or a WeberVST P10Q. If you want some british tone, you might check out Celestion's new Silver series or WeberVST's Blue Pup and Silver Ten.
Footnote 02-17-97 David followed the procedure outlined above and corrected the rubbing problem, thus preserving the value of his vintage speaker.
I think I understand your explanation of the differences between AlNiCo and Ceramic magnets in speakers, but what do people mean when they say modern AlNiCo is different than old AlNiCo or that it has a half-life?Shawn Bolton
Shawn, I hadn't heard about the belief that old and new AlNiCo are different, however they are one and the same. For speaker applications, AlNiCo 5 is the best choice in the AlNiCo family of alloy magnets. Its peak energy product is just right for loudspeakers where we need to concentrate high densities of magnetic flux in the gap around the voice coil. AlNiCo 5 is an alloy made up of 8% Aluminum, 14% Nickel, 24% Cobalt, and 3% Copper. The cobalt is what makes the AlNiCo expensive. Most of the world's supply comes from the African country of Zaire. Besides that country controlling the market, cobalt is also a strategic metal used in missiles and other weapons systems. It currently sells for about US $32 a pound.
As far as it having a half-life, that's another new one on me. When a speaker is put together, the magnet is initially uncharged, or unmagnetized. Then, at the end of the assembly line, just before testing, the speaker is passed under a huge electromagnet that zaps the magnet with about 10 to 20 times the magnetism that would be required to saturate the magnet. After the electromagnet is turned off, the speaker magnet immediately loses about 2% of its magnetism and then stabilizes. In the next year, it drops another 1%, and then is essentially stable for thousands of years. Unlike a flashlight battery, the magnet is not being used up or depleted of its energy while in use. All that has happened is that we have forced many of the tiny molecular magnets called domains to realign themselves in one direction. Once 'flipped', they reach an equilibrium and stay there. Besides intentional demagnetization with a demagnetizer, there are three situations that can cause a magnet to become partially or fully demagnetized, and this may be what the half-life comment is all about. The first one is excessive heat. That's not a consideration for speakers because the temperature to demagnetize an AlNiCo magnet (called the curie point) is over 1000 degrees F. The second is from a large, changing magnetic force. This could happen in a speaker. The typical case would be where a person blows out the speaker by using excessive power. The high value of magnetism produced by the voice coil could partially demagnetize the magnet. This is why you should always ensure that anyone who is going to recone a speaker for you have a magnetizer to renew the charge on the magnet, just in case it got partially demagnetized. The third and final situation is physical shock. If you dropped an AlNiCo magnet speaker and it happened to land on the corner of the magnet frame, it could partially demagnetize the magnet.
I need information on how to wire speakers to obtain 2, 4, 8, & 16 Ohm loads. A schematic for each configuration would be prefered. This has been a puzzle to me for a long time.Graham Kahnt
Graham, let's look at the definition of impedance first, then go from there. You'll often see a speaker or other device rated at a 'nominal' Z, or impedance. The word 'nominal' comes from the Latin word 'Nomen' which means simply 'name'. An example of where you may have heard this term used in another context is during a space shuttle mission. During the initial ascent, you'll often hear the astronauts say "all systems nominal", or "mission nominal". What they mean is that everything is going as planned, as written, or as described. In the case of a speaker, we are calling, or naming the device a certain impedance. The electrical quantity of impedance is made up of resistance, which doesn't change with frequency, and reactance which does change with frequency. So, impedance is the combination of the two at a particular frequency. Remember in the movie Wizard of Oz when the Scarecrow finally got a brain, he immediately started reciting some bizarre formula about "The sum of the squares of the sides of a right triangle...."? He was reciting the Pythagorean theorem for right triangles. We can use that formula to calculate impedance. Think of a flagpole with the sun casting a shadow on it. The height of the flagpole would represent the reactance, and a line between the base of the flagpole and the point on the ground away from the flagpole where the shadow stops would represent the resistance. If you connected a string between the top of the flagpole and the point on the ground where the shadow stops, the length of the string would be the impedance quantity and would be longer than either of the other two. So, what am I getting at? A speaker that is said to be 'nominally' 8 Ohms will have a resistance lower than 8 Ohms. So, a rule of thumb is that if it measures lower than a commonly used name, such as 8 Ohms, but not lower than the next lower commonly used name, such as 4 Ohm, then you would call it the higher name, or 8 Ohm. Many 'nominal' names have been used over the years, including 2 Ohm, 10 Ohm, and 15 Ohm. 4, 8, and 16 seem to have been standardized though, for the past 30 years or so. The main reason for the variance between several coils, all of which would be called 8 ohms, for instance, would be because each might have a slightly different DC resistance due to the length of the coil, size of the wire used, etc. In each case, though, the DC resistance would be in the range of 5.5 to 6.5 DC Ohms, so it would be called an 8 Ohm device. Another method for naming the impedance of a device is by actually measuring the impedance, or AC resistance of the device with special test equipment. Many have used 400 hz as the test frequency, while others have used 1,000hz as the test frequency. Some have derived the name from an impedance plot such as the one shown in Figure 1 below. They declare the nominal impedance to be the impedance at the 'first dip after the first peak'. Notice the large peak at the speaker resonance around 100 hz. Then, it drops dramatically, dips, then starts going back up. It would be the impedance at the lowest point of the dip that would be the declared 'nominal' impedance. It's an interesting exercise in using an impedance bridge, but the old rule of thumb we discussed earlier works just fine for declaring a nominal impedance.
Do the holes in the speaker frame affect the sound any?Glenn Ewban
Glenn, that's a good question. Technically, in an open backed or ported enclosure the size and shape of the cutouts would make a difference in the sound because there is actual movement of the air mass behind the speaker as the cone moves. The air movement consists of a pressure waveform that changes with the frequency and intensity or loudness of the sound from the speaker. In a sealed enclosure, such as a 2-12 arrangement in a piggyback amp, the cutouts don't affect the sound as much because the waveform reaches a constant pressure or equilibrium and remains fairly constant. Of course, even closed back, or sealed cabinets have to leak a little. If they were hermetically sealed, barometric pressure changes would bias the speaker cone in or out and disturb the operation of the speaker. Over the past 30 years or so, most speaker designers haven't given much thought to the issue of designing basket cutouts because the necessary metal stamping and tooling for speaker baskets is very expensive. Therefore, a few different types have been essentially standardized for the industry. You'll notice that most speakers made in America are either a 4 or 6 trapezoidal hole arrangement, and baskets from most manufacturers look about the same. British speakers tend to use the larger 4 hole trapezoidal arrangement.
