decato
01-11-2007, 10:39 PM
Introduction and Goals
The RTA 15TL was Polk’s top of the line non-SDA model, manufactured from 1990-1992. It features a vertical array of drivers with the tweeter positioned in the center, a configuration known as MTM or D’Appolito, named after its designer Dr. Joseph D’Appolito. The cabinet houses two independently tuned passive radiators, one “high-resonance” in the front and one “low-resonance” in the rear.
After quite some time, I finally got around to working on the RTA 15TL. I listened to them with the stock crossover parts and SL3000 tweeters before making any modifications. I concluded that the tweeter was working pretty hard, probably due to a low crossover point, which made some recordings sound harsh, especially those with a saxophone or electric guitar. Even some vocalists came across with an unpleasant timbre. It also sounded as though the tweeter was a bit too loud, or forward. I had some spare RD0198-1 tweeters lying around, so I gave them a try. To my surprise, they actually sounded worse overall! They were cleaner but even more strident.
The bass of the RTA 15TL was not what I expected. I measured the response of both passive radiators and observed that they have similar output, except that the “low-resonance” one always puts out 3-4 dB less than the “high-resonance” one. (Julian Hirsch reviewed the RTA 15TL in December 1990 and measured a 5 dB variation in output levels. I have attached a PDF of this review for you to read.) I also noticed that the speakers boomed dreadfully when placed close to a wall. There was an excess of midbass, especially in the 100-200 Hz range, but output below 100 Hz was noticeably less. My hypothesis on the abundance of boomy midbass is as follows:
Both passive radiators, despite their different moving masses, operate over the same basic range. However, they are out of phase with each other. If placed close to a rear wall, the waves from the backward-firing passive radiator will hit the wall and be reflected toward the front, now in phase with the front-firing passive radiator but slightly delayed, resulting in slow, boomy bass.
While this type of bass is great for reproducing organ pedals, whose sounds rise and fall slowly, it is less than desirable for almost everything else.
So, I can sum up my evaluations as follows: I like the sound of the RTA 15TL. I think that it can be made to sound even better with some effort. The crossover design could benefit from a little tweaking, and I would like to try a few tricks to improve the bass.
Parts and Procedures
The original design of the crossover (Figure 1) yields the transfer function you see in Figure 2, assuming a DC resistance of 6.8 ohms for the 4 MW6503 drivers in a series-parallel configuration and 5.6 ohms for the single RD0198-1 tweeter. After several iterations of design and evaluation, I have modified the crossover schematic to that of Figure 3, with the new transfer function shown in Figure 4. I did not make a “drastic” alteration but rather a subtle change to the slopes. The tweeter does a bit less work, while the woofers’ high end rolls off more gradually. The crossover point has been shifted up several hundred Hz. The tweeter's phase is pretty close to that of the woofers.
The parts needed for this project came from the following vendors at a total cost of $201.50 + shipping:
RD0198-1 (SL3000 Replacement) Tweeters
Polk Audio
1-800-377-7655
www.polkaudio.com
2 x RD0198-1, $96 (with Club Polk discount)
Solen PB Series Film Capacitors and Air Core Inductors
Madisound Speaker Components, Inc.
608-831-3433
www.madisound.com
4 x 12 uF, $20
2 x 10 uF, $9
2 x 1.75 mH, 16 awg, Sidewinder, custom-wound $30
2 x 0.3 mH, 19 awg, inductor, $5.90
Mills MRA-12 Resistors
Sonic Craft
1-800-689-4434
www.soniccraft.com
2 x 2.7 ohms, $7.00
2 x 0.5 ohm, $7.00
Fee for order less than $25, $5.00
Waldom Universal Connector
Wayne Electronics
973-839-5888
2 x 628-6, $10
Foam Sheets and Damping Material
Local craft store
2 foam sheets, $1.60
Large bag of polyester sheets for a baby’s crib, $10
The two SW105 passive radiators are, unfortunately, priceless. I have a few spare parts that I collected over the years packed away in boxes. I'll talk more about the passive radiators shortly.
The following figures show how the crossover looks with the new components in place. Sorry, but most of the labeling has been worn off the capacitors from conducting too many experiments.
Figure 5: RTA 15TL Crossover with Waldom Connector (makes experimenting easier)
Figure 6: RTA 15TL Crossover, Topside View
Figure 7: RTA 15TL Crossover, Underside View
The new components are significantly larger than the stock parts and require a new layout. The thick gauge 1.75 mH inductor is probably the least useful of the modifications, resulting in just about 1 dB more bass at very low frequencies. Eliminating this component makes the modification easier (and less costly), as no change to the layout is required. I have not tried more expensive capacitors, but I have tried Daytons. I cannot recommend using them, as they sound “lifeless”.
