Cardas Internal Wire Replacement For The SDA SRS 1.2TL
I decided to take a cue from Roger Russell and rewire my SDA SRS 1.2TL's with Cardas copper Litz wire in 15.5 gauge. The Cardas wire is slightly larger than the stock wire, which is 16 gauge. This Cardas wire is the same wire that Mr. Russell uses in his $18,000 IDS-25 loudspeaker system. However, Mr. Russell does not believe the Cardas wire offers any sonic advantage over ordinary wire. He just uses it for marketing reasons. [Link: Cardas wire and the IDS-25]
There is nothing wrong with the stock wire. I just wanted to see if the better wire would make a difference in this application. The Cardas wire is a lower noise design due to the Litz configuration (reduction of strand interaction by insulating individual wire strands) and due to the polished surface of the conductors. AudioQuest uses similar technologies in their cables.
The wire was conditioned for four days (96 hours) on the Audiodharma Cable Cooker prior to assembly.
The Cardas wire measured better with respect to conducted noise, but actually sounded much worse than the original wire.
Figure 1. Main ingredients: wire and connectors.
Figure2. 1/4" WBT quick disconnects, which are gold plated copper.
Figure 3. I found that I could use the WBT disconnects with the narrow tab on the RD0198 tweeters if I
gently pried up the middle band on the back of the disconnect and insert the tweeter tab under it.
Cardas wire has each individual strand coated with a clear polyurethane varnish. I tried two different Dremel attachments to remove the varnish:
428-Carbon Steel Brush - "Remove dirt, rust, and corrosion from most surfaces. Clean electrical connectors and other metal parts, remove buildup from faucets, clean camp stove burners and more."
530-Stainless Steel Brush - "Use on stainless steel, pewter, aluminum, silver and other white metals. Use for cleaning, deburring and surface finishing."
The carbon steel brush worked faster and better.
Figure 4. The carbon steel brush is attached to the Dremel. The stainless steel brush is at the lower left.
The work surface is the bottom of an aluminum baking pan.
Figure 5. Driver wiring harnesses - Cardas on left, original on right.
Figure 6. Tweeter wiring harness.
Figure 7. Binding post and SDA inductor wiring harness.
Figure 8. Cardas wire installed in the right 1.2TL.
Grinding off the clear, invisible insulation from each end of a piece of wire was a very tedious process. Added to that was the need to test for continuity with a multimeter after grinding.
Figure 9. Cable Cooker output signal.
Figure 10. Cable Cooker output signal from original internal wiring.
Figure 11. Cable Cooker output signal from Cardas wire after four days of conditioning.
The conditioned Cardas wire showed less ringing than the original wire.
The right speaker was done first. However, since the results were not favorable, Cardas wire was not installed in the left speaker. With Cardas wire in the right speaker the following sonic results were obtained:
1. The sound stage on the right contracted 4 feet.
2. The right side sound level was apparently lower.
3. Less tactile sensation on the right side.
4. Less image weight on the right side.
5. Less bass definition on the right.
6. Less overall detail on the right.
7. Less tactile sensation on the right.
I was expecting the Cardas wire to either sound better or the same as the original wire, not worse. I don't mean to discourage anyone from using this wire as many people, even members of this forum, have achieved good to excellent results with it. It just didn't work for me in this particular application.
Further Study - Cardas Wire Only On Tweeters
The Cable Cooker accelerates cable break in with a high-power, wide-bandwidth conditioning signal. The rule of thumb is 1 hour on the Cooker is roughly equivalent to 3 hours of music play. Recommended cooking times for interconnects is 36 to 60 hours and 48 to 84 hours for speaker cables. I went 96 hours (4 days) with the Cardas chassis wire because I wanted to be sure of adequate cooking time. In this case, it was impractical to follow my usual regimen of listening-cooking-listening because that would have required multiple tedious iterations of removing and reinstalling the drivers, tweeters, crossovers and binding post plates. The 96 hours of cooking time is roughly equal to 288 hours of music play time.