I have a 20/50 watt THD 'plexi' amp which I run either 2-EL84s or 2-EL34s. In order to get rich distortion and compression, it must be played loud. I would like to get that sound at lower levels, so I intend to purchase a THD HotPlate or similar attenuatoSteve
Steve, that's true about the F-M curve and the correction needed at the lower volumes. I would have to recommend though, before using a power attenuator that you go through the calculations or have a professional amplifier technician advise you on its use with regards to steady state power handling and thermal dissipation capabilities of your output circuit. There's some 'sonic psychology' involved here, in that as you are listening comfortably to the lower, attenuated level of acoustic power, the output circuit is still giving it all it's got. So, since your ear hears the lower level, it isn't quite so evident that the output circuit is working so hard. The bottom line is that the output transformer is heating up, it might be in saturation a lot of the time, you lose some bass due to the saturation as well as increased resistance of the windings due to heating, etc. As far as speakers go, depending on how much attenuation you want to dial in, you'll probably want to get the most efficient speakers you can find. The bottom line is that you are attenuating the big amp down to the power level of a small amp, so you need a speaker that will make the small amp sound huge. If you want to restore a lot of the highs, you might go with a Kendrick 10, which is very efficient at the high end. As far as AlNiCo vs Ceramic goes, since you aren't going to really blast the speakers with a high average volume, that is to say a lot of your tone (distortion plus compression) will be coming from your output circuit, I think Ceramics would work just fine, and save you some money.
Additional comments submitted by Joe Pampel: Another option you might consider is the use of a closed back or sealed cabinet. With an open back, you lose a little bass due to cancellation between the rear and front. In a closed back, that problem diminishes, however it takes more power to get the same loudness. Also, rather than using high efficient speakers as suggested above, you might actually prefer lower efficient speakers. Assuming you'll have plenty of power to put into either the power attenuator or the speaker, you can put less in the attenuator and more in the speaker for a given loudness. You might find that this arrangement will give you more overall control of your tone at the loudness you desire.
Tell me about speaker fatigue. I've heard that it has to do perhaps with heat buildup in the voice coil after heavy use (i.e. after a 4 hour gig at volume). Also, does an older speaker that's been used alot at volume necessarily see speaker fatigue and neBill Abernathy
Bill, speakers have always been quite resilient and forgiving considering the beating they take in the way of cone flexing and vibrations. The weakest mechanical links in the chain are the corrugations in the surround at the periphery of the cone. At this point, the cone is the thinnest, yet takes a lot of vibration. So, like bending a wire until it breaks, the corrugations will often begin to tear after long term use. Since the cone is made from paper pulp, other considerations such as temperature and humidity play a part in weakening the cone. Many speakers are a little tight when new, and seem a little sterile. After being driven at high levels for awhile, they change character slightly as the surround corrugations begin to loosen up. The spider also loosens from the constant flexing, so the net result is a little more low end and the cone breakup characteristics change. Most people acknowledge that this change is for the better as far as enhancing the desired tone for guitar work.
Heat buildup in the voice coil is definitely a problem, because it will eventually cause the shellac (or other chemical used to hold the wires in place) to soften and allow the voice coil former to deform. We can't afford to have the former deform much, because the gap between the outside of the former and the plate hole (called 'ring') is somewhere in the neighborhood of .010 to .012 inch. Many attempts have been made over the years to get the heat out of the voice coils, including the use of large aluminum dustcaps or regular dustcaps with vent holes. Several different former materials have been used also, including aluminum and Kapton.
One of the big problems I've seen with vintage speakers is with the spider shifting in position. This shifting will allow the voice coil to shift, causing a voice coil rub. Many vintage baskets were painted, so the glue holding the spider down wasn't applied directly to the metal. Over time and with alot of vibration, the glue would actually pull up the paint.
Vintage speakers used various forms of glue, from animal based to other chemical concoctions. Those glues were no match for the CA (super glue) type glues available today. Today, we use a brown, rubbery glue we call 'elephant snot' to attach the spider and the underside of the cone to the basket. For attaching the voice coil to the cone neck, we use CA-77, a type of thick super glue. For the top of the cone to the gasket, we use a white glue similar to Elmer's glue. This is for appearance only, since it dries clear. We still use the flexible black rubbery glue to hold down the voice coil wires where they pass through the cone and connect to the tinsel foil wires that run to the terminals. The dustcap is held down with either the black glue or the super glue, depending on dustcap style.
Your question about using vintage speakers and the resulting fatigue is a tough one for me to answer. Because of the vintage glue and pulp cone ageing problems, the chances of failure when used regularly at high average volumes is considerable. If I had a vintage amp with original speakers, I'd be inclined to remove the originals and, assuming they were in good shape, pack them away in safe storage. Then I would install replacement speakers, such as Kendrick, Mojo, Naylor, Celestion, or WeberVST. Because of the high intrinsic value of original speakers, I'd be much more comfortable beating a replacement speaker with high volume.
I recently purchased a '60s Ampeg SB-12 amp. As I understand it, the amp was designed to be used for either guitar or bass. But, when I play bass through it it sounds distorted. I had a tech check it out, replaced tubes, etc., but still had the distortionJames Miades
Jim, while using a speaker with a low bass resonance and given the fact that lower frequencies tend to require more power, it makes sense that you would get more volume with that type of speaker, especially if you could get one with a high SPL (more sensitivity). However, one of the most popular amps of all times -- the Fender '59 Bassman -- started out as a bass amp and then became popular for guitar, and it has 4-10" low power speakers. If you were going to use the stock speaker in the Ampeg for guitar only, I'd suggest trying to repair the small tear with the popular concoction of tissue paper and Elmers or with silicon gel. Try your best to close the gap in the tear before glueing in an effort to maintain the same position and 'parallelogram' established by the cone, spider, and voice coil. If you want to use the speaker for bass, though, I'd recommend reconing it. The excursions (overall cone movement in and out) at bass frequencies will probably strain the patch, causing the cone to shift slightly at that point during peak excursions, and the result could be voice coil rub during those peaks. This would put high frequency scratchy sounds on your bass note and would be very annoying. If you do have it reconed, you can ask the reconer to select a cone and spider that will yield a lower resonance and better low end response at the expense of losing a little of the highs.
When I was restoring an old phonogragh I noticed the magnets on the speakers were those with the thin magnet inside a small metal box with two sides missing. What are these speakers called? Are they worth reconing?David
David, I believe you are describing AlNiCo plug type magnet speakers. The small metal box with two sides missing is called the field case. The sides or legs of the field case direct the magnetic field from the top of that round magnet around to the bottom or base plate. In the middle of the base plate is a hole. Attached to the bottom of the magnet is a piece of round steel, and that round piece fits inside the hole in the base plate. There is a gap or space between that pole and the hole in the base plate. The voice coil that is attached to the cone slides in and out of that gap or space and fits around the round metal piece attached to bottom of the magnet. The speaker is assembled very carefully so the voice coil doesn't touch the pole or the sides of the hole. All old AlNiCo speakers are worth reconing in my opinion, because they are AlNiCo and have that vintage sound. The only drawback is the expense of getting them reconed. Many record players of that era used similar speakers, so you might check out your local thrift shop, garage sales, etc. for a similar unit. You'll probably be able to buy a complete unit with good speakers for less money than you'd spend reconing the ones you have.
I noticed in your original note you said you are 12 years old. Congratulations on your restoration project and your success with it. You've chosen a great and rewarding hobby, in my opinion.