To reduce the muddiness of the bass, I added 2 extra sheets of damping material behind the drivers and 1 extra sheet in the center of the cabinet behind the front-firing passive radiator. As mentioned earlier, the “low-resonance” passive radiator has less output than its sibling, so I removed it from the cabinet. Then I moved the SW100 from the front to the back and installed a lightweight SW105 (used in the Monitor 7 and 10 Series 2) to the front. I obtained a 3 dB boost in bass below 60 Hz and lost 2 dB around 150-200 Hz with this modification. I wish I had detailed specifications about the various passive radiators to scientifically justify my observation.
Figure 8: Dense Damping Material behind Drivers
Figure 9: Damping Material in Bottom of Cabinet
Figure 10: Front-firing SW105 Passive Radiator
Figure 11: Rear-firing SW100 Passive Radiator
To improve imaging while simultaneously masking a defect in the vinyl around the tweeter, I cut out a foam pad and gently attached it to the area around the tweeter with a glue stick.
Finally, I applied 3 coats of Tung oil to the veneer.
Figure 12: Foam Pad around Tweeter
Figure 13: RTA 15TL Front Baffle
Figure 14: Tung Oil Finish
Question: Why are the speakers sitting in the lids of plastic containers?
Answer: To avoid scratching the floor more than I already have!
Listening Evaluation
I am quite pleased with how this modification turned out. The midrange tonality is more natural, the bass is fuller, and the tweeter output is no longer abrasive. By using Solen capacitors, I have experienced the same improvements as many other forum members, so there is no need to describe those sonic changes here. The RTA 15TLs cast a beautiful center image if toed-in properly. The imaging is, indeed, very close to that of the SDA series, except that sound never comes from the far left or right of the soundstage. I still recommend keeping them as far away from walls as possible.
Does this mean that I am 100% finished with this project? No. Those of you who know me recognize that I am almost never fully satisfied with my projects. I always tend to revisit them, especially when new information becomes available or I learn a better method to solve a problem. In this case, my next goal is to obtain more detailed information about the drivers, tweeter, and passive radiators. However, for now, it is time to sit back and enjoy the music!
- Brian Borowski
The RTA 15TL was Polk’s top of the line non-SDA model, manufactured from 1990-1992. It features a vertical array of drivers with the tweeter positioned in the center, a configuration known as MTM or D’Appolito, named after its designer Dr. Joseph D’Appolito. The cabinet houses two independently tuned passive radiators, one “high-resonance” in the front and one “low-resonance” in the rear.
After quite some time, I finally got around to working on the RTA 15TL. I listened to them with the stock crossover parts and SL3000 tweeters before making any modifications. I concluded that the tweeter was working pretty hard, probably due to a low crossover point, which made some recordings sound harsh, especially those with a saxophone or electric guitar. Even some vocalists came across with an unpleasant timbre. It also sounded as though the tweeter was a bit too loud, or forward. I had some spare RD0198-1 tweeters lying around, so I gave them a try. To my surprise, they actually sounded worse overall! They were cleaner but even more strident.
The bass of the RTA 15TL was not what I expected. I measured the response of both passive radiators and observed that they have similar output, except that the “low-resonance” one always puts out 3-4 dB less than the “high-resonance” one. (Julian Hirsch reviewed the RTA 15TL in December 1990 and measured a 5 dB variation in output levels. I have attached a PDF of this review for you to read.) I also noticed that the speakers boomed dreadfully when placed close to a wall. There was an excess of midbass, especially in the 100-200 Hz range, but output below 100 Hz was noticeably less. My hypothesis on the abundance of boomy midbass is as follows:
Both passive radiators, despite their different moving masses, operate over the same basic range. However, they are out of phase with each other. If placed close to a rear wall, the waves from the backward-firing passive radiator will hit the wall and be reflected toward the front, now in phase with the front-firing passive radiator but slightly delayed, resulting in slow, boomy bass.
While this type of bass is great for reproducing organ pedals, whose sounds rise and fall slowly, it is less than desirable for almost everything else.
So, I can sum up my evaluations as follows: I like the sound of the RTA 15TL. I think that it can be made to sound even better with some effort. The crossover design could benefit from a little tweaking, and I would like to try a few tricks to improve the bass.