Originally Posted by codyc1ark
Figures 12-15 illustrate the changes in signal quality after cooking. A Tektronix TDS 2012 oscilloscope was used to take the time domain measurements in figures 12-15 and the frequency domain measurements (Fast Fourier Transform) in figures 16-19. It is easier to see differences in figures 14 and 15, figures 16 and 17 and figures 18 and 19 if they are saved and viewed in succession.
Figure 12. Cable Cooker output signal.
Figure 13. Cable Cooker output signal from Cardas chassis wire at the beginning of cooking.
Cardas chassis wire uses Teflon insulation which is a harsh sounding dielectric, until the molecules become polarized and saturated with energy.
Figure 14. The Cardas wire's ringing and distortion were significantly reduced after 24 hours of cooking. I did
not see any further improvements in the waveform after an additional three days of cooking. This plot is at the
96 hour point.
Figure 15. Cable Cooker output signal from the original SDA SRS 1.2TL internal wiring.
Comparing figures 14 and 15, the Cardas wire shows considerably less ringing on the falling edge of the conditioning signal pulses.
Yes, I have a little more to say.
Originally Posted by codyc1ark
While reviewing the Fast Fourier Transform plots, i realized that the Cardas wire showed lower noise with the 1000 Hz test signal than with the 100 Hz test signal. Another listening evaluation was done with the original tweeter wiring harness replaced with the Cardas wiring harness. Again, it is easier to see differences in figures 16 and 17 and figures 18 and 19 if they are saved and viewed in succession.
Figure 16. FFT plot of 100 Hz test signal from original internal wire.
Figure 17. FFT plot of 100 Hz test signal from Cardas internal wire.
The Cardas wire demonstrated greater noise density and higher amplitude harmonics about 100 Hz than the original wire. The Cardas wire major noise amplitudes were +6 dB at the 50 Hz sub-harmonic, +3.5 dB at the 200 Hz second harmonic and +1 dB at the 300 Hz third harmonic.
Figure 18. FFT plot of 1000 Hz test signal from original internal wire.
Figure 19. FFT plot of 1000 Hz test signal from Cardas internal wire.
The Cardas wire demonstrated much lower noise density and about 1000Hz, and much lower noise density overall, than the original wire. The Cardas wire major noise amplitudes were -0.5 dB at the 200 Hz second harmonic and exactly the same as the original wire at the 300 Hz third harmonic.
With only the tweeter wiring harness replaced with Cardas wire in the right speaker, improvements similar to the "TL" modification for the SDA CRS+ and SDA 2B models were realized (4.1TL Modification-SDACRS+). The improvements were more discernable in stereo mode than in mono mode.
Mono mode with right speaker tweeter wires changed to Cardas:
1. Slight improvement in sustain and decay in stringed instruments (violins, guitar, bass, piano) and percussion instruments (shakers and cymbals) on the right.
2. The right speaker sounded louder.
3. More bass definition on the right.
4. More tactile sensation on the right.
Stereo mode with right speaker tweeter wires changed to Cardas
1. Moderate improvement in sustain and decay in stringed instruments (violins, guitar, bass, piano) and percussion instruments (shakers and cymbals) on the right.
2. More image weight on the right.
3. More bass definition and articulation on the right.
4. More tactile sensation on the right.
Since the right speaker delivered Such Good Sound with the Cardas wire in the tweeter section, I enthusiastically made a Cardas wiring harness for the left speaker.
Stereo mode with both speaker's tweeter wires changed to Cardas:
There were no changes in sound stage dimensions. There was an improvement in the sense of space between sound images (particularly front to back) and significant improvement in detail and clarity of sound images at the sides and rear of the sound stage. The sound stage was more holographic and three-dimensional. In addition to the improvments listed above, with both speaker's tweeter wires replaced, I realized:
1. More overall holography.
2. More detail in the sounds of ambient room reflections in the recording.
Figure 20. The topside of the WBT 1/4" quick disconnect will fit the RD0198 tweeter's narrow tab.
Figure 3 in post #1 shows how the bottom of the WBT 1/4" quick disconnect was adapted to fit the narrow positive tab of the RD0198 tweeter. It turns out that the little extensions at the bottom of the narrow tab are just the right size to fit into the front of the disconnect. I found that the disconnect's topside curved pressure plates needed to be gently bent downward with pliers for a snug fit on the positive and negative tabs of the tweeters and drivers.
