dB717TL

Page Created 12/25/04; Last Revised 5/29/06

 
This page shows a new speaker which in essence is a modification to the dB616TL. The difference is that it makes use of the new Dayton RS180 aluminum cone woofer and several high performance dome tweeters.

This project can be done as an upgrade to existing dB616TLs or as a new project using the basic dB616TL construction with some changes in dimensions (a new drawing is shown below). The new enclosure is wider to account for the increased woofer size, and accounts for the large magnet assembly. If you build the project from ground up and don't use the add-on baffle, be sure to move the first reflector in the transmission line rearward to make room for the larger magnet assembly.

For the first version I used the Vifa D25AG-35-06 aluminum dome tweeter and the initial description covers this version. But then, in pursuit of a fully shielded version, I looked for shieded tweeters and found the Usher 9950C-15 and the SEAS 27TFFC/TV. I ended up trying the SEAS tweeter and was very pleased with the results. Then, in mid-year 2005, the Dayton RS28AS shielded aluminum dome tweeter became available. The results were very nice and this is the version I am using now for my home theater. I should note that, since, at the time I tried the RS28, I also tried a Cauer-eliptical crossover, the RS tweeter project is shown separately at the bottom of this page with its own parts list.

So, click here for: D25AG-35-06 version 27TDFC/TV version RS28AS version...or just scroll down.
 
Click here to see some comments and pics of dB717TL projects built by some avid DIYers.
   

Here's a pic of the dB717TL, RS28 version, in my garage.

   

   

Dayton RS180S-8

Vifa D25AG-35-06

 
In designing this project it must be kept in mind that a low crossover point will help deal with the aluminum woofer cone's sharp resonant breakup which occurs at around 7 KHz. One thing that will help with this is a tweeter with a low Fs (under 1 KHz). Fortunately, many of these are available and many sound very good. Now, since the RS woofer is shielded, it makes sense to use a shielded tweeter if at all possible. Unfortunately, the shielded version of the Vifa D25AG has an Fs much higher and that more or less rules it out. It may be that the shielding magnet and cup preclude use of the dampening chamber that the unshielded version employs and that's why the elevated resonance frequency.

So, the search for a shielded tweeter continued. The Usher 9950 (Fs = 700 for the shielded version) comes to mind, although it's quite a bit more expensive than the Vifa. Another is the SEAS 27TDFC/TV (Fs = 700), shielded version of the TDFC which has been praised by quite a few DIYers. It's priced similar to the Vifa too.
   

Usher 9950C-15

SEAS 27TDFC/TV

 
In mid-year 2005, the RS28 tweeters became available from PartsExpress in both shielded and unshielded versions. The shielded version has an Fs of 700 Hz and although more pricy than the SEAS or Vifa tweeters is still reasonably priced.
   

Dayton RS28AS

 
   
So, what follows below is, first, a description of the enclosure and baffle modifications, then crossover designs for the Vifa and SEAS tweeters with a parts list, and finally the Cauer-eliptical crossover design for the Dayton tweeter with its own parts list. (I should mention that the Cauer-eliptical design isn't specifically called for or needed in the RS tweeter version. I just decided to try the C-e design at the same time.) By the way, and to preclude any extra confusion, I never tried the Usher tweeter.
   

ENCLOSURE -- MODIFIED dB 161TL
If you're modifying/upgrading dB616TLs you built, the best way to mount the new drivers may be to use an add-on overlay baffle. Even though there's just enough space on dB616TL's baffle for the drivers, enlarging holes is difficult to do. Also, the tweeter's faceplate is smaller than the Dayton Silkie's so you'd end up somehow filling the gap around the tweeter.

The new baffle is 1/2" wider than the enclosure and is intended to be mounted so that 1/4" sticks up above the top of the enclosure.

   
I think the appearance with the add-on baffle isn't all that bad. If you disagree, you can run it all the way down to the bottom of the speaker and the "monolithic" tower appearance will be maintained. I leave it up to you.