I have two 8 Ohm speakers that I can wire either parallel or series to my tube amp for an impedance of either 4 or 16 Ohm. My amp has both 4 and 16 Ohm output taps. Are there any sonic differences or benefits of series over parallel wiring or vice versa?Gerald C. Lopez
Gerald, connecting two speakers in parallel is an old trick to smooth out speaker response and enhance the damping of either speaker. HIFI designers took it one step further by connecting two speakers of different sizes in parallel. A speaker has a large impedance increase at its fundamental resonance, and depending on the installation, this can cause the speaker to sound boomy or out of control. By connecting two speakers in parallel, particularly two speakers of different sizes with different resonant frequencies, each speaker will tend to quench or dampen the boominess of the other. Since no two speakers are exactly alike, even two of the same size, that damping will occur, however slight, for any speakers connected in parallel. For speakers connected in series, there appears to be less control, and more of what is called 'back EMF' from the speakers fed back into the output circuit. While that seems rather chaotic, many players prefer the series connection, as it gives them a more textured tone, enhanced breakup, and overall a more desirable tone for guitar work. It's totally subjective, of course, and many factors affect the end result, such as voice coil size, gap energy, closed back/open back, output circuit damping, etc. The best thing to do, in my opinion, is try both arrangements since you have the luxury of impedance tap selection, and go with the configuration you like the best.
I am interested in AlNiCo magnets. There are many kinds of AlNiCo magnets, my question is what kind are used in those old Altec, Western Electric, JBL, etc. speakers?David Abe
David, an entire family of alloy magnets employing Aluminum, Nickel, and Cobalt were developed. The most common ones are 2, 5, and 8. Each has its own unique characteristics which generally has to do with how strong they are when magnetized (similar to how many volts a battery has at its terminals), and how easily the magnet can be demagnetized (how much current the battery can put out before its voltage begins to sag). AlNiCo 5 was almost universally chosen for loudspeaker use because it has a high flux (like high battery voltage), and under normal circumstances of use, the speaker wasn't intended to be driven hard enough to affect the magnetism of the magnet by the voice coil magnetism. Of course, all guitar players discovered early on that driving a speaker hard into this region of what is called 'recoil' yields a smooth compression of the magnet field and as it turns out is very desirable for guitar tone. Taking the magnet to the 'edge' of recoil is not a problem for AlNiCo 5, because it springs back to its operating point as the signal to the voice coil is reduced. However, excessive power to the voice coil, enough that would probably cause damage to the coil in the form of excessive heat would most likely generate enough magnetism to partially demagnetize the AlNiCo magnet. What's interesting about magnets and magnetic circuits is that they have a load line just like tubes and transistors, and these load lines are employed when designing the loudspeaker, motor, or other magnetic device.
I was wondering if you could tell me anything about magnetic shielding on speakers (i.e. center channel speakers) because I'm building my own and I have no clue how to shield them?Alex Rush
Alex, the reason we have to shield the magnet on the speaker is because not all of the magnetic flux is focused in the gap as useful energy for use by the voice coil. The wasted, or unusable flux is called leakage flux. Ceramic magnet speakers and modern speakers with large gaps tend to have more leakage flux than vintage AlNiCo magnet speakers with tight gaps. An example of a typical ceramic magnetic circuit and the location of the leakage flux is shown in Fig. 1 below. Shielding the near environment of a speaker from the leakage flux is often important, especially near a television since the picture tube in a TV uses magnetism to direct the electron beam to draw the picture on the screen. The constant leakage flux from the speaker would affect the scanning beam and distort the picture as well as the colors, etc. The best shielding is a canister or pot-shaped cover made of a high permeability metal called mu metal. It is basically a short circuit to the leakage flux so the flux can't get through the mu metal and radiate into the surrounding air. What happens is that the leakage flux tries to magnetize the mu metal, the mu metal generates a magnetic field of its own that resists the leakage flux. All of the magnetism stays within the mu metal so it isn't radiated. The foregoing was a simple explanation of how all magnetic shielding works. With the popularity of surround sound, center channel, etc. for TV's today, many speaker companies are starting to use 'pot' covers for their larger, higher powered speakers that will be used in close proximity to a TV. If you want to try to make your own, you can get a coffee can or other metallic canister that will just slip over the magnet as diagrammed in Fig 1. Apply some silicon gel to the perimeter of the magnet to hold the can in place. The only precaution you need to take is that you want to make sure your canister doesn't come in contact with both the back plate and the front plate that sandwich the magnet, which would cause a shunting of the magnetic flux away from the gap where it is needed. Another method of containing the leakage flux that is gaining in popularity these days is the use of a bucking magnet. The speaker manufacturer takes a magnet the same size of the magnet used in the speaker, magnetizes it, and glues it to the back plate. It looks strange as heck to see this ceramic ring magnet hanging off the back of the magnet assembly, but it is quite effective in cancelling the leakage flux from the main magnet and in some cases, can actually add to the flux in the gap. Strangely enough, when you add up the cost of a metallic pot cover of sufficient size, the cost to paint it, and the cost of the glue or fastener required to install it, in many cases the extra ceramic magnet used as a bucking magnet is actually cheaper. TV's are not the only place mu shielding has been used. I was recently told by an acquaintance that when he installed a new speaker in his amp (which had a rather large magnet), his channel switching quit working. Apparently, the speaker was located close enough to the channel switching relay that the leakage flux affected its operation.
I don't know a lot about guitar speaker cabinet design. I take it that open back cabinets take advantage of the drivers free air resonance, but how do manufacturers come up with the dimensions of sealed enclosures? It seems to me that the specs on the indT. Naumann
I sure wish I could state an answer as well as you posted the question. Unfortunately, I have never met, talked to, nor read any articles written by the original designers of speaker cabinets for guitar amps with regards to their philosophy of design, etc. Like most, I do have my own observations and opinions on them, though. No doubt, in a combo amp an open back enclosure is imperative to allow air to flow in and out for cooling the tubes and other electronic components. Having the back open also allows for a convenient storage space for the power cord, etc. The amp would sound louder overall at the expense of losing a little on the low end, the extra loudness being a plus for the marketing of the amp. Most combos I've seen are designed with just enough room for the speaker to be mounted with a symmetrical border of baffle around it, while not interfering with any of the chassis components. Having been involved in production engineering as well as cost accounting, I can assure you there was also some thought given to how much wood and covering would be used for a given size cabinet using the standard sizes of material stock, while yielding the least amount of scrap. Another issue would be total shipping weight, since that would have to be factored in when arriving at a price for dealers. In a sealed enclosure, it generally takes more electrical power to get the same acoustic power as a similar sized open back. The sealed air in the cabinet behind the speaker acts as a resistance to cone movement, the smaller the enclosure, the more the resistance. Think of blowing up a small, tight balloon. You have to blow like heck to get it started compared to a large loose balloon. So, the bigger the box, the better, generally. Now think of the early '60s 'piggy back' amps. It wasn't long before players were standing them on end with the amp head sitting on top. Was it for better sound projection or because it looked cool? Of course, manufactures saw that, started making the 'stand-up' models, and made the cabinets larger but still used the same size speakers. Next is the question of designing the box to match the speakers technical specs. I've often wondered that myself, and have never really arrived at a good answer. Some recent amp manufactures, such as Dr. Z are using ported cabinets and are getting very high marks from reviewers. Nevertheless, these designers, like all guitar amp designers, are designing and building for a particular sound and performance as a guitar amp, not a HIFI speaker enclosure. In other words, the issue here is music 'sourcing', not music 'reproduction'. Your last question is one I don't think I'll ever be able to answer, even as a speaker manufacturer. I've had people call me and tell me "I gotta tell you, this is the worst speaker I've ever heard in this amp, way too dark". Yet, I've had others tell me the same speaker is the best they've ever had, and they've been trying for a long time to find the right speaker for that particular amp. To design a cabinet that matches the speakers technical specs will certainly smooth out its performance, make it a better mathematical model, so to speak, but on the other hand, much of the desirable guitar tone is in the comb filtering, cone breakup, and other distortions that occur when a speaker is taken to the edge of its design operating parameters. It's apparent this is why serious players buy, sell, and trade speakers, like amps, until they get the tone they are after.