Parts and Procedures
The original design of the crossover (Figure 1) yields the transfer function you see in Figure 2, assuming a DC resistance of 6.8 ohms for the 4 MW6503 drivers in a series-parallel configuration and 5.6 ohms for the single RD0198-1 tweeter. After several iterations of design and evaluation, I have modified the crossover schematic to that of Figure 3, with the new transfer function shown in Figure 4. I did not make a “drastic” alteration but rather a subtle change to the slopes. The tweeter does a bit less work, while the woofers’ high end rolls off more gradually. The crossover point has been shifted up several hundred Hz. The tweeter's phase is pretty close to that of the woofers.
The parts needed for this project came from the following vendors at a total cost of $201.50 + shipping:
RD0198-1 (SL3000 Replacement) Tweeters
Polk Audio
1-800-377-7655
www.polkaudio.com
2 x RD0198-1, $96 (with Club Polk discount)
Solen PB Series Film Capacitors and Air Core Inductors
Madisound Speaker Components, Inc.
608-831-3433
www.madisound.com
4 x 12 uF, $20
2 x 10 uF, $9
2 x 1.75 mH, 16 awg, Sidewinder, custom-wound $30
2 x 0.3 mH, 19 awg, inductor, $5.90
Mills MRA-12 Resistors
Sonic Craft
1-800-689-4434
www.soniccraft.com
2 x 2.7 ohms, $7.00
2 x 0.5 ohm, $7.00
Fee for order less than $25, $5.00
Waldom Universal Connector
Wayne Electronics
973-839-5888
2 x 628-6, $10
Foam Sheets and Damping Material
Local craft store
2 foam sheets, $1.60
Large bag of polyester sheets for a baby’s crib, $10
The two SW105 passive radiators are, unfortunately, priceless. I have a few spare parts that I collected over the years packed away in boxes. I'll talk more about the passive radiators shortly.
The following figures show how the crossover looks with the new components in place. Sorry, but most of the labeling has been worn off the capacitors from conducting too many experiments.
Figure 5: RTA 15TL Crossover with Waldom Connector (makes experimenting easier)
Figure 6: RTA 15TL Crossover, Topside View
Figure 7: RTA 15TL Crossover, Underside View
The new components are significantly larger than the stock parts and require a new layout. The thick gauge 1.75 mH inductor is probably the least useful of the modifications, resulting in just about 1 dB more bass at very low frequencies. Eliminating this component makes the modification easier (and less costly), as no change to the layout is required. I have not tried more expensive capacitors, but I have tried Daytons. I cannot recommend using them, as they sound “lifeless”.
To reduce the muddiness of the bass, I added 2 extra sheets of damping material behind the drivers and 1 extra sheet in the center of the cabinet behind the front-firing passive radiator. As mentioned earlier, the “low-resonance” passive radiator has less output than its sibling, so I removed it from the cabinet. Then I moved the SW100 from the front to the back and installed a lightweight SW105 (used in the Monitor 7 and 10 Series 2) to the front. I obtained a 3 dB boost in bass below 60 Hz and lost 2 dB around 150-200 Hz with this modification. I wish I had detailed specifications about the various passive radiators to scientifically justify my observation.
Figure 8: Dense Damping Material behind Drivers
Figure 9: Damping Material in Bottom of Cabinet
Figure 10: Front-firing SW105 Passive Radiator
Figure 11: Rear-firing SW100 Passive Radiator
To improve imaging while simultaneously masking a defect in the vinyl around the tweeter, I cut out a foam pad and gently attached it to the area around the tweeter with a glue stick.
Finally, I applied 3 coats of Tung oil to the veneer.
Figure 12: Foam Pad around Tweeter
Figure 13: RTA 15TL Front Baffle
Figure 14: Tung Oil Finish
Question: Why are the speakers sitting in the lids of plastic containers?
Answer: To avoid scratching the floor more than I already have!
Listening Evaluation
I am quite pleased with how this modification turned out. The midrange tonality is more natural, the bass is fuller, and the tweeter output is no longer abrasive. By using Solen capacitors, I have experienced the same improvements as many other forum members, so there is no need to describe those sonic changes here. The RTA 15TLs cast a beautiful center image if toed-in properly. The imaging is, indeed, very close to that of the SDA series, except that sound never comes from the far left or right of the soundstage. I still recommend keeping them as far away from walls as possible.
Does this mean that I am 100% finished with this project? No. Those of you who know me recognize that I am almost never fully satisfied with my projects. I always tend to revisit them, especially when new information becomes available or I learn a better method to solve a problem. In this case, my next goal is to obtain more detailed information about the drivers, tweeter, and passive radiators. However, for now, it is time to sit back and enjoy the music!
- Brian Borowski