Systems List › DIY › Dayton UM18-22 Sealed
NotesThe Dayton Audio UM18-22 driver, loaded into the standard 4.2 cubic foot sealed test box used to measure 18” drivers, results in an impedance that starts at about 4.2 ohms near DC and then rises towards 10Hz where it reaches about 5.1 ohms. The system resonance is at about 34Hz where the impedance rises to 25 ohms. The minimum impedance above the system resonance is 5.9 ohms at 75Hz where the impedance starts slowly rising again towards 200Hz. There is a minor blip in the impedance at 175Hz which may be related to the enclosure and a major issue at 500hz which is driver related. The moderate angle of impedance rise above the minimum at 75Hz indicates that the inductance on this driver is managed well.
The basic on axis response shape and sensitivity of the system, as measured in a ground-plane setting, shows that the sensitivity of the UM18 driver in a sealed box of this size is around 89dB over the 35-125Hz range when stimulated with a 2 volt signal. The response peaks at a hair over 90dB at 45Hz and drops about 2dB down by 100Hz. The top end does not exhibit the sharp roll off of the upper bass frequencies commonly associated with long throw subwoofer featuring heavy voice coils. This is due no doubt to the multiple Faraday rings inside of the motor. However there is a very large rise in the response near 300-500Hz, peaking at 420Hz and then falling off sharply, corresponding with the 500Hz spike in the impedance measurement, so this driver is best used with a steep low pass filter to keep its reproduction limited to the subwoofer region. The time domain measurements show nothing worth worrying about below 200Hz with a clean group delay well under 1 cycle and no significant ringing. Of note is that the 420Hz response peak does show up as latent energy. This driver is best used under 150Hz.
The long term output compression sweeps start at a drive level of 4 volts, which produces 50Hz at roughly 90dB at 2 meters for the base measurement. The UM18 in the sealed test enclosure responds in a linear and relaxed manner to the next couple of increases in voltage corresponding with a 5dB increase in output demands from the speaker. Finally after a 20dB increase in level to a 40 volt drive level which should produce around 110dB at 50Hz, the UM18 starts to exhibit about 1dB of output compression in the upper bass bandwidth. After another full 5dB increase in output demands, requiring a drive signal of 71 volts, The UM18 finally starts to significantly compress the output by about 1.5-2dB. At this level there was some amount of excursion noise and distortion from the driver, so this testing was stopped after a further increase of only 2dB to the drive signal. At +27dB over the base 4 volt measurement, the drive signal had now ballooned to about 90 volts. If the system was perfectly linear and uncompressed the output should’ve been around 117dB at 50Hz by this point but it can be seen on the output compression magnitude graph that the output was now being compressed by about 2 to 2.5dB broadband. The driver had also made some overload noises and high levels of distortion during this measurement so it was wise not to push things any further. The repeat 4 volt base line measurement conducted after the 90 volt measurement did not show much change in the base line sensitivity, perhaps a loss of about 0.5dB broadband, which is pretty good. This indicates that the coil and motor has held up to this amount of input signal without heating up excessively. During the 90 volt measurement the UM18 produced about 115dB from 30-120hz and just over 106db at 20hz.
The distortion results are quite good. Once again owing partially to the multiple shorting rings in the motor no doubt and also helped by the extended xmax of this driver. At a 22.5 volt drive level the THD is held below 10% above 17Hz and well under 20% even at 10Hz. Above 30Hz THD is generally 2% or less. At a 40 volt drive level THD is still held below 10% above 22Hz and under 20% down to the 10Hz area indicating that the drive r is still quite linear at this level. Above 30Hz THD is still around 3% or under. Another 5dB increase in signal to 71 volts and we can see that the driver excursion is now getting outside of its linear envelope in the deepest bass as the THD is now in the range of 30-45% below 20Hz. Above 32Hz THD is held below 4% still. The measurement with about 90 volts applied indicates that the driver is out of excursion in the deep bass at this level. THD reaches almost 75% and around 40% at 20Hz. Above 32HZ THD is still well below 5% even at this level which is impressive. Looking at the harmonic makeup of the 90 volt measurement shows that the 3rd harmonic is the primary distortion in the deep bass where excursion is high. Above 30Hz where excursion demands are reduced the primary distortion is the second harmonic which is fairly benign subjectively. The 3rd harmonic does briefly creep back into the picture at 150Hz. All other harmonics are well down in level.
The CEA-2010 short term, distortion limited, burst output measurements once again confirm that the UM18-22 driver does indeed have a lot of linear excursion. It produces a passing result at 10hz of about 90dB with a 56 volt input and over 94dB at the 12.5Hz band with a 69 volt input. Neither bandwidth is an easy pass for a driver in a sealed box of this size. At 20Hz the distortion drops notably and the UM18-22 comes up just shy of 105dB which is excellent for a driver of this size and cost. Moving up to the 31.5Hz band the output continues to increase and distortion continues to drop as the system is now operating at its system resonance, so the power input is very low. The UM18-22 was able to grunt out just shy of 118dB at 31.5Hz with passing distortion. At 50Hz and above the driver is able to produce short term levels of 123-126dB with the K20 amplifier being pushed into clipping (Calculated voltage applied is north of 400 volts) . In the deepest bass, where excursion is high, the output is limited by 3rd harmonic distortion. At the 31.5 and 40Hz bands the system was limited by driver excursion rather than distortion. At 50hz and above the amplifier was the limiting factor for the recorded SPL.