Additional comments submitted by Joe Breher: Jim Marshall, the man who gave us the world's most emulated sealed box guitar amp speaker enclosure, was recently interviewed by a popular guitar magazine. In the interview, he stated that the dimensions for his cabinet were dictated by the smallest practical cabinet that would fit 4-12" speakers. Evidently, no thought whatsoever was given to the acoustical properties of the cabinet.
I'm building a 1x12 guitar cab and have made the baffle from 3/4" plywood. I have not yet installed it and just read online about a guy using 3/8" ply for his baffle. I have a 70s Fender Princeton and it's baffle is 3/4" that's where I got the figure fromBob Thomason
Bob, let's look at the reasons for the different woods, construction techniques, and how they affect tone. In a HIFI enclosure, we want the cabinet to be sonically dead, so some cabinets are felt lined, or other forms of deadening material are used. In a guitar amp, to a certain degree, we want the opposite characteristics. Because, as we discussed in the previous note, we're sourcing music rather than reproducing it, a live cabinet adds to the texture of the guitar tone by generating standing waves and minor resonances that produce a comb filtering effect, hence the added texture of the overall tone. That's why there has always been an ongoing debate on the use of plywood versus pine board, finger joints versus lap or butt joints, etc. Of course, we don't want the cabinet to be so lively that it buzzes on big bass notes, etc. I wouldn't recommend the use of a 3/8" board for the baffle, though, especially with vintage speakers, because of the tight gaps and lightweight speaker components. Any flexing in the baffle could twist or strain the basket and cause a voice coil rub. Another consideration when using a thicker baffle board is that since the speaker is located farther back from the front surface of the baffle board, it tends to decouple the speaker from air mass resonances that occur at the opening. The 3/8" difference between a 3/8" and a 3/4" board may not seem like much, but it can and does make a difference. Those who remember the tone ring used in some 60's amps are aware of this effect. Vibrations on the surface of the baffle board can intensify the effect of the air mass resonances. The thicker baffle board increases the mechanical resistance to those vibrations, thereby reducing their intensity. I've built solid pine, particle board and plywood cabinets and have been happy with the results, however, I have always used a 3/4" baffle board. Perhaps some other readers can give us their opinions on this, which I would be happy to post here.
I would like to get into reconing speakers and was wondering if you could tell me how to get started. Where to get parts, can you order parts from the speaker companies,etc. If you knew the answers to these questions, it would be great!Mike
Mike, there is a company in Ft. Wayne, Indiana called WVS which sells an instructional video tape on reconing. They can also provide parts and customer support. Their web page is www.reconer.com You can also contact Image/Waldom in Chicago to see about purchasing speaker reconing products from them. They require a membership and have protected territories for existing members, so your success in being able to buy from them will depend on how many of their members are already established in your area. Technically, speaker reconing is very easy, and with practice, you can recone a speaker in about 15 to 20 minutes. It seems, though, with any business, there is always a 'gotcha!' In the case of speaker reconing, it is imperative (in my opinion) that you have a demagnetizer/magnetizer. It's extremely important to get any debris out of the voice coil gap prior to rebuilding the speaker. If any of the debris happens to be ferrous, (material that is attracted to a magnet) the only way you will be able to remove it is with the magnet demagnetized. After the rebuild, you recharge, or remagnetize the magnet to it's full state of magnetism (saturation). This is also very important, especially in the case of an AlNiCo magnet. If the speaker was blown by the use of excessive power, the magnet could have been partially demagnetized. So, you need to ensure the magnet is 'renewed' in any case. At the very least, your customer deserves to have his or her rebuilt speaker as clean as when new, and a full charge on the magnet.
The 'gotcha!' I referred to is the cost of the magger/demagger. A commercial unit to do all of the more popular speakers today costs about $14,000. You can buy them used or as surplus for about half that, and can get some even cheaper that need repaired. Talk about lethal energy, though, WOW! Let me give you an idea of how much energy is in a magger/demagger. If you're familiar with a typical Fender 40 Watt amp's power supply, you might see a high voltage circuit of around 350-450V with 40 to 80uF (microfarads) of filtering. Our big magger for production goes up to 800 Volts and has over 80,000uF of filtering (storage). Needless to say, that's alot of juice, and it all gets dumped into a coil around the speaker magnet in less than 1/50 of a second. Rare earth magnets require even more energy than that. The coil and wires shake violently, and we have to hold down the speaker with quite a bit of pressure so it won't get tossed around. The whole thing sounds like a muffled explosion when the coil gets blasted with that much energy. Best of luck to you with your reconing business. If we can be of any help, please don't hesitate to ask. I don't think there will ever be a shortage of speakers that need to be reconed.
One of the two speakers in my amp has apparently developed a buzz or rattle. I don't think it is a voice coil rub because I can't hear the noise at low volume and it only does it when I hit certain notes. Any idea what the problem might be?Richard Canton
Richard, it sounds like the magnet assembly has worked loose and is resonating against the basket plateau. Let's look at how the magnet assembly is attached, then we'll try to find a resolve for the problem. For the past thirty years or so, most speaker manufacturers have used a staking process to attach the magnet assembly to the speaker basket plateau. When the front plate of the magnet assembly is fabricated, a punch press is used to punch out the round front plate. Then, a second operation punches the hole in the middle where the voice coil fits. A third operation punches either four or six holes in a three inch pattern around the voice coil hole. In this operation, the four or six punches don't go all the way through the plate, only about 3/4 the way. This results in four or six plugs or stakes to appear on the other side of the plate. At the speaker assembly shop, the basket is placed down over these stakes and a press flattens or mushrooms them out to fasten the plate to the basket. Some speaker manufacturers use a gasket between the plate and the basket, while others use a bead of glue. These are used in an effort to stop any rattle that may occur. The problem is that due to production tolerances, punch wear, etc., some of the stakes may not get pressed or mushroomed all the way out. After awhile, with vibration, these can work loose enough to cause the buzz or rattle.
