Peerless BC25SC08 vs XT25SC90

Peerless BC25SC08 vs XT25SC90 These two tweeters from Peerless are quite interesting. Though both are around $20, don’t let the low price fool you. They sound excellent if used properly. Common features between the two are the use of neodymium magnets and 25mm voice coils. Apart from that, their designs differ greatly. The BC25 model uses a small waveguide and the voice coil is immersed in ferrofluid. It even has tiny fins for cooling. The XT25, on the other hand, uses Peerless patented ring radiator with a pointed central waveguide. No ferrofluid but has a rear chamber to extend the low frequencies. How do they match up?

The Frequency Response Plots above are from Peerless. On the left is the BC25. The right is the XT25. The most telling difference is the impedance peak. The BC25 is damped whereas the XT25 is rather high and sharp. As for response flatness, it appears the BC25 wins. But that can be deceiving because the scaling for the two responses don’t seem to be the same. Let’s see how they measure when mounted onto a box.
(Fig 1) BC25SC08 Response (No Smoothing, mic at 0.5 meters on axis, 5 msec gating)	BC25SC90 Frequency Response in Super-Micro In Fig 1, the RAW response (Black Plot) records a kink at 5kHz. Frequencies below that are lifted. This is the effect of a 5″ wide baffle. The Red Plot is with a 12dB filter at 4kHz. The hump now changes to a plateau. It is an improvement but the asymptote is wrong. The Blue Plot is with a 18dB/oct filter at 5kHz. Now, the roll-off looks much better. This network includes an EQ to flatten the top (from 8kHz to 15kHz).
(Fig 2) XT25SC90 Response (No Smoothing, mic at 0.5 meters on axis, 5 msec gating)	XT25SC90 Frequency Response in Super-Micro The kink now appears at 4kHz. Frequencies lower than 4kHz are also lifted but they are not as abrupt as the BC25.Unlike the BC25 which has a plateau, the XT25 roll off (Red Plot) at 12dB/oct (fc=4kHz) is more reflective of a crossover slope.
Does it mean the XT25 is superior to the BC25? Not really. Other factors are involved. The XT25, for example, has a much higher impedance peak than the BC25. If that peak is not treated, it can induce distortion. Trouble is, adding an LCR to damp the peak increases the network cost. Might as well use a more expensive tweeter that doesn’t need an LCR. Having worked with the BC25SC08 and XT25SC90, I would reserve them for situations where I want to cross at 3,500Hz of higher. The filter network is easier and cheaper this way.