Speaker Design Proposal Thomas Young
Functional Description This speaker system is a replacement for the front left/right speakers of an Onkyo HTS-790 home theatre in a box system. The existing speakers feature 5 inch reinforced paper woofers and 1 inch dome tweeters and sound decent for a theatrical setting, however they are very muddy and distant when used for music listening. The ideal replacement speakers should improve the musical listening experience, without sacrificing the film experience. Requirements The desired product should be reasonably close to bookshelf speaker size, and should not be exceedingly heavy, for portability. As they will be integrating into a system which already contains an effective subwoofer, performance below 80 hz is not needed. The speakers should have an emphasis on time domain accuracy, including a tight low end response. As these speakers will be primarily used for home theatre and personal listening, they will not need to be exceedingly loud, the ideal listening volume would be 85dB, which when combined with a 20dB crest factor, would be 105dB. They will be a driven by a 100w/ch home theatre receiver. If possible, it would be preferable that the speakers be shielded, as they will be used close to a TV for the near future. Design Process After specifying the requirements and goals for the resulting speakers, several hours were spent looking at available drivers on the Madisound and Zalytron websites. With a total working budget of around 400 dollars, a guideline of a maximum of $200 for 2 tweeters and 2 woofers was set, or 100 dollars per speaker for drivers. In order to obtain an 80 hz F3 point with a sealed cabinet (as discussed below), woofers smaller than 6.5 inches were not considered. Enclosure Vifa Bookshelf Example [audiodiycentral.com] According to the loudspeaker design cookbook, for a set of sealed bookshelf speakers, an f(sc) of 50 hz or less is adequate [Dickason, 36], however considering that these speakers will used with a subwoofer at 80hz, the decision was made that a smaller box would suffice. The decision to utilize a sealed enclosure was made for several reasons. First, sealed speakers are much more accurate in the time domain, with a Speaker Design Proposal 2
good deal sharper transients. Second, because this system will be tied with a subwoofer, the low frequency performance is less critical, since it will be high passed at around 80 Hz anyway. Upon consideration of the size constraints of this project, along with some simple calculations as to frequency needed [Dickason, 36], it was decided that a box with a volume of.5-.6 cu ft would meet the needs of this project. Pictured above are an example enclosure for the P17*J woofers. [Audiodiycentral.com, Vifa Project] Woofers With these guidelines, several woofer and driver models were selected (also considering price), including the Vifa P17SJ (shielded 6.5 ), the Vifa P17WJ (unshielded 6.5 ), the Peerless PL18 (unshielded), and the Seas Prestige CA18RLY (unshielded 7 ). A sealed box was chosen due to its tighter transient handling and the lack of a need for lower bass frequencies, due to the subwoofer assumption. Working with the assumption of a 0.5 cu foot sealed box, all of these were modeled in WinSpeakerz, with all producing an F3 point of around 80 hz. The exception being the Seas, which in these circumstances, had an F3 of 60 hz. (Attached are the plots for the Vifa P17SJ in both a sealed and vented box and the Seas in a sealed box, plots 1, 2 and 3 respectively). While all the drivers provided a fairly flat response from 150 to 3k [Attached info sheets], the P17SJs (seen above) were chosen for their flatter frequency response, particularly around the critical 2k range, and their lower cost. This can be seen in the included sheets 4 and 5. These drivers cost $58 a driver, meet the specifications of being affordable yet still having a decent frequency response. Additionally, their shielded magnet and the reasonable Vas, making it work well in a smaller enclosure designed for home theatre work. Tweeters The selection process for tweeters was significantly simpler than that for the tweeters. Based upon the class discussions about the Vifa D27TG45, the decision was made early on in the design process to utilize these drivers as they are a smooth tweeter. Because they feature a 1 inch soft dome, they are well suited to use for casual monitoring and can be crossed over at a lower frequency. [ldsg.snippets.org Appendix B] The stock specifications for this tweeter specify a free air resonance of 650 hz, which means that a crossover frequency of 1.4 khz would be possible, however further reading suggests that new models Speaker Design Proposal 3
