Tests and Analysis by David L. Ralph
The 7100 is an interesting, but in my opinion, over-priced tweeter. The 6600 is a better choice. My thanks to Mark K. for loaning this unit for testing. As done for the 6600, test results, analysis and a couple of tweaks are presented. The latter help to correct some frequency response issues.
The initial quasi-anechoic response made on a 2mx2m baffle shows a somewhat ragged response on-axis. However, the impedance shows the flat impedance in the upper range that is indicative of a good, low-distortion motor.
The off-axis response maintains a bit of the raggedness, though it does improve with increasing angle. As will be seen a bit farther down, much, though not all, of this is due to faceplate geometry. Part of it is from internal reflections.
The cumulative spectral decay (CSD) or waterfall plot shows what one would expect from seeing the frequency response, a ridge of energy and a dip as well that is a consequence of a reflection.
The CSD is not overly useful, especially for design. It is simply another way to examine the SPL response. I find it to be mostly a confirmation of what is generally obvious from examining the SPL response. LAUD does make one display of the CSD useful and that is in cancelling the display before the full CSD is created. This displays the SPL response as it would be if the start time marker were moved forward by a corresponding amount of time. It's most easily done on a slow computer due to the slow update during the CSD creation. The next set of graphs were generated from the same impulse response as the original SPL response. The CSD was cancelled at successively longer times in the impulse response. The impact of the reflection causing the severe dip is make more apparent.
A second reflection appears to be developing around 2K or below, but no ridge actually shows up in the full CSD. The benign nature of the dips (somewhat distributed in frequency) are apparent due to the lack of any serious anomolies in the frequency response. They could be nothing more than anomolies due to the measurement, since this is getting into the gray area of limited resolution of the MLS measurement.
As always, once I had all of the baseline measurements made, I opened up the driver to check out the internals more closely. The photos below show the motor from several perspectives.
The first one shows the driver from the side without the faceplate, exposing the front of the diaphragm assembly. The second one is a closeup of the motor pole-piece vent and gap area.
Next up are photos of the felt dome over the pole-piece vent. It's supported from the center by a piece in the vent opening rather than by silicone dabs at three points around the edge as is the manner used for the 9300/9500 series drivers. This method looks to be better and consistent in positioning. The silicone method sometimes left the gap under the dome too small for good air flow into the vent.
Unfortunately I wasn't able to remove the rear cover "chamber" area without fear of damaging the driver.
Like the 6600, the basic diaphragm assembly for the 7100 is not new, having been introduced in the Vifa DX25. The dome material and doping compound are different from both the 6600 and the DX25.
The diaphragm has a similar treatment as those for the DX25 and 6600. Each one has differences, however. The dome material as well as the treatment appears to different for each one. There are some common characteristics, though. Each one has a form of tip area treatment that I suspect is aimed at damping the signal at higher frequencies. This can be seen as a ring on the dome, at a different position and thickness for each driver (DX25, 6600 and 7100). The 6600 seems to have the best result of the three.
Each of the drivers that share the same diaphragm mold, DX25, 6600 and 7100, have different faceplates. This is more influence than one might think, keeping in mind that the examples are on-axis only where the impact is strongest. I had all on-hand so I tested the response with each one. The faceplate from the DX25/XT25 seems to be better shaped with the results shown below.
The diaphragm assemblies for the DX25, 6600 and 7100 are all interchangeable from what I can tell, so I tested a DX25 dome I had on-hand in the 7100 motor, results below. Which one might be considered better is a bit of a toss-up. Which faceplate is better is not in my opinion. The DX25/XT25 is the winner. The best one happens to be that of the 6600 (not shown).
I've not had a pair of these for auditioning in a system, so I can't comment directly on the sound. What bothers me is the anomolies in the upper frequency response that might result in some linear distortion. Reports I've read, though, haven't shown this to be a problem.