The biggest challenge faced by full range driver enclosure designers is producing enough bass output to match the rising midrange response typically generated by the low Qts values of full range drivers. If the designer is really serious about using only one driver, without any passive circuit elements between the driver and the amplifier, to cover the entire audio spectrum then in most cases a hybrid transmission line / back loaded horn enclosure design is required. Getting a hybrid transmission line / back loaded horn enclosure to work well is not a simple task. If a passive correction circuit (really a modified BSC circuit with additional midrange attenuation) is acceptable to the designer then a resonant style of enclosure, a bass reflex enclosure or some form of quarter wave enclosure, can be used to produce a very balanced SPL response at the expense of some loss of efficiency. The second approach was used in my Projects 3 and 4 and is also applied in Projects 5 and 6.
An alternate design approach that introduces a second driver into the system utilizes a larger diameter high efficiency woofer to produce the demanding bass frequencies allowing the full range driver to cover the remaining majority of the audio spectrum. This is an appealing idea since the large excursions required to produce significant bass output are removed from the smaller diameter full range driver which also tends to have Xmax values of less than or equal to 1 mm. A couple of years ago, I decided to try this approach for my next project. After purchasing four high efficiency 15" diameter drivers the project started and stalled a number of times, as I ran into design obstacles, before arriving at the open baffle concept which was eventually built during December of 2005.
I searched the Internet for 15" and 18" high efficiency woofers to pair with my Lowther full range drivers. My search criteria was an efficiency of greater than 95 dB, for 1 watt input at a 1 m distance, and a Qts value of approximately 0.4 so that a resonant enclosure could be easily designed. I looked at a number of drivers from manufacturers like Eminence, JBL, Electro-Voice, and Dayton all of which could be purchased from Parts Express.
After trading off the different driver parameters, I down selected to the JBL 2226H and the Dayton 15" high efficiency woofers. Comparing the specifications between the two drivers showed almost identical values. Comparing the prices showed that I could buy four of the Dayton drivers for less than a pair of the JBL drivers. I started to think about an Isobaric design which would cut the final size of the cabinet in half. Following this logic, I purchased four of the high efficiency Dayton 15" cast frame drivers with 4" voice coils. The build quality of the Dayton drivers was excellent and after running them for many hours of break-in I measured the T/S parameters of the four drivers. The measured T/S parameters were in close agreement with the Parts Express data sheet and the drivers themselves were very consistent.
Using the measured T/S parameters, I started investigating different Isobaric enclosure designs. This is where the design problems (challenges?) started to surface. There were three basic issues which eventually killed this design concept. First, the isobaric woofer system was only going to be ~96 dB before accounting for a baffle step loss which meant that any of my Lowther drivers with efficiencies greater than 95 dB would need some form of attenuation to bring the system SPL response into balance. Second, any bass reflex or transmission line enclosure for the bass drivers would result in a double humped impedance curve making a passive crossover (I had envisioned using a single SS 200 watt amp) very difficult in the frequency range between 100 Hz to 200 Hz. And finally, the enclosure was still going to be large, heavy, and difficult to construct and seal. The idea stalled for over two years while I thought about and analyzed different enclosure and passive crossover options.
Over the past two years, I have been adding baffle step response calculations to my MathCad worksheets. Once a baffle step calculation algorithm is programmed, switching to an open baffle calculation is very easy. As I continued to work on my MathCad transmission line worksheets, an open baffle worksheet was written in parallel and I started to experiment with different sized baffles and driver placements.
To get calibrated, the response of the Lowther PM2A and a single Dayton woofer were calculated for an infinite baffle mounting. The calculated infinite baffle SPL responses are shown in Figure 1. From these plots it can be seen that both drivers are about 97 or 98 dB efficient but start to roll off before ever reaching down into the lower bass frequency range.
After spending some time trying different sized baffles and driver positions, the arrangement shown in Figure 2 seemed to provide a good result. One goal was to position the Lowther at about ear level and at least one of the woofers near the floor boundary. The final baffle dimensions were 48 inches tall and 60 inches wide, unfortunately not very small.
Figures 3 and 4 show the driver SPL response for each driver at a 1 m distance on the axis of the Lowther PM2A driver. Figure 5 combines the responses for the two Dayton woofers. Figure 5 is the key result that helps this open baffle project produce adequate bass output. The bass output from the two Dayton woofers produces a SPL of 106 dB before rolling off, crosses the 100 dB level at 50 Hz, and is down 6 dB at approximately 40 Hz. By using a crossover with a 12 dB/octave slope located at a frequency between 100 and 200 Hz the higher output of the woofers can be gradually attenuated while maintaining the SPL of the lower bass output.
Figure 6 presents the final system combined SPL response on the axis of the Lowther PM2A driver at a distance of 1 m for a 1 watt input. The crossover conditions are also listed in Figure 6. This open baffle speaker system is approximately 100 dB efficient down to nearly 50 Hz and is 6 dB down at about 35 Hz. Another real benefit to this approach can be seen in Figure 7, the displacement requirements for the Lowther PM2A are very small. By crossing over to a second set of drivers that produce the bass frequencies, the Lowther is operating with very little required displacement to produce its contribution to the total system's SPL output.
