About 20% of 1801a listeners provided specific feedback about the sound pressure level of the OW1 tweeter in the 1801a.  None of them conveyed that the 1801 needed more sound pressure from the tweeter.  About 10% of the 1801a owners conveyed that the 1801a needed less sound pressure from the tweeter.  Herein I will describe some of the pertinent dynamics, my comments, and a summary of the resistor options that affect the tweeter level.  I wish the art/science of building speakers was simple, but there are some very worthy considerations when placing a speaker in a system.

There are three resistor options for the 1801b tweeter circuit.

1.  7, 15 (7 ohms in series and 15 ohms in parallel) - flat tweeter

2.  8, 15 (8 ohms in series and 15 ohms in parallel) - about 1db down tweeter

3.  8, 12.5(8 ohms in series and 12.5 ohms in parallel) - about 2db down tweeter

Series and Parallel Resistor Explanation

Reading this section is not necessarily mandatory.  I convey this information because a few folks have wondered how this series and parallel "stuff" works.  I scratched a picture that might help those who are visual learners.

Please think of electricity like water. Flowing electricity is like flowing water. Adding resistance to electricity is very similar to placing your thumb over the end of a garden hose. Your thumb will resist the flow of water. Consider electrical PRESSURE in the same way as water PRESSURE. Consider electrical FLOW in the same way as water FLOW.

This paragraph will address the effect of the series resistor. The first number (i.e. 7 or 8 ) is the value of the series resistor. As this value becomes larger the spl of the tweeter decreases. This is because a larger resistor will absorb a greater amount of energy from the signal. A HUGE resistor would absorb all the energy. No energy/heat will remain for the tweeter and no sound will come out of the tweeter. NO resistor will soak up NO energy/heat. All of the sound will come out of the tweeter.

This paragraph will address the effect of the parallel resistor. This resistor is effectively placed across the terminals of the tweeter/driver. This is the second number (i.e. 12.5 or 15). As this number becomes larger the amount of electricity traveling through the tweeter will increase. As this number becomes smaller the amount of electricity traveling through the tweeter will decrease. This is because the voice coil in the tweeter will has roughly 6 ohms. If a straight wire, no resistance, is placed across the terminals of the tweeter then ALL of the electricity will go through that straight wire. Electricity applies pressure evenly to both the tweeter and the straight wire. There is no resistance in the straight wire, therefore all of the electricity will go through that straight wire.

There is a more complex reason that both of these resistors exist in the circuit. Please don't read this if you want to "keep it simple".  I don't think there is a simple mechanical way to convey this. The purpose of having both the series and parallel resistors in the tweeter circuit is to obtain the proper level of impedance in that circuit. This way the impedance of the tweeter and the woofer will be the same at the crossover point. An even impedance transition makes the phase matching easier. This is because phase shift across a capacitor or inductor is determined by the inverse tangent of the AC versus DC resistance. The AC resistance is frequency dependent. The phase shift direction depends on whether a capacitor or inductor are providing the AC resistance. Capacitors shift phase forward and inductors delay the phase. The DC resistance is provide by the voice coils, resistors, and inductors in the circuit. Capacitors do provide some consistent DC resistance, but it is effectively negligible. Therefore, getting the phase relationship correct at the crossover point requires the even resistance at the crossover frequency.

You are welcome to question me on the information in this section, but my response will not be quick. I'll likely have to do some research. Also, I generally don't do any of the above calculations manually. I did this once and it was waaaay too much work. I allow my measurement jig to determine proper impedance and phase. It is significantly  easier. I thank God for my measurement jig - when it works. Sometimes, well, my jig gets a little frustrating.

Hopefully these remarks are helpful.

