Since our Tandem rack review, at 2011, the truth is that not many things have changed towards the laboratory assessment of similar systems. We don't have a widely adopted methodology associated with the needs and characteristics of an audio device support system. During this lab evaluation, we tried to find some data on how the Statement and Ultimate worked, and if they perform the way Tandem says they should, aiming to form an objective basis on which to support some conclusions. Measurements were made with an acceleration meter which was placed at two points, directly on the stand surface and on the top of the chassis of a power amplifier. The amplifier used was our standard reference amp (Parasound HCA3500) at its normal position in our listening room, i.e. where it is always during our listening sessions. This is important in the case where an excitation through the loudspeakers is included during the measurement.
The two stands replaced the one that is usually used for the support of the amplifier (a simple wooden shelf with castor wheels from Exelixis). The whole procedure included both measurements with the amplifier driving the loudspeakers at a normal reference level (73dBSPL/2.8m with a pink noise signal at -20dBFS level) and with the amplifier idling, without any input signal.
Before one proceeds to study the relevant graphs, however, some time should be spent to gain a sense of the vertical scale used (which is the same as that used in the loudspeaker cabinet coloration measurements). The unit supported by the measurement tool is the somewhat unusual dBm/S² (decibels of acceleration) and relates the sensitivity of the accelerometer with its output at each measurement frequency and the system reference level. The sensitivity of the accelerometer is 1mV/m/S² and the system reference level is 1V, so if one does the necessary math, he should find that the -120dBm/S² corresponds to an acceleration of about 0.001m/S² or 0.0001g and that the -30dBm/ S² corresponds to 31.6m/S² or 3.1g.
The first set of measurements was conducted on the stand surface, after both stands were set-up properly and without the amplifier mass loading them. The relevant graphs show that the most "uncomfortable" (i.e. noisy) region is defined by an octave either side of 50Hz, a finding that raises some suspicion of a leakage of electrical noise in the measurement system. However, measuring the output of the accelerometer "in the air" (standing completely still and without contacting any surface at all) showed that the noise is not electrical but comes -indeed- from the stand surface and its contact with the floor. Here, one can observe that the Statement and Ultimate are consistently better than our reference in the entire range of the measurement, i.e. they carry less floor vibrations to the device through their surface. In addition, the Ultimate is more efficient than the Statement, not unsurprisingly since it is the top and more sophisticated and expensive model.
The same measurement with the amplifier mass loading the stand but turned off, shows what one would normally expect, based on intuition: The stands show much smaller acceleration under the heavy weight of the power amplifier, sufficient to minimize the environmental noise regardless on what stand is used. At very low frequencies, however, it can be seen that the Ultimate remains the stand with the lower acceleration figures.

When the amplifier is turned on but without any signal feeding its inputs, (so its power supply is now a mechanical noise source) the situation has not changed substantially in general (mind that the vertical resolution in this diagram is larger): Towards the very low frequency part of the spectrum, (10-35Hz), the Ultimate shows the lowest acceleration, but a new finding makes its appearance: Strong cancellations at 100Hz and 200Hz, with both the Statement and the Ultimate to be significantly better than t he reference in the first case and the Statement to excel in the second. One possible explanation for this behavior might be the effect of the amplifier support points on the behavior of the whole system as they are placed, in effect, between a source of vibration (the power supply transformer) and the accelerometer.
This explanation is, probably, supported by the measurement of the acceleration in the amplifier chassis: This behavior in this case is also different, both in terms of the frequency at which the cancellations occur (the lower frequency is now below 100Hz) and in terms of the their level.
What conclusion can one draw of these findings? A possible one is that in addition to the design of the stand, the final result depends also on the amplifier support points properties which are in general unknown variables of the system and in fact different each time, depending on the device they are attached on. Perhaps the next step here is the removal of the standard support points and their replacement by a part with known properties to ensure optimum interfacing with the given stand.

Measurements with the amplifier driving the loudspeakers, as occurs during the normal operation of a system, offers an opportunity to check the stand behavior under "full" load, meaning that there are two types of vibrations to be managed. The inbound, from the loudspeakers and the environment and the outbound from the amplifier itself. Furthermore, regarding the inbound vibration, there are two transmission paths, through the air and through the mechanical coupling between the loudspeaker, its stand, the floor and the amplifier stand under test, so the analysis of the phenomenon becomes even more complex.
The acceleration measurement at the stand surface, shows that the main design goal as described by Tandem has achieved: Both the Statement and the Ultimate exhibit smaller acceleration than the reference stand, in almost all of the range of measurement, a behavior apparent particularly in the 20Hz-35Hz and 150Hz-800Hz parts of the spectrum. According to Michelios, the stand inertness is the main criterion of its good behavior, an evidence that it does what it was designed for, i.e to transfer any vibration to the floor as fast as possible. This behavior is more clearly showed when one compares the acceleration of the amp chassis to the acceleration of the shelf surface (blue and violet curves, respectively, in the second graph set at the bottom of this page).
The measurement of acceleration in the chassis of the amplifier shows a more complex behavior: The stand/amplifier system both in the Statement and the Ultimate cases, seemed particularly capable to attenuate very low frequencies (20-30Hz), but it appears less effective in areas where the chassis shows its more vibration-prone character, as it is evident by the peaks around 100Hz and 300Hz. Instead, it has a clear effect on the region between 100Hz-200Hz and above the 500Hz, up to the limit of the measurement.
The bottom line here is that the stands appear to be functioning the way the company promises (transferring vibrations as fast as possible compared to a simple stand, given Tandem's criteria) and that this behavior translates to a corresponding reduction of vibrations in the amplifier chassis in specific frequency ranges, probably in dependence with the amplifier's supporting points.

Previous | Next