Low Pass Filter - 2nd Order L-R at 100 Hz with a 3.5 dB Boost
High Pass Filter - 4th Order L-R at 100 Hz
i
The complete simulation of this Lowther open baffle speaker system using the Dayton woofers is attached as a pdf file of the MathCad worksheet. This worksheet is the default configuration that can be downloaded from the Upgraded MathCad Computer Models page found elsewhere on this site.
A 60 inch wide open baffle was going to be too big to have in my listening room so I decided to used hinged wings that could be folded back when I was not listening to music. With the wings folded back the front baffle is reduced to a more manageable 24 inch width which is still quite big but small enough to allow me to get behind the speakers to access the wiring and power cords for the amps and crossover. A picture of the folded back wings is shown below. A completely dimensioned drawing in pdf format is available for people interested in more details of the speaker construction. Since this was more or less an experiment, I painted the baffles to match the wall color. My goal was to have them blend into the room and not be so noticeable, it didn't work.
As I stated early on, in my opinion using a passive crossover was not really feasible for this type of design due to the significant peaks in the electrical impedance curves for the woofer drivers at low frequencies. At about the same time I was formulating this design my almost 20 year old Adcom 200 watt amp and in particular the preamp started to show signs of age and replacing them could be justified. Looking at the various active crossovers available, it became obvious that an amp with XLR connections would be an advantage if I wanted to use one of the reasonably priced pro audio active crossovers.
After much searching and comparing cost/performance trade-offs I decided to upgrade my preamp to the Rotel RC-1090, upgrade to a pair of Rotel RB-1080 200 watt per channel amps, and add a RT-1080 tuner to completely replace my old tired Adcom units. I auditioned these in my home before buying and found that the Rotel set-up had the same level of detail as the now popular T-amp designs (I was using a pair of TEAC A-L700P amps as monoblocks for a while during the transition) but with the solid deep bass output I felt most of the early T-amps lacked. The performance standard I was using as a benchmark was a Bryston set-up I had heard connected to my Lowther ML TL speakers in a friend's home. I would have really liked to have gone with the Bryston preamp and Bryston amps but the cost was more then I could justify. The Rotel was close enough in performance to satisfy my needs but at a significantly lower price.
The search for a crossover was a little frustrating. My original goal was to use a Behringer DCX2496 Ultradrive Pro Digital Crossover but it has been unavailable for well over six months. While looking for alternatives, I found the dBx Driverack PA which appeared to provide most of the features of the Behringer with the addition of an automated room EQ function. I have to admit I had an immediate attraction to the dBx product having owned a few original dBx products that still completely function 20 years later. Probably not the same dBx company, but the name brings back many good memories.
Finding reasonably priced XLR cables was the final challenge. I searched the various on-line audio vendors for XLR cables and was astounded at what was being charged for some of the boutique cable products. So I turned to E-Bay. I found an E-Bay seller who listed very reasonably priced XLR cables of different lengths and in various quantities. These XLR cables were the final piece that allowed me to connect the amplification and the active crossover to finally hear my Lowther Open Baffle Speaker System.
The bottom line is the sound. These speakers produce a very large sound with incredible dynamics and power. The characteristics of the presentation are very similar to my Lowther DX3 ML TL speakers but the effortless dynamics are much better. By having such a large effective cone surface area, for the bass frequencies, the reproduction of a drum solo is getting closer to the point of feeling and hearing the drummer. The bass is not overwhelming or boomy, it is tight and in proportion to the mid range and high end SPL response. These speakers will not shake the house but do generate an effortless reproduction of acoustic jazz, with great imaging, that I really enjoy.
One of the interesting things I noticed right away was that the driver cones barely moved even at very loud volumes. You have to touch each driver to convince yourself it is connected and contributing to the system SPL output. The next interesting observation occurred while turning the system off late one night, I shut off the amp driving the woofers before the one driving the Lowthers, the next night I did the opposite. This is a classic example of the whole being greater than the sum of the parts. There is no way I would have ever believed that the muted and one dimensional sound coming from these two halves of the speaker system could combine to produce the dynamic engaging performance I was hearing just five minutes earlier. This has been a fascinating result and a common demo I play for guests.
One of the down sides to this set-up is a slight background hiss produced by something in the audio chain. While this is not really obvious with my lower sensitivity Lowther ML TL speakers, it can just be heard from the open baffle system when no music is being played. The volume of the hiss is less than what I have heard in a few high end audio emporiums but it is greater than what I was experiencing with the TEAC A-L700P amps. Not objectionable, but not totally silent. I am not sure what is causing this slight hiss but I think that the almost 100 dB system efficiency is exaggerating the hiss and allowing it to become audible. When music is being played it is not a factor. I intend to continue investigating this and try to determine if it is the crossover, the amps, the preamp, or the interconnects.
The second down side is the size. While this was a very easy speaker system to construct the final result is huge. Even with the wings folded back the front panel is bigger then anything I have ever had in my stereo system in the past. My room is probably a little too small for such a large speaker but to be honest I don't really hear any significant problems caused by room size. I am enjoying the open baffle experience while it lasts before they have to go to make room for the next speaker project.
