Vifa XT-25
A ring-radiator tweeter that DIY builders kept reaching for - and kept arguing about its impedance.
Overview
The XT-25 is a tweeter, not a loudspeaker. That matters here because most of what's documented about this driver comes from DIY construction articles that used it as a component, not from any standalone product literature. What the surviving documentation shows is a ring-radiator tweeter that earned a reputation among hobbyist builders for unusually good performance relative to its cost. It also carried at least one specification that experienced users quietly disagreed with.
That disagreement is the most concrete thing in the record. A builder writing in Audio Express (October 2003, p.45) noted that the XT25's impedance looked low enough that he'd call it a 3 Ω tweeter rather than the 4 Ω it was specified as. He flagged it as probably the wrong choice for a single-ended-triode amplifier. The same article built a full Zobel impedance-compensation network for the tweeter before doing anything else with the crossover. The driver has no ferrofluid damping at low frequency, so the impedance rise at resonance is big enough to interact with the filter unless its corner frequency sits several times higher. The published impedance plot for the XT-25 bears this out, and the predicted curve after the Zobel values were selected shows how much correction was needed to get anywhere near flat.
The same builder had originally avoided the XT25 on theoretical grounds, and because of doubts about the published response curve. Testing changed his mind. He reported that the actual tweeters measured well and sounded better still, with the off-axis response, measured at 15°, coming out smoother than the on-axis result. His stated view was that off-axis performance in the 15° to 30° range correlates better with perceived sonic quality than on-axis data does, on the reasoning that off-axis behavior drives the total power response in a listening room. That is the specific reason this driver got recommended.
The XT-25's reputation for value comes up again and again in the documentation. The builder who adopted it chose the Vifa because it offered unusually good performance for the money, with a low resonance frequency and flat, extended response. Cost control mattered in this design. The reference version of the same speaker used a Scan-Speak SS9800 instead. The two drivers were described as having nearly identical impedance curves, similarly flat response, and near-identical efficiency, which made the Scan-Speak essentially a drop-in upgrade. But at nearly $200 apiece, the builder said it raised the stakes considerably. He used the Scan-Speaks in his own systems.
That speaker project went through four revisions, and the XT25 only appears in the last of them. The MkI used a Focal 8V4211 and an MB Quart MBTTR1, both discontinued for DIY availability by the time the design was finished. The MkII abandoned cost constraints entirely for an Eton 8-800 and an Accuton C23-6. The builder was dissatisfied with the cost-to-performance trade-off. The Eton showed a cone mode around 1400 Hz in the impedance curve, and the Accuton didn't measure as smoothly or as extended as its published curves suggested. The MkIII and MkIV shared the same HiVi Research M8a woofer with rear-firing ports and differed in front-panel layout and tweeter. The MkIII used a Focal Tc120dx2, another discontinued part. The MkIV, the version documented in detail, used the Vifa XT25 ring radiator. So it was the cost-conscious pick, chosen after two more expensive tweeters had been tried and set aside, and after one cheaper alternative had gone out of production.
The crossover the XT25 ended up in was only slightly more complex than a similarly implemented Linkwitz-Riley fourth-order network. That's not an unusual configuration for a two-way, but it delivered real benefits in driver power handling and system behavior thanks to the steep slopes achieved in the first 48 dB of attenuation, plus a notch tuned to the low-frequency driver's peak. Component cost in that network was flexible. Surplus GE polypropylene capacitors from Madisound were cited as very good quality and a way to cut expense, as was winding your own inductors. The builder chose off-the-shelf parts instead, on the theory that easy sourcing from multiple vendors would make life simpler for the average constructor.
One practical note from the construction documentation: the XT25 has no ferrofluid damping at low frequency, so the impedance peak at resonance is high. That's why the Zobel network was treated as mandatory rather than optional. Builders who recycle crossover designs across projects were warned that if you plan to reuse a network, compensating the tweeter's impedance is a must. The Zobel board itself was mounted on the rear of the built-up front panel behind the tweeter, with clearance carefully controlled. The documentation advises leaving final port assembly until after the board is mounted and wired if there's any doubt about room inside the cabinet.
What the record does not contain is nearly as notable: no production years, no original retail price for the driver itself, no manufacturer's specification sheet, no independent laboratory measurements from the audio press, and no collector market data. The XT-25's documented life is almost entirely as a component inside other people's projects.
eBay Listings
As an eBay Partner, we earn from qualifying purchases. This helps support our independent vintage technology research.