Room Dynamics

The amount of absorption material in the room will determine the amount of sound that is soaked-up.  Listening rooms vary from very dead to very live.  Nearly all absorptive material absorbs more high frequencies than low frequencies.  Hence, the in-room response of the tweeter is slightly lower in a room with full carpet and soft furniture.   Such a room would be termed "dead".  Based on feedback, my experience, and good science from Everest's The Master Handbook of Acoustics, a room with nice thick carpet, heavy cloth furniture seems to favor a slightly higher tweeter level.  This translates to less resistance behind the tweeter and a more flat response.  In some very extreme circumstances a slightly elevated tweeter might be desirable, but this has never been the case with the 1801.  I believe this is due to the very good off axis energy from the smaller dome of the OW1.  A room with wood floors and wood chairs will sound thin with a flat tweeter.  This is because the none of the high frequency  is absorbed and those high frequencies bounce until arriving at the listeners ear.  This translates to the desire for a lower tweeter level and more resistance behind the tweeter.  There is much more information involved, but the jury agrees that a room with more absorbency also sounds more dead, and will accommodate a slightly higher tweeter.  For more information, I highly suggest Everest's The Master Handbook of Acoustics.

Recording Preference

Initially I was fully convinced that a perfectly flat woofer/tweeter balance was key to good sound with any recording.  This is surely a purist perspective.  Dennis Murphy maintains that a flat tweeter simply conveys what is on the recording.  I certainly agree with this, but don't prefer a flat tweeter.  Based on my experience, and feedback from others (my wife included)  I am safe to convey that a flat tweeter response is preferred only when listening to very good recordings.  Specifically, very good symphony recordings from Delos, Reference Recordings, and Telarc sound good when listening to the 1801 with a flat tweeter.  Dennis Murphy likes it perfectly flat.  I prefer 1db down with very good recordings.  This preference changes dramatically when listening to other recordings of the more common variety with a slightly exaggerated tweeter on the recording.

I have an older Celine Dion CD that is a fairly decent recording but the high frequencies are very hot on this recording.  Even with the resistor values at 8, 12.5 my wife insisted that the level of the tweeter was significantly too high.  I agree.  This is because the Celine Dion recording is very hot in the tweeter.  It sounded very out of balance and quite annoying.  This trend was present, but less obtuse with several other very decent recordings from John Lee Hooker, Andrea Crouch,  and Point of Grace.

Source Equipment

I will first describe my good system(s) and the impact on resistor level.  I will then describe my perception of cheap electronics and the impact on tweeter level.  Lastly, I will roughly describe the system of one 1801 listener that is very pleased with a flat tweeter.

I have essentially two systems for listening and testing.  Both of these systems are of acceptable quality or better.   I have one system upstairs, and one system downstairs.  My downstairs system is also my testing system.  The downstairs system consists of a 125wpc Kenwood KVR-126r receiver (@1988), and a SB-Live video card.  I generally use zip cord for speaker wire and the interconnects are certainly not fancy.  My upstairs system consists of a Bryston 3B-ST (or Jolida 302b), Creek OBH-12, modified Anthem CD-1, home brew Belden wires and Goertz bi-wire for the speakers.  Regardless of what system I use for testing, the balance of the tweeter remains unchanged.  This includes the Jolida 302b with El34 tubes in a push-pull configuration.  I hear the same levels in all viable combinations of components.  This might seem surprising, but this is what I hear.  I will also convey that I have no desire to change the tweeter level while using any of the source electronics described above.  With any acceptable electronics my preference for the tweeter is unchanged.  I prefer the mellow sound of 8, 12.5 - even with tubes.

However, even this resistance is not enough for an older XXXX  (@1983?) stereo integrated amplifier that I own.  This integrated amplifier was rated at 100wpc and retailed for about $650 back in 1983.  This XXXX amplifier has dynamics galore.  It also has some very significant grit and noise.  This noise isn't visible in the specifications, but is very audible to the ear.  I have heard this grit and noise on other lesser source equipment previously.  Some of the equipment was new, and some if it was slightly old.  With this combination, I prefer slightly less tweeter level overall because the noise is also subdued.  I cannot explain why my old XXXX amplifier conveys this grit and noise, but it is very present.  I don't purvey a set of resistors appropriate for such an amplifier.  This is primarily because such an amplifier should never be used with the 1801s, and secondarily because 8, 12.5 is already comfortably mellow.