have a free air resonance of 850 hz, so the crossover frequency will need to be a bit higher, probably 1.7 khz. The D27TG45s lip around the tweeter improves it s diffraction control, which will be good for use in smaller rooms, additionally, due to its lower fs, it will be easier to design the crossover, which meets another goal of this project. [ldsg.snippets.org, Vifa Info Page] Enclosure Revisited Now that drivers have been chosen, the design of the cabinet needs to be fine tuned. Upon modelling in winspeakerz, a tweaked cabinet size of.55 cu ft seems to be approximately right for the P17SJ, giving it an F3 point of approximately 81 hz. This can be seen in the P17SJ winspeakerz model (sheet 1). This model was then converted to a 4th order vented box, with a port tuned at 43 hz. This yielded an impressive F3 of 56 hz, which would be very decent as a standalone bookshelf system. While for now, the project will remain a sealed box, the box design will be designed so that a port can be added later if desired. [Murphy, 29] In order to obtain an internal volume of.55 cubic feet, careful calculations had to be made to so that the internal reflections of the cabinet were not going to be a problem. To this end, internal dimensions of the following were made. A height of 15.5 inches, a width of 9 inches and a depth of 6.8 inches. The reason these were chosen was so that there were no common dimensions (to cut down on standing waves). Additionally, careful consideration was made so that no frequencies were directly at the crossover frequency, which was achieved. In particular, the depth of the box was tuned so that it was not at the frequency of the crossover point. With a bookshelf system like this, it is difficult to completely avoid the crossover, due to the sizes involved. To clean up the diffractions of the cabinet, the edges of the front of the cabinet will be rounded, which should clean up a good majority of the diffractions created by the tweeter, which will be flush mounted onto the baffle. With regard to the actual construction of the box, the actual outward physical appearance of the cabinet will be important, as these will be sitting prominently in a living room on the home theatre/music system. To this end, the outward appearance shall be a hard, multi-plywood board, of atleast 1/2 thickness, depending on the availability of wood in the class. This will be stained for a dark high gloss finish in the end. This outer construction shall be laminated onto 1/2 medium density fibre board (which will be the inner layer). The front panel of the speaker will be 2 1/2 sheets of MDF, painted black, to cover up the rounded corners. This should provide a solid enclosure, with a decent amount of dampening, while still maintaining a relatively light and portable enclosure. Careful consideration was given to tweeter placement. The decision was made that the tweeter will be centered on the front baffle, so as to keep all reflections simple. The tweeter will be placed as close as Speaker Design Proposal 4
possible to the woofer, to maintain as much phase coherence as possible. Both drivers will be flush mounted into the front baffle. The cabinet will be internally braced, with the brace cutout behind the woofer to avoid reflections. Crossover Filter The speaker system will feature a passive crossover to each of the drivers. The decision to do this was based on the higher cost of having an active system and that it will be integrating into an existing home theatre system. The crossover frequency will be at 1.8 khz. This decision was made by looking at the frequency graphs for the tweeter and the woofer. The woofer loses it s flatness around 1.6 khz, and the tweeter is flat down to 1.5 khz, so the decision was made to place the filter in the middle of this. The crossover shall be a 4th order Linkwitz-Reilly type. If needed, an inductance compensator will be added to the woofer to counteract the rising inductance of the woofer. According to the LDSG, newer D27TG45s are more harsh sounding than some of the previous models. If this becomes an issue, an inductance compensation circuit [Murphy, 109], will be added to the tweeter to smooth out the tweeter s sound. Summary In summary, this project will create a pair of replacement home theatre monitors, with optimized performance for music. They will be partially shielded for placement near a TV, and will have a frequency response from 80 Hz to 20 khz, +/- 3 db. Speaker Design Proposal 5
Sources Dickason, Vance. The Loudspeaker Design Cookbook. Audio Amateur Publications. 2005 Murphy, John L. Introduction to Loudspeaker Design True Audio. 1998 Kreskovsky, John. Vifa P17WJ-00-08/D25AG-35-06 Two -way System. 10/1/2007. http:// www.audiodiycentral.com/sd200_johnk.shtml LDSG Snippets. http://ldsg.snippets.org/ Vifa Spec Sheets Peerless Spec Sheets Winspeakerz Modeling Speaker Design Proposal 6