   
My modified dB616TLs with the D25AG35-06 on the right, 27TDFC/TV on the left.

   

ENCLOSURE -- dB717TL NEW CONSTRUCTION

   
If you're building the enclosure from scratch use the drawing to the right. It's basically the same as the dB616TL except that it's widened by 7/8". You can refer to dB616TL for construction details and techniques. Note the dimensions of the first reflector because of the large motor structure on the RS-180.

 

 

MATHCAD SIMULATION (Courtesy of Martin King's great work which you can see at www.quarter-wave.com)

   
This is a MathCad simulation of the transmission line response of the dB717TL. Ripple in the response is a function of the line properties and the amount of stuffng used. Generally though, these variations are minor compared to those caused by the room boundaries. Placement of the speaker in relation to the rear wall will have a large effect in the overall bass response. My experience from listening is that placing the speakers 2 to 3 feet from the wall behind produces the best sound while still providing solid bass down to 30 Hz.

 

   
 

 



CROSSOVER FOR D25AG35-06

   
Please be aware that the impedance of this speaker, like many dual woofer designs, dips to slightly less than 4 ohms. You should make sure your amplifier is 4 ohm capable.

The following shows the crossover development for the RS180s with the Vifa aluminum dome tweeter. I had originally thought that I would develop both the Vifa and the SEAS concurrently, but I found that it becomes a bookkeeping nightmare...		  
   
This is the initial crossover schematic. Note that inductor values are in uH so that, for example, 900.0 uH is 0.9mH.

 

   
This is the Speaker Workshop simulation using measured spl at 30 degrees off axis and impedance. The rise at 1 KHz, just before the low pass roll off is typical and a characteristic, I believe, of instituting a steep filter at this crossover frequency. In addition to the possible sonic effects due to the rise, there is a phase problem which causes the tweeter and woofer to sum poorly. It's not seen in this pic (because I used Woofer Tester impedance with the phase zeroed), but measurements of the actual speaker/crossover showed it clearly. In addition, based on listening, the tweeter level needs to be reduced.

 

   
Addition of a series notch filter at 980 Hz plus a 3 dB L-pad is shown here.

 

   
This is the resulting simulation.

 

   
And the system measurements. The 1 KHz rise is improved and the crossover summation problem is gone.

   

CROSSOVER FOR SEAS 27TDFC/TV

   
This is the on-axis measurement of the system with the 27TDFC/TV tweeter using the same crossover.

   
Adjustments to the RC network and addition of a trap filter to reduce the rise at 3500 Hz resulted in this crossover. However, even though the measurements showed a very flat response through this range, the midrange sparkle and dynamics that existed before were gone.

   
Staying with the adjustment to the RC network and going without the trap sounds best right now.

The parts list should now be correct for this crossover. I also corrected a previous error on the list and on the schematics which showed Rz = 15 ohms and Cz = 4 uF. These number were reversed. Rz should be 4 ohms and Cz should be 15 uF.

Parts and Cost for One Speaker
(Note different parts for the Seas TD27TDFC/TV version)

Item

Quantity

Cost/Each

Price

Dayton RS180S-8 7" Woofer

2

$30.74

$61.48

Vifa D25AG-35-06 Tweeter
(Seas 27TDFC/TV Tweeter)

1

32.45
(27.50)

32.45
(27.50)

3/4" MDF, 4' by 8' Sheet

1

20.00

20.00

Polyfil Stuffing, 20 Ounce Bag

1

2.89

2.89

Terminal Cup

1

3.95

3.95

Crossover Parts (Parts Express part number)  

 

 

C1 (027-426) 8.2 uF

1

2.75 

2.75 

C2 (027-432) 15 uF

1

4.10 

4.10 

C3 (027-350) 33 uF

1

 .85

 .85

C3b (027-424) 6.8 uF

1

2.25

2.25

Ce (027-418) 3.0 uF
(027-427, 6.2 uF for Seas 27DTFC/TV)

1

1.67
(2.15)

1.67
(2.15)

Cz (027-432) 15 uF

1

4.10

4.10

Cn (027-348) 22 uF

1

.75

.75

Cnb (027-424) 6.8 uF

1

2.25

2.25

L1 (255-216) 0.27 mH

1

2.34

2.34

L2 (255-246) 0.9 mH

1

3.39

3.39

L3 (255-244) 0.4 mH

1

2.58

2.58

Ln (255-242) 0.8 mH

1

3.22

3.22

R1 (005-3) 3 ohm non-ind.