Traditionally, speaker manufacturers tested each speaker by sweeping a frequency from low to high and then checked the response with a standard as a pass/fail indication. This was time consuming, and with the advent of fast computers capable of doing fast mathematical analysis, most speaker manufacturers today use a composite test tone package, display the result on a computer screen, and compare the plot to an upper and lower plot also plotted on the screen. The composite test only takes about 1/2 second and resembles the sound that would occur if you could push all the numbers on your touch tone telephone at the same time. While this method saves alot of time, it doesn't always catch the problems such as the one you are probably experiencing with the magnet/basket rattle. So, what do we do about it? If you have an audio oscillator, connect it to one of the amp channel inputs. Next, turn up the amp to a comfortable level and slowly sweep the audio oscillator from a low to a high frequency. If the buzz doesn't occur, try turning up the volume and do the sweep again. Speaker manufactures tend to get carried away with their power ratings, so you don't want to leave a steady tone at high power level driving the speaker for very long, just long enough to detect the buzz or rattle. OK, let's assume you found the frequency at which there is a pronounced buzz. Place your fingers on the front side of the magnet and your thumb against the basket plateau. Slowly move your hand around the perimeter of the magnet while maintaining a constant pressure with your thumb against the basket plateau. When you find a spot where the rattle of buzz stops, mark it or remember it, and turn off the audio oscillator and amp. Next, get a safety razor blade, and with two pair of pliers, break the razor blade in half lengthwise. With one pair of pliers, or needlenose pliers, force the sharp side of the razor blade into the gap between the front plate and the basket plateau. If it's too loose, you may have to double up on the blades, or use a thicker piece of metal. I like to use a razor blade since the edge is sharp and will work into the gap much more easily than a blunt edge. Do the sweep test again and see if this cured the problem. If it did, use some silicon gel or other thick glue to hold the razor blade in place.
If, by the way, you don't have access to an audio oscillator, you can use a microphone and whistle a tone from low to high to perform the sweep test. A few manufacturers still bolt their magnets to the baskets. The reissue Bulldog, for instance, uses brass bolts to attach the plates and large ring magnet to the basket. WeberVST uses bolts and glue on all their products out of necessity since their plates and poles are machined rather than stamped.
How do I determine the best match, power wise, between amp and speakers? For example, I have a 100 watt SF Twin and want to replace the two 8 Ohm speakers. How do I calculate the wattage ratings needed for each speaker to best match the amp? Also, how doChuck Minahan
Chuck, normally you would use two speakers of the same power rating. For instance, in your situation, you would use speakers which would be rated at 50 watts minimum. Together, whether they are wired series or parallel, they would handle the 100 watts since the power would be divided evenly between them. Speakers with higher power ratings generally have larger magnets since the wire in the voice coil is a larger diameter, is heavier, and requires a larger magnet to get the magnetic energy required to get the heavier voice coil moving to achieve high frequency response. Low frequency response is also improved.
As far as determining the specifications of an unknown speaker, about the only way to do that is to take it to a professional reconer who can perform sweep and T-S parameter tests on it. Mixing different types of speakers in the same installation will work as long as you know the power ratings of each speaker and don't exceed those ratings in either speaker. Many players I have talked to achieve the tone they are looking for by mixing speakers with different characteristics. For instance, one might have clean, tight low end, with smooth top, while the second one would be very efficient in the upper mids and high end. I know of one player who has a 2x12 amp with a P12B Blue Dog on one side and a P12N on the other. According to him, it gives him just the right blend and texture for his particular playing style.
I have seen special devices that mimic the frequency response of vintage Celestion speakers, etc. for the primary purpose of running a guitar preamp directly to a mixing board for recording or P.A. It seems that you could accomplish the same goal by eitheAlbert Tejera
Albert, the speaker plots we are all familiar with and often see published are single tone sweep plots, usually performed with 1 watt of power applied to the speaker. This has become a standard in the industry and allows us to compare one speaker to another on the same terms. Other tests performed include 2nd and 3rd harmonic distortion plots, and single tone sweep plots at higher levels, e.g. 25 and 50 watts. While these tests allow us to compare apples to apples, so to speak, they don't really address the issue of how the speaker reacts to a composite signal made up of many frequencies. When you apply a composite signal to a speaker, the low frequencies may move the cone a certain amount, while the higher frequencies riding on the low frequencies move the cone in small steps back and forth. The cone system itself, because it has mass (or weight), will continue to move by itself and actually works against the signal commanding it to move by generating a voltage in the voice coil that counteracts the signal from the amp telling the voice coil to move. This is called motional impedance. As you can imagine, this is a very complex effect and is not very friendly to mathematical analysis. This is why the selection of speakers, by necessity, is very subjective. What appeals to one player may not appeal to another. This motional impedance also plays a big part in how a speaker responds to a particular amplifier output circuit or enclosure. Many speaker emulators on the market today use inductors, capacitors, and resistors in an effort to model a speakers voice coil, but they stop short of emulating the motional impedance. Others use opamp filters as well as other shaping networks in an effort to put peaks or dips in the response curve. The only way you could truly emulate a speaker would be to have some kind of electromechanical device in the signal path that would generate the 'back EMF' and in effect, the unpredictable motional impedance. Several methods could be used to do this. You could build a dummy load/voice coil emulator and use a small portion of the signal to drive two 4 or 6" speakers mounted face to face. The low powered signal would drive one of the speakers, while the other one acts as a microphone to pick up the acoustic signal of the driven speaker. This 'microphone' signal would then be amplified and mixed in with signals from other parts of the circuit, and you'd then have a fairly complex signal at the output. Another method would be to remove about half of the laminations from an iron core choke and use that in the design of the dummy load/voice coil emulator circuit. The laminations would vibrate and emulate motional impedance. Of course, the laminations would need to be fairly small and lightweight so they would vibrate sufficiently to affect the inductance of the choke.
The foregoing description of emulating motional impedance assumes you have a power amp to drive the emulator circuit. Since your question was about using a preamp output for DI, the only choice you really have is to use filters or EQ to emulate the single tone sweep response, short of using a complex algorithm employing a Digital Signal Processor IC. As far as plots on the Internet, I have seen plots for HIFI drivers, but as you have discovered, I haven't found any for musical instrument speakers. I characterized and studied every vintage loudspeaker I could get my hands on while doing the research for our speakers, and plan to add a page to our website to display those plots and their T-S Parameter info. I will also post as much information as I can about guitar amp speakers available today. We also plan to have a page that will provide information for building guitar amp speaker enclosures using the T-S parameter info for these speakers. We will have info for sealed (closed back) as well as ported cabinets.
I have a ceramic Jensen gold and brown label speaker (220523) that is stamped a "C12PS". I have seen 12P's, N's, Q's, etc., but what does the "PS" mean?Ron
Ron, according to the old Jensen data that I have collected, the AlNiCo 'P' model (P12P) is similar to the 'N' model in that it uses a 1-1/2" voice coil and has essentially the same cone dimensions. In your case, it is a later model employing a ceramic magnet. For this reason the energy in the gap of the 'P' model is about 2/3 that of the 'N' model. It also has a power rating a couple watts less than the 'N' model. The 'S' suffix is something I'd have to guess on, though. For special or custom design applications, many manufacturers used one of their base models and made changes in the magnet, spider, etc. to get the frequency response or other characteristic the customer desired. In this case, the customer probably specified a looser spider in an effort to get the same sensitivity with a smaller (less expensive) magnet. That could be where the 'S' came from. It could also designate a particular customer, or simply the next letter in a sequence for modifications of a particular base model. If I run across some definitive information on this particular model suffix, I'll post it here. Incidentally, Jensen made a model C12N in later years, but if I'm not mistaken, it is quite a bit different than the P12N in most respects.