Steven Harrison has a dedicated listening room and the very best equipment attached to any 1801 to my knowledge.  His source equipment is all BAT. He listens to classical music and attends the symphony on a regular basis.  Steven enjoys a the highest viable tweeter level on his 1801a.  This tweeter level is very similar to 7,15 in the 1801b.

Harmonics and Detail

Detail results from linear motors, stiff cones, coherent phase, damped diaphragms, strong cabinets and many other sources.  However, this section will address the perception of detail via fundamental and harmonic sound pressure.  This is pertinent because the relative tweeter level will change the level of detail perceived in the music.

The sound produced by any instrument can be described by its production of fundamental and harmonic frequencies.   The fundamental frequency is the musical note written in the page.  The fundamental note is the one played by the musician or sung by a vocalist.   Let's consider a middle-"c".  Many different  instruments can play a middle-"c", but all of these instruments sound different.  A vocalist can sing a middle-"c" using an a, e, i, o, u... but these sounds are all different.  The reason they are different is because the HARMONICS are different.  A single note from an instrument or voice covers a huge frequency spectrum.  There are many ways to test this.  The easiest is to unplug the tweeter from your speaker.  When listening to music without a tweeter the sound from even the male vocals will sound dull.  Female vocals sound  horrid, and a violin is non-recognizable.   This is because the tweeter effectively carries the harmonics of the music.  The perception of detail results from clear reproduction of music harmonics. The FUNDAMENTAL tone provides a sense of impact, but the  harmonics provide a sense of detail. A downward frequency response slope will have nice authority, but lack in perceived resolution. An upward frequency response will sound detailed, but will lack impact.  This is true with subwoofer bass, vocals, violin and all instruments.  The correct balance of fundamental and harmonic frequencies results in the correct sound.

Someone once told me that I could not determine what a speaker would sound like from looking at a frequency response curve.  While other issues also effect the sound of a speaker, nothing could be further from the truth.  Changes in the response curve do effect the sound signature of a speaker.  Unplugging a tweeter can prove this before the worst pair of ears. The frequency response curve is certainly a factor in determining the sound of a speaker.

Graphs

This section will grow as I add feedback from other folks and their thoughts on tweeter balance.

The graph above is a 3 sweep multi-graph 1/3rd octave smoothed frequency response plot of the resistor options The upper sweep has the lowest resistor values, and the bottom curve has the highest resistor values.  I purvey these resistors with the 1801 in either kit or completed speaker form.  I smoothed these graphs because the jagged ripple made the graphs very difficult to read.  These graphs are taken about 15 degrees off axis to greater simulate the in-room response.  At 45 degrees of axis the 3khz - 5khz dip actually becomes a 1db bump.

The frequency response curves on the top of the graph appear darn close.  None of these resistor choices is terribly wrong.  However, the resistor changes are audible to my ears when listening to music. 

The top MLS sweep represents the 7, 15 resistors.  My wife likes this setting with good symphony recordings.  Dennis Murphy likes it too.  I agree that it sounds nice with good symphony recordings.  The sound isn't terribly forward, but there is slightly too much tweeter for my ears.  Nonetheless, a well recorded violin sounds very nice using this tweeter level.  Conversely, my Celine Dion CD is utterly piercing in the treble region.  Every treble sound is in the foreground IMO, but the detail is nice.  It sounds very revealing, but too forward for my ears.  I have no desire to entertain this resistor choice in my living room.  Nonetheless, it is a viable option for classic music lovers with good equipment.  Relative to the other choices, this one sounds clinical.