1

3.50

3.50

R2 (005-15) 15 ohm, non-ind.

1

3.50

3.50

Re (004-2) 2 ohm, non-ind.
(004-1 for Seas 27TDFC/TV)

1

.98

 .98

Rz (004-4) 4 ohm, non-ind.

1

1.25

 1.25

Rs (004-2) 2 ohm, non-ind.

1

.98

.98

Rp (004-20) 20 ohm, non-ind.

1

1.25

1.25

Rn (004-6) 6 ohm, non-ind.

1

1.25

1.25

Grille Cloth

1/2 Yard

7.50

3.75

Grille Guides, Small

8

1.90

1.90

Glue

1/2 Bottle

4.00

2.00

   

 

 

   

Total

$171.38
($166.91)

Notes:

C3 and C3b paralleled = 39.8 ohms to meet 40 ohm value.

Cn and Cnb paralleled = 28.8 ohms to meet 30 ohm value.

 
   


CROSSOVER FOR RS28AS
   
This is the basic Cauer-elliptical crossover schematic. Note that the tweeter polarity is reversed.

   
The SpeakerWorkshop simulation shows a slightly rising high frequency response and a peaking at around 10K Hz..

   
Adding Rx and Cx to tilt the tweeter's rising response down by a couple of dB.

   
Adding Rn, Cn, and Ln as a notch filter centered on around 10K Hz.

   
This is the SpeakeWorkshop simulation with the notch filter which very effective flattens the high frequency response.

   
On-axis (green), 30 degrees off-axis (red) and 60 degrees off axis (yellow) shown here.

   
Overall the sound of the dB717TL with the RS28AS tweeter is exceptionally open without any tendancy to glare. Although the measurements show the effect of the notch filter, I can't really hear a major difference with or without it. Your ears may vary, so my suggestion would be to try it yourself. You may also want to try the design without the Rs/Rp L-pad.

The pic shows the crossover. Wired with reversed polarity for the tweeter.

 

   
 

Parts and Cost for One Speaker

Item

Quantity

Cost/Each

Price

Dayton RS180S-8 7" Woofer

2

$30.74

$61.48

Dayton RS28AS-4

1

49.45

49.45

3/4" MDF, 4' by 8' Sheet

1

20.00

20.00

Polyfil Stuffing, 20 Ounce Bag

1

2.89

2.89

Terminal Cup

1

3.95

3.95

Crossover Parts (Parts Express part numbers)      

C1 (027-428) 10.0 uF

1

3.15 

3.15 

C2 (027-432) 15.0 uF

1

4.10 

4.10 

C3 (027-352) 47.0 uF (NPE)

1

 .90

 .90

C4 (027-350) 33.0 uF (NPE)

1

.85

.85

C5 (027-418) 3.0 uF

1

1.67

1.67

Cx (027-418) 3.0 uF

1

1.67

1.67

Cz (027-348) 22.0 uF (NPE)

1

.75

.75

Cn (027-414) 2.0 uF

1

1.52

1.52

L1 (255-216) 0.27 mH

1

2.34

2.34

L2 (255-234) 0.6 mH

1

3.02

3.02

L3 (255-260) 1.5 mH

1

4.47

4.47

L4 (255-234) 0.6 mH

1

3.02

3.02

Ln (255-202) 0.1 mH

1

1.83

1.83

Rs (004-1) 1 ohm non-ind.