Why is it that some speakers sound great in a closed back or sealed cabinet, but sound terrible in an open back cabinet, and why is the opposite true with other speakers?Chris Eades
Chris, excellent question. When a person designs a speaker, one of the considerations is how much the cone will move and the subsequent loudness for a given amount of power driving it. If the designer uses a loose spider and fairly flexible surround on the cone, the voice coil will have an easier time moving the cone and the speaker will be loud with a small amount of power. The problem with a loose spider and surround is that as the power is increased, it takes the cone to its mechanical limit of movement and it gets harsh. This is an underdamped system and can also sound ringy. If you put this speaker in a sealed or closed back cabinet, the air that is trapped in the cabinet acts as a spring or big balloon that the back of the cone pushes against. This air mass helps control the movement of the cone and also affects the damping of the system, making the speaker sound more in control. A designer who uses a big magnet, tight surround and tight spider is designing for a low system Q, good damping, and good transient response. In other words, when you hit it with a sharp attack note, it responds and then stops very quickly rather than ringing on and on. This speaker would work well in an open back. However, in a sealed or closed back, the air mass spring we discussed earlier would add to the damping, potentially causing the speaker to be overdamped, compressed, and lifeless. It's a difficult design task to design a speaker that is a good compromise for all shapes, sizes and types of cabinets.
I saw a ceramic magnet speaker the other day that had two ceramic magnets stacked together rather than the usual single magnet. Does that mean the magnet is twice as strong or is there some other reason for that? Thanks.Bill Whicker
Bill, that sure looks impressive, doesn't it? In order to get more bass you either have to go to a larger cone or you have to move the cone farther in and out. This is because the radiation impedance of air is very low at low frequencies. If you stick your hand in a tub of water and move it forward and backward very slowly, you'll notice the water doesn't resist your hand movement, nor do you create much of a wave. So, you either need a bigger hand, or you need to move your hand farther with each stroke. The same thing happens in a speaker. With a small speaker such as a 6" or 8" you typically see connected to long tubes in car audio for subwoofer applications, the cone has to move a lot to reproduce the very low frequencies. This means the cone has to move as much as one inch in and out. A typical guitar amp speaker only moves about 1/8" in comparison. So, in order for the voice coil in the small speaker to move that much, there has to be a lot of room between the voice coil and the back plate. Unfortunately, stacking ceramic magnets, as impressive as it looks, only yields about 20-25% more magnetism than a single magnet. This is due in part to the extended pole required and the loss of flux associated with it. The bottom line, then, is that the main purpose for stacking magnets is to facilitate extended cone movement.
I want to build a cabinet for an amp head I bought and was wondering about the various speaker configurations, etc. For instance, if I use two ten inch speakers, is that the equivalent of one twenty inch speaker? I guess what I'm asking is what is the relJohn Zwern
John, historically, musical instrument amps have used a single full range speaker rather than multiple, specific range speakers with crossover networks such as that used in HIFI installations. In order to reproduce the high frequencies, the musical instrument speaker must have a fairly low mass (weight) cone system so it can vibrate fast enough and with sufficient movement to reproduce the high frequencies with adequate volume. Connecting two of these speakers in parallel will increase the output at the low frequencies thereby emulating a larger speaker, while retaining the low mass properties of each speaker. Since the cone is round or circular in shape, the math used to describe the effective area of the speaker involves pi (3.14) and the square of the radius of the cone. At low frequencies, where the entire cone is moving in and out as one piece, the effective area of the two speakers in parallel is approximately equal to the square root of the number of speakers times the diameter of one of the speakers. So, two ten inch speakers in parallel would be approximately equivalent to 10 x 1.414 (the square root of two) or a single 14 inch speaker for the low frequencies. So, in theory, you would get the low frequency response of what you would expect from a big 14 inch speaker, while retaining the high frequency response of the lower mass ten inch speakers. An important point we need to make here is that the low frequency response is determined by both the size of the cone and the movement of the cone. So, it requires more power to reproduce the signal as the frequency goes down. This is particularly true below the resonant frequency of the speaker, and the output falls quite rapidly below that resonant frequency. If the 14" speaker we calculated had a lower resonant frequency than the 10" speakers, which in all probability would, it will have a better bass response than the two 10" inch speakers, despite what we calculated. As far as comparing relative or calculated equivalent sizes, I emphasized that this equivalency would occur at low frequencies where the cone is moving in and out as one piece, because at higher frequencies, say above 1Khz, the cone vibrates in sections. At low frequencies, the sound from the speaker covers a very broad, even pattern in front of the speaker cabinet. At higher frequencies, this pattern is reduced to smaller, narrow patterns call lobes. The collision of these smaller lobes from one speaker with the other one add texture and a certain amount of comb filtering to the overall sound coming from the cabinet. This is why some players hear what they describe as a kind of reverb, phase shifting, almost stereo sound coming from a multiple speaker cabinet as opposed to a single speaker cabinet.
I have picked up some Trapezoidal enclosures (made out of Birch plywood). for sound reinforcement. I am interested in reloading them with JBL speakers. The bottom end boxes had two 18"speakers (HxWxD - 48"x24"x24") The top boxes had two 15" speakers and aChris
Chris, that's quite a setup you have there! Let's go through some preliminary information, then I'll point you to a website where you can get all the math and other information you'll need. Trapezoidal or any shape of cabinet other than square and rectangle really does serve a purpose other than looking cool. In theory, an unusual shaped cabinet like that has fewer parallel walls, so fewer standing waves are set up that can affect the sound in an adverse way. As far as port and cabinet dimensions go, you really have to know the technical parameters of the loudspeaker(s) you are going to use so that you can tune the cabinet and port(s) properly. While it is possible to experiment endlessly until you are happy with the performance, it would be much better to calculate it as close as you can, then make minor tweaks, just like an electronics experimenter does. In guitar work, as opposed to HIFI work, we often have the unusual arrangement of connecting several drivers or speakers in parallel rather than using specific range speakers with crossovers, etc. While this may seem to complicate matters, it really doesn't. For the port, you can simply calculate for one speaker, then add a port of the same dimensions for each of the other speakers, or you can use only one port, increase its diameter by the square root of the number of speakers connected in parallel and adjust the length accordingly. Probably the easiest and fastest method would be to just calculate for one and add one for each speaker. 4 speakers, 4 ports. Here then, is a website you can go to that will lead you to more information on cabinet building than you could ever imagine, including pictures, the math, design philosophy, suggestions, parts sources, etc.
Best of luck to you with your cabinet project.