The middle MLS sweep represents the 8, 15 resistors.  This is my choice for a dead room, good recordings, and good equipment.  I would choose this configuration if I wanted to sell 1801s commercially.  The OW1 does sound darn good, and a very slightly hot tweeter sells well in hifi shops even before critical audiophile ears.  This resistor level is a fairly nice balance and will be tolerable with lesser recordings.  I'd chose this resistor level if I only listened to good symphony recordings.  I characterize this resistor choice as a "comfortable level of detail".   Relative to the other choices, this one sounds accurate.

The bottom MLS sweep represents what I use in my living room.  It might be too mellow for some folks.  The sound is a bit dead, but the balance with most recordings is correct IMO.  The background treble remains in the background and the foreground treble remains in the foreground.  Classical recordings from Telarc, Reference Recordings, and Chesky sound a little dull because the violin harmonics aren't pronounced, but not all violins are pronounced at the symphony.  The Chesky Jazz recordings still put the drum hat in the front.  I characterize this resistor choice as "smooth and lovely".   Relative to the other choices, this one sounds smooth..

Arriving at these decisions regarding the tweeter balance of the 1801 has taken several years.  Initially a more prominent tweeter sounds very good.  This is primarily because the OW1 is a very good driver.  The detail is quite stunning.  After a few months the "good driver sound" wears a little thin.  After several years I have overcame the desire to hear good drivers.  I now prefer to hear good music, properly balanced, all day long.  As such, I am very comfortable sending 8, 12.5 resistors will all of my kits. 

Following these remarks I asked Dennis Murphy for his feedback.  I wanted to know if his thoughts were similar to mine.  I also wanted to know if my remarks accurately reflected some simile of truth in the matter of resistor padding and the 1801.  The exchange will be in a question and answer format between Dennis and I.

Dave, "Hey Dennis, what do you think of the remarks on my web page about resistor padding on the 1801b".

Dennis, "Hey--you gotta stop listening to bad recordings.   I think the advice should be--6 + 15 = absolute flat response, probably only appropriate for the best recordings. 7 + 15 = normal setting for normal rooms  7 +
12.5 = attenuated highs for brighter rooms, poorer recordings.  8+ 12.5 for extreme conditions."

Dave, "Okay, and, there might be a plausible reason why we are hearing things the way we are (i.e. slightly different).  I  have wood floors with a  8x11 throw rug backed by a 1/2" carpet pad.  Do you have that nice wall to wall carpeting stuff?"

Dennis, "Wall-to-wall carpet--acres of it--and real fir paneling on the rear wall.  And lots of dog hair.   Very warm acoustics...." 

Dave, "Well, my guess is that we are hearing about the same thing in the 1801b tweeter.  The difference is not even 1 db.  This is rather small.  This could easily be attributed to the dog hair on the walls, or that fancy  wall to wall carpet."

I know that other thoughts exist on this matter.  Therefore, I offer resistor options.  Providing a customer with ONLY the information necessary to purchase a product is a sad dilemma.  The truth regarding the resistor level is that there is no correct option for all rooms, systems, and recordings. 

After 18 months of offering resistor options, the consensus is twofold.  First, there are very slight audible differences between the resistor options.  The differences seem barely audible.  Second, most folks seem very pleased with 8 & 15 ohm resistors in their 1801s.

Upon purchase of the 1801, please let me know what resistors you would like installed/delivered.

Mark Johnson's comments

Hi Dave,

I have not emailed you for a long time, but I did say I would provide feedback after ordering a pair of the 15 resistors to try out. Well, I had some amp problems that took an awful 5 months to get fixed (the pitfalls of buying a foreign built product) and then I left the new resistor combo in for a couple more months and then recently switched back to the original 12.5 and plan to leave it that way. I have a fairly large room with little treatments and didn't find the difference too dramatic, but some harsher recording seem a bit easier to listen to and I don't feel the -2db output sacrifices too much detail.

So....it was a long experiment with no switching back and forth, but you can add me to the group that thinks the combo the kit ships with is the way to go. On some recordings I'd like the extra output, but the difference is so small and it's not worth listener fatigue on other stuff. So I guess you know what you're doing :-)

Hope that helps,

--Mark.