The initial trial with the right speaker re-wired with Cardas chassis wire produced disappointing results with an overall loss of detail and detriments to spatial properties. A review of measurement data indicated that the Cardas wire might be better suited for the high frequency section alone. This was verified in listening evaluations. Prior experiments in changing internal speaker wiring provided no audible difference. In this case the results were mixed: detrimental results with a complete replacement using Cardas wire and beneficial results when Cardas wire was only used in the high frequency section.
The Cardas wire was difficult to work with due to the tedium of grinding off the polyurethane insulation, but this would not deter me from using it again in another project.
This experiment was precipitated by the good results achieved from switching to higher quality AudioQuest binding post jumpers (AQ Binding Post Jumpers). Now, I find that I must give some serious consideration to trying a higher quality Litz wire, such as AudioQuest's GO-4. I expect that the G0-4 will be much easier to work with as it is constructed of four Teflon insulated conductors that can be easily striped (2 x 17 AWG + 2 x 20 AWG for an effective AWG of 15).
Further Study: Replacement of Binding Post and SDA Inductor Wiring Harnesses
Replacing the original tweeter wiring harnesses with Cardas wire harnesses resulted in sonic improvements, so I decided to revisit the Cardas wiring harnesses for the SDA inductor and the binding post plate. The Cardas binding post harness was reinstalled in the right speaker and compared to the left in mono and stereo.
1. There was a modest further improvement in bass weight, bass articulation and bass detail.
2. The right side sounded slightly louder.
3. Subtle background percussion sounds and background vocals had more detail
Reinstalling the Cardas wiring harness for the SDA inductor resulted in a further enhancement of depth. Depending on the recording, some sound images moved forward 1 foot in the center and at the sides.
The bad sound I experienced earlier has been isolated to the wiring harness for the stereo and dimensional drivers. I suspect that the length of the wires and the capacitive properties of the wire may be contributing factors. The SDA inductor harness wires and the binding post harness wires were each 24 inches long. The tweeter harness wires were
39" for tweeter 1 (top),
35" for tweeter 2,
31" for tweeter 3,
27" for tweeter 4.
The wires for the driver wiring harness comprise 19 feet 5 inches of wire (115" for positive leads and 118" for negative leads) for the stereo drivers and 19 feet 5 inches of wire (118" for positive leads and 118" for negative leads) for the dimensional drivers. This is a total of 38 feet 10 inches of wire per speaker for the drivers.
I have sent an email to Cardas, along with a wiring diagram showing the lengths of each wire segment, asking if the length of the wire is an issue in this application.
I had some concern that I might not like the Cardas wire because some reviews said it sounds "warm" and "laid back" and I like neutral. The Cardas wire does not sound warm or laid back to me. It just sounds more detailed and clear than the original wire.
AMP Pin Housing Pin Removal
You may want to reuse an AMP pin/socket housing and need to remove the contacts. The pin and socket contacts snap in place with two opposing metal flaps. AMP sells a $57 contact removal tool, but I was able to find a 10 cent option: using a modified socket contact.
Figure 2. AMP socket contact pried open on left, regular socket contact on right.
I took a socket contact, stood it on its rear end and gently tapped a small screwdriver down the seam. The goal is to open it just wide enough so that it fits in the space between the socket to be removed and the housing. It may take some trial and error expanding the socket and then closing it down a bit with pliers.
Figure 3. Push removal socket all the way in.
The AMP nylon housings are translucent, so you can see the retaining flaps and properly position the removal socket to cover them. If your housing is solid colored, like the stock SDA housings, you will have to use trial and error until you luckily cover the flaps.
Using a small pair of needle-nosed pliers, push in the removal socket all the way to the housing. Sometimes, just pushing in the removal socket pushes out the socket or pin. Sometimes, the retaining flaps are dug into the top of the housing sleeve and I need to rotate the removal socket a 1/4 or 1/2 turn, then I can easily remove the pin or socket from the top of the housing with ordinary pliers.
Figure 4. Pin easily pulls out once retaining flaps are held in by removal socket.
Figure 5. All done and housing is ready for reuse.