I'm interested in the differences between edgewound versus regular voice coils and how this factor effects the sound of a speaker. I'm particularly interested in this subject with regard to bass guitar speakers. I've heard that edgewound coils generally hLarry Weissenborn
Larry, speaker design involves a lot of physics theory because we are trying to accelerate a mass (weight) in a short amount of time. This is necessary to move the speaker cone in and out to reproduce high frequencies. In the early days of speaker design, many attempts were made to build full range speakers, rather than specific range speakers that are combined with other speakers and networks to achieve the full frequency range. Again, in order to get the full range, we have to make all of the components in the speaker system very lightweight. Figures 1 and 4 below are examples of strands of copper wire as they would look when wound on a voice coil former. Notice the wasted spaces between the windings, due to the curvature of the round copper wire. Figure 2 shows how we can eliminate most of the wasted space by using square wire. If we take it one step further and flatten the wire into a ribbon shape, we reduce the wasted space and also add more current (and power) capability, as shown in Figures 3 and 5. This also allows us to have more windings for a given coil length. Most voice coils use copper wire, however, aluminum wire was used in some ribbon or edgewound coils to save even more weight. Many of the original edgewound coils were short coils, usually just the length of the width of the front plate. These are called 'even hung' voice coils since the ends of the coil are even with the edge of the front plate. Although even hung coils exhibit more sensitivity, they run out of coil in the gap during big excursions of the cone. The result is square-waving of the mechanical signal (force) to the cone. In other words, it just stops moving. If the cone is very lightweight and has high top note efficiency, this square-waving can cause ringing and ghost notes in the cone and it can sound very harsh.
I have heard various views on impedance mismatches between the amp and the speakers. One is that you should always match the impedance (4 ohm amp = 4 ohm speaker or two 8 ohm speakers in parallel), or you can blow your transformer. The other is that it isChuck
Chuck, technically, you should always provide a load that is recommended by the manufacturer of the amp. The designer of the amp chose a particular output device (tube) and specified all of the operating voltages for the output stage so the tube would work at its optimum efficiency while delivering maximum power to the load with minimum distortion. Ok, so let's discuss the problems associated with mismatches. When you use a load that is lower than the intended load, the output has to drive the load (speaker) with more current because it is a lower impedance than is expected. Two inherent problems associated with transformers are flux leakage and regulation. Flux leakage is also referred to as leakage inductance. It is related to the current in the secondary, and these problems increase as the current increases. As the current draw in the secondary increases, the primary has a more difficult time transferring the signal to the secondary, so the secondary signal to the load gets squashed, or 'soft-clipped'. This soft clipping is called regulation. While regulation is desireable in a power supply, it is undesireable in a transformer. In other words, in a power supply, if the input voltage or the output load current changes, we don't want the output voltage to change. In a transformer, we want the output voltage to follow the input voltage and not regulate at all. When you put a heavier load on the output than was intended, it will pull the output voltage down, hence regulation. The leakage inductance problem arises because the current from the heavier load causing the regulation to occur reduces the efficiency of the transformer by not allowing the output to follow the input. Transformer designers simulate or view this problem as having extra inductance in series with the primary. The extension of this idea then, is that with the heavier load, you could affect the efficiency of the transformer, alter the frequency response (due to the extra leakage inductance in series with the primary), and cause other distortions to occur. OK, on to mismatching the other way. A speaker is a current operated device in that it responds to the current through it to generate a magnetic field that works against the magnetic field of the speaker magnet to make the cone move in and out. Thinking in very short amounts of time, when the output charges up the voice coil with current, then the signal goes away or gets reduced, the cone system moves the voice coil back to its home or resting position. As it is moving back, it generates a voltage that is fed back up the line into the transformer and appears in the output circuit of the amp. This generated voltage is often referred to as flyback voltage, because we are charging up an inductor, then when we disconnect or stop charging the inductor, the magnetic field in the inductor collapses and induces this big voltage into itself. This big voltage then 'flies back' to the source of the charging current. There is a mathematical formula to determine how big the voltage is and it is related to the inductance of the voice coil, the amount of time it was fed current, and how much current it was charged with. The bottom line is that the voltage fed back to the output circuit is oftentimes much higher than the voltage that was used to drive or charge up the voice coil initially. This voltage gets transformed up by the turns ratio of the output transformer, and in many cases can be over 1,000 volts. What happens then is that arcing can occur between the pins on the output tube socket. Once this has occured, a carbon path forms on the tube socket between the pins. The carbon path allows a steady current to flow between the pins and eventually burns up the socket due to the heat that is generated. For example, it wouldn't be too uncommon to see a transformer turns ratio of 30:1. If we had a voltage fed back from the voice coil that was around 50 volts, 30 times 50 would be a 1,500 volt spike at the plate of the output tube. This is why you often see designers connect diodes in a string between the output tube plates and ground. They are trying to suppress these spikes and dissipate the energy in the diodes rather than allowing an arc to occur at the tube socket. So, when you use a higher impedance load on a lower impedance tap, the turns ratio is higher and resulting fed-back (flyback) voltage gets multiplied up higher than what it would have been with the correct impedance load.
It's just about impossible for me to answer how long an amp would last under these conditions. It all depends on how the designer took these potential problems into account in his or her design with regards to the quality of the tube sockets, the use of stringed diodes, the output circuit operating voltages, etc.
I hear alot of talk about the magnet cover making a big difference in the overall tone of a speaker. Is that true, and if so, how much of an effect does it have? Should I try to make magnet covers for my speakers?Thomas Sherill
Tom, I'm walking on thin ice answering this one because it involves some beliefs that are closely held by many. A speaker is a dipole device, meaning the same amount of air is moved when the speaker cone moves back as when it is moved forward. So, since a stream of air is generated towards the back of the speaker as the cone moves back, it stands to reason that we would want the path to be as smooth as possible with the least amount of interference to the airflow as possible. Any interference would cause reflections of air back to the speaker which would affect the overall tone and character of the speaker tone. OK, now that we have the technical description of what's happening out of the way, is it relevant? No, not really. The undulations or variations in the surface of the magnet circuit would be the concern, but the aggregate of these undulations is insignificant compared to the slots and struts in the basket. Most magnet circuits are smaller in diameter than the plateau where the magnet circuit is mounted, further evidence that the effect is negligible. This kind of belief reminds me of the beliefs concerning paper versus other voice coil formers, and special kinds of wire compared to high grade standard copper wire. Technically, the mathematics is there to substantiate a claim, but the measured differences are out 5 and 6 decimal places, rendering the issue a non issue to the average ear. Magnet covers, or pot covers are and always have been used for cosmetic reasons. The one exception is as I pointed out in an earlier Q&A question. Pot covers have recently been resurrected for use in surround sound systems to provide magnetic shielding.