I have two experiments in mind for these speakers. I am going to get a third Rotel RB-1080 amp since the dBx Driverack PA can handle a three way crossover. Then I am going to study two upgrades to the system. First, a super tweeter will be added to the top of the baffle and crossed over from the Lowther someplace above 10 kHz. It will be interesting to hear (or maybe not hear) what potential benefit a super tweeter might bring. Second, I plan to split the control of the two woofers so I can roll them off at two distinct frequencies. By applying more boost to one of the woofers I hope to push the usable bass down lower and have the system response remain flat to below 40 Hz before it starts rolling off. I will add the results of these experiments at a later date.
This is a slight tangent from my original planned path of upgrades, as described above, but it is too interesting to not document. On the DIYaudio forum a few weeks ago, a question came up about the possibility of designing an inexpensive open baffle system. Having recently completing this open baffle project, I performed some quick simulations using a few different 15 inch woofers paired with Fostex FE-167E full range drivers. The results looked quite promising with the Eminence Alpha 15A woofer, which has a Qts of approximately 1.2 and is very reasonably priced, supplying the bass below 200 Hz. The relatively small size of the required open baffle to achieve bass extension down to 50 Hz really peaked my interest in this Eminence woofer. While the inexpensive open baffle system will have to be the topic of another project at some later date, the Eminence woofer was an interesting potential substitution into this Lowther open baffle speaker system.
I have to admit, I like working with low Qts drivers but I could not resist running a few more simulations pairing the Lowther PM2A full range driver with two Eminence Alpha 15A woofers just to see if some additional bass extension was possible compared to the response shown above in Figure 6 using the lower Qts Dayton woofers. Just to get a handle on the SPL response differences between the two woofers, the infinite baffle SPL responses of the Dayton and the Eminence woofers are shown in Figure 8. As can be seen in Figure 8, the Eminence woofer is not as efficient as the Dayton above 100 Hz but below 100 Hz the response is much flatter and extends down to 40 hz before rolling off at 12 dB/octave. I could not resist and decided to try the higher Qts Eminence Alpha 15A woofers and ordered four from Parts Express.
I think it is a safe bet to speculate that the Dayton and Eminence drivers are manufactured by the same company. The build quality of the Eminence drivers was also excellent and the driver cut-out and bolt hole pattern matched the Dayton drivers resulting in an easy swap without requiring modifications to the big white baffle. After running them for several hours, I measured the T/S parameters of the Alpha 15A four drivers. The measured T/S parameters were in close agreement with the Eminence data sheet and the drivers themselves were very consistent.
After substituting the measured T/S parameters into the Open Baffle MathCad worksheets, the crossover frequency and required boost applied to the woofers were irterated to obtain a flat SPL frequency response. Figure 9 shows the calculated results and the details of the crossover filter settings for the dBx Driverack PA. Comparing Figures 6 and 9 shows that the Eminence woofers produce a flatter and deeper bass response. This version of the open baffle speaker system is approximately 100 dB efficient down to nearly 40 Hz and is 6 dB down at about 30 Hz. This calculated improvement in bass response matched what I heard while playing my collection of reference CDs one evening after the swap.
Low Pass Filter - 2nd Order L-R at 100 Hz without any dB Boost
High Pass Filter - 4th Order L-R at 100 Hz
It is extremely easy to adjust the crossover settings on the dBx Driverack PA while listening to music. With the pressing of a few buttons and the spinning of a dial one can move the crossover's cut-off frequency, change the slope, and add or subtract a few dBs of SPL. On the fly, you can fine tune your speaker performance to sound its best in your room. I modified the original settings of the crossover for this new combination of Lowther and Eminence drivers and arrived at an adjusted system response shown in Figure 10. This is the combination that sounds the best to me in my room ..... at least today.
Low Pass Filter - 2nd Order L-R at 200 Hz with a 2.0 dB Boost
High Pass Filter - 4th Order L-R at 200 Hz
The differences in the performance with the two different woofers can easily be described. With the Dayton woofers, the bass was very tight but a little lean as a result of the early roll-off of the low Qts driver really limiting the low frequency output. The Eminence woofers provided a much deeper and warmer bass presentation. The Eminence bass output might not be quite as tight as the Dayton but the added depth and warmth were more to my liking. I think the Eminence drivers were a definite upgrade to the Lowther OB system at a reduced cost, a winning combination. The complete simulation of this Lowther open baffle speaker system using the Eminence woofers, consistent with the results shown in Figure 10, is attached as a pdf file of the MathCad worksheet.
One last observation/sytem improvement that resulted from this upgrade. When I swapped out the woofer drivers I also upgraded the "internal" wiring of the OB speaker system. Basically I had taken a short cut back in December and quickly wired the speakers using whatever speaker wire I had on hand and some cheap crimped connectors. It was a real collection of mismatched parts. When I had the speakers apart I took the time to carefully redo the "internal" wiring with lower gage cables and the correct connectors. This seems to have significantly reduced the hiss I had originally noticed. Some hiss is still present but it has definitely been reduced in volume level.