1

1.25

1.25

Rp (004-8) 8 ohm, non-ind.

1

1.25

1.25

Rx (004-2) 2 ohm, non-ind.

1

1.25

1.25

Rn (004-4) 4 ohm, non-ind.

1

1.25

1.25

Grille Cloth

1/2 Yard

7.50

3.75

Grille Guides, Small

8

1.90

1.90

Glue

1/2 Bottle

4.00

2.00

       
   

Total

$179.71

Notes:

Non-polarized electrolytic caps (NPE) may be upgraded or bypassed with film and foil or metallized poly propylene caps. If you use bypass caps, parallel the NPE cap with a value of approximately 0.1 percent of the NPE. Upgrading caps can get expensive. For example, C1 (10 uF) is $3.15 for the Dayton poly versus $43.13 for an AudioCap Theta. I leave this up to you!

I don't recommend changing inductors to larger wire gauges. The inductors used in the design are 18 ga. and their resistance (DCR) has been taken into account and used to make the speaker's response what it is.
 
   
   
   
   
Comments by Jeremy N:

Sorry, I emailed with some questions probably a good six months ago concerning the db717tl speakers I built--then I built them, and now I'm realizing I totally forgot to send you both pictures and a thanks.

The speakers themselves are completely assembled/debugged (had some relatively minor issues w/box stuffing, fitting the crossover in the box, etc--but overall, the plans were very good). I used the RS180/RS28 combo with the second variant of the crossover (the one without the notch filter to flatten the hump in the response at 10,000 Hz, because I like things bright =) in MDF enclosures.

The only oddity was the amount of box stuffing; i ended up removing most of it because I found that it sounded better without a lot of stuffing. not sure if this is my individual taste, or if something is screwy, but they definitely don't sound "boomy" in that configuration. They sounded really muted when I overstuffed them (*really* muted), but upon fixing that, it was liking opening the door to the chicago symphony or something similarly melodramatic. They sound really, really great, and I couldn't be happier with them.

The only thing left is to finish the cabinets, which I've omitted doing for the last four months because I've been a completely swampped at the office; I want to sand them down, primer them, and apply car paint.

   
Here's some information from Mark C:

Mark tried out the dB717TL crossover on his Dr. K MTMs. I haven't tried doing any upgrades to the crossover myself, but from what Mark says, it should definitely be worthwhile. He initially experienced some harshness in the sound but it went pretty much away after bypassing the 22uF, 33uF, and 47uF electrolytic caps. Here's what he said.

"I added a .47uf bypass cap to the last 47uf NPE in and effort to tame the tweeter down a little bit and it seems to have knocked out most of the remaining harshness. I've used all Sonicaps for the bypass caps and I think the speakers are just about where I want them now. I have a rebuilt Dynaco ST70 (DIYTUBE.COM board), an ef86 linestage, and a modded Sony SACD player...not the most expensive equipment out there but pretty solid.

"I spent a lot of time battling this harshness before and I'm pretty sure I have the source end of it ironed out. Sometimes it still turns out to be a small bias adjustment. I usually toe the speakers out just a bit so I'll play around with position on them some. I posted on the PE tech board a while back asking for some help with these speakers and have gotten a couple of emails from folks asking what Xover I finally settled on. I hope you don't mind that I recommended your schematic as it might result in a few more emails questions for you. :)

Mark"

   
Here's a pic of Mark's speakers. The Dr. K MTM is a vented design and the question comes up of adapting the dB717TL crossover. Dr. K (Darren Kuzma) designed this speaker for the pre-built PartsExpress 1.0 cu.ft. MTM enclosure, however, I don't know if Mark built his to the same dimensions and port tuning or not. In any case, the baffle width is fairly close to that of the dB717TL so lower midrange/upper bass response should be similar. The offset tweeter will change the midrange and higher response but that will also vary depending on how the speakers are placed -- toed in or straight ahead.

   
Back to Speakers

Back to dB616TL