I've always wondered how the ports in a speaker cabinet work. Also, how do they design them or decide how many to use?Jim Chaverel
Jim, the purpose of the port is to replace the bass that is lost from the main speaker as its efficiency drops off at lower frequencies. As mentioned in an earlier Q&A answer, in order to reproduce strong bass, we either need a larger speaker, more movement of the cone, or both. Since the bass output of the speaker is determined by both the movement of the cone and the size of the speaker, you can see why the bass drops off quite rapidly as the frequency goes down. OK, on to the design of the port. What we do is size the speaker cabinet and the port so that the port will resonate at the frequency below the resonant frequency of the speaker where the speaker output power or loudness drops to half. So, we are taking the energy (in airflow) from the back of the speaker cone, since it moves just as much as the front side, and blowing it into the port. This is no different than blowing across the top of a coke bottle. The idea then, is that this 'free energy' is put to use by replacing the sound that is decreasing from the speaker. Of course, nothing's free or perfect. Since the port is only tuned to one specific frequency, it has a resonant frequency where it works best and provides the most help to the main speaker. A problem associated with ports is due to the fact that we are blowing air through the port. The result is wind noise due to the tumbling of the air through the port. Designers refer to this as port viscosity. What it means is that the walls of the port are a different resistance to the air movement as the air in the middle of the port. So, tumbling of the air occurs. One way to cure this is to use multiple ports to split up the airflow and divide it among the ports. Many designers of multi-speaker cabinets use one port for each speaker. Some use as many as four ports for a large base woofer. Although porting has shown up in guitar amp speaker cabinets over the years, it has been used mainly in HIFI systems. Lately, some of the more popular specialty guitar amp manufacturers are offering ported cabinets and are receiving good reports from customers.
Alot of information about ports and port design can be found at: Do It Yourself Audio Page
I have several old Fender Champ amps from the 50's and 60's. I thought I had the speaker impedance thing all figured out until I saw that you offer 8" speakers in both 3.2 and 4 ohm. I always thought they were all 4 ohm, and the 3.2 was either a misprintMitch Cornell
Mitch, pretty confusing, eh? I'm not sure where the best starting point is, so I'll just lay out the facts and hope it makes sense. First and foremost, speakers for guitar amp applications have always been designed by subjective analysis rather than objective. In other words, tone. A typical guitar amp speaker has a short voice coil so it will be very sensitive, and a thin, paper cone so it will be very lightweight and have a wide frequency response. The efficiency (loudness) is both for marketing of the amp and the fact that we need high sensitivity for it to be a wide-range speaker since it is all by itself, i.e. no help from bass woofers and high frequency tweeters like the setups used in HIFI speaker systems. So, we typically use what's referred to as an even hung voice coil. This means that the length of the voice coil is the same as the width of the front plate of the magnet circuit as shown in the diagram below:
Now, let's talk about that front plate. Four thicknesses of steel have essentially been standardized for a long time for the front plate. They are 1/4" (.250), 5/16" (.312), 3/8" (.375), and 1/2" (.500). Many of the early guitar amps speakers, both AlNiCo and ceramic, used the 5/16" front plate. Also, in America, the impedances of 4, 8, and 16 were pretty much agreed upon as being the standard impedances for speakers. Now, if we wind an evenhung voice coil for, say, 8 ohm impedance, it would measure about 6.4 ohms DC resistance. If we change the wire size so we can make it a 4 ohm coil without changing the length of the coil, it works out about half, hence 3.2 Ohm DC resistance. Going the other way, a 16 ohm coil would be about 13 ohms DC resistance. OK, now let's say we want to use 1/4" steel for an 8" speakers front plate to make it lighter and cheaper, yet we still want to retain the evenhung coil. I call it the 80% rule. The 1/4" steel is about 80% the thickness of the 5/16" steel. So, reducing the coil length to 1/4" rather than 5/16" to match the new width of the front plate, we can simply take the original 4 ohm impedance and multiply it by 80% and we get 3.2 ohm impedance. We can find the DC resistance change by taking the 3.2 ohms DC resistance of the original 4 ohm coil times 80% and get around 2.6 ohms DC resistance. So, you see, it all follows the thickness changes of the front plate, and that's why I call it the 80% rule. Unfortunately, the confusion that is widespread is the coincidence that the 4 ohm coil has a DC resistance of 3.2 ohms, and when we reduce it by the 80% rule, we also get the number 3.2, only this time we are talking about impedance rather than DC resistance. Many suppliers of modern replacement speakers make a point in their advertising of a 4 ohm speaker by saying (3.2 ohm DC resistance). The difference between using a 4 ohm and a real 3.2 ohm speaker on a 3.2 ohm output is small, to say the least. Weber uses .312 (5/16") front plates on all their products, including the 3.2 ohm models. In this case, the coil is actually an underhung coil since it is shorter than the width of the front plate. The result is a more low end, more linearity, and a little wider cone excursion range.
Why aren't more speaker frames made out of Aluminum?Tim Nickel
Tim, there are several reasons why stamped aluminum frames, or baskets, are not used.
- Aluminum is softer than steel, so a stamped aluminum basket could bend or twist causing a voice coil rub.
- Steel is much less expensive than aluminum.
- The weight of a typical steel basket, or frame, is insignificant compared to the weight of the magnet circuit. Therefore, there wouldn't really be a weight advantage using aluminum for a stamped basket.
Of course, there always seems to be an exception to the rule. Cast aluminum frames have been used over the years in some of the higher priced speakers where lightweight but very rigid baskets were necessary because large and heavy magnet circuits were used with very tight voice coil gaps. The advantage here is that the aluminum casting can be quite complex and beefy for the added strength where a similar steel casting would be too heavy. Note these are the high end speakers such as EV, JBL, and others. The implementation ratio of stamped steel baskets to cast aluminum baskets is huge.
This may seem like a "non question" to one looking for good sound but, I am really curious to know if a standard has ever been established regarding the polarity of the magnet in a loudspeaker. In other words, does the "north pole" of the magnet always faRalph Nichols
Ralph, that's a good question. I suspect at some point in the early years of speaker development there was a standard, but it has never stuck. Virtually all custom speaker manufactures conform to the customers wishes rather than their own, so for whatever reason, the customers have flipped and flopped over the years both in the polarity and the actual marking of the plus terminal. This could have something to do with the customers assembly line installation of the speaker harness, or just about any reason, really. Speakers that are/were intended to be used in a single speaker installation, such as in an elevator, etc., for background music may not be marked at all. In fact, I saw some identical speakers going down two separate lines at a speaker company once, and noticing the line equipment was slightly different, I checked out the magnetizers and final test stations. Sure enough, the two identical speakers were magnetized oppositely. I asked my host about it, and learned about the doesn't matter application. Of course, I already understood that in a single speaker application such as a guitar amp using a single 10 or 12" speaker, the polarity doesn't matter as far as hooking it up, but I always thought manufacturers would at least magnetize everything in one direction. I also asked, as you have, and looked in several books that I have found on the subject and have never been able to find a specific agreed-upon standard for polarity, either at the magnet N/S position, or at the terminals with the polarity indication. Out of curiousity, I have run a bunch of tests with different speakers and amps to determine if I could hear a difference, knowing that the real physics of it says there shouldn't be. I couldn't hear a difference in any of the setups, and I even had a person who I couldn't see flipping the switch so I couldn't know which way it was while I was listening. I must admit, though, I was biased by the fact that the physics of it says there should not be a difference. Some people say they can definitely hear a difference, though, particularly when they have their favorite guitar, effects, and amp in the equation.