As most of my high-end speaker projects are still unfinished, my reference speakers are still the T+A T160E. My only mod there was to replace the cheapo caps and coils with better devices, carefully paying attention to series resistance.
Now I had the idea that this would be the quickest way to audition the XT25 in an otherwise familiar setup. The original tweeter is a Vifa D27TG07-06. As far as I can tell from quick-und-dirty measurements, this is the D27TG05 with a different faceplate that is flat on two sides so it can be mounted closer to the two midranges.
Turns out the spacing of the three screws is the same, and so is the diameter of the plastic cover of the VC assembly. The hole in the D27 face plate is maybe 1-2 mm smaller, but also, due to some spacers integrated in the screw holes, the XT25 will be about 2 mm further away from the hole compared to mounting it in its own faceplate. So the smaller face plate cutout is no problem, and the gap is quickly sealed with foam tape. (On a side note: both the original D27 and XT25 assemblies have a cavity between the plastic of the faceplate and the VC assembly, and both originally have a gap between the faceplate and the edge of the dome of 0.5 to 0.8 mm -> doesn't this cause problems?).
The magnet is exactly the same in size, only the connectors stick out a little more on the XT25, so the cutout in the wood needs to be extended in two points with a jigsaw.
Electrically, matching is also quite easy. The D27 has 4.6 R_dc and 92 dB/2.83 V sensitivity. The XT25, in spite of its lower 3 R has about the same voltage sensitivity (91.5 dB). In spite of it having no ferrofluid, the impedance peak is only slightly higher than in the D27, but it occurs at 550 rather than 950 Hz.
The original crossover is standard CLC 3rd order with a -3db point of 3.5 kHz. Impedance compensation is primitive, consisting of just a 10R resistor in parallel to the tweeter. This reduces the impedance peaks of 16 R (for both tweeters) to about 6 R, no matter where they occur. I gather the the slope is high enough so it does not matter whether the residual peak occurs at 950 or 550 Hz. So with nothing changed, I would expect generally 0.5 dB less tweeter sensitivity and a 1-2 dB additional loss between 3 and 8 kHz because of the lower load. If I really wanted to get the same response, I could probably replace the 10R resitor with an RCL circuit tuned to the resonance of the XT25.
Anyway, I replaced the tweeter on one side only, and I cannot detect any changes in tonality or balance. Also, I cannot yet say that the sound is fundamentally different, in spite of all the difference between the tweeters (dome vs. ring, ferro vs. no ferro, standard pole piece vs. completely copper-covered pole piece. I will replace the other tweeter also and listen more carefully...
Eric
Now I had the idea that this would be the quickest way to audition the XT25 in an otherwise familiar setup. The original tweeter is a Vifa D27TG07-06. As far as I can tell from quick-und-dirty measurements, this is the D27TG05 with a different faceplate that is flat on two sides so it can be mounted closer to the two midranges.
Turns out the spacing of the three screws is the same, and so is the diameter of the plastic cover of the VC assembly. The hole in the D27 face plate is maybe 1-2 mm smaller, but also, due to some spacers integrated in the screw holes, the XT25 will be about 2 mm further away from the hole compared to mounting it in its own faceplate. So the smaller face plate cutout is no problem, and the gap is quickly sealed with foam tape. (On a side note: both the original D27 and XT25 assemblies have a cavity between the plastic of the faceplate and the VC assembly, and both originally have a gap between the faceplate and the edge of the dome of 0.5 to 0.8 mm -> doesn't this cause problems?).
The magnet is exactly the same in size, only the connectors stick out a little more on the XT25, so the cutout in the wood needs to be extended in two points with a jigsaw.
Electrically, matching is also quite easy. The D27 has 4.6 R_dc and 92 dB/2.83 V sensitivity. The XT25, in spite of its lower 3 R has about the same voltage sensitivity (91.5 dB). In spite of it having no ferrofluid, the impedance peak is only slightly higher than in the D27, but it occurs at 550 rather than 950 Hz.
The original crossover is standard CLC 3rd order with a -3db point of 3.5 kHz. Impedance compensation is primitive, consisting of just a 10R resistor in parallel to the tweeter. This reduces the impedance peaks of 16 R (for both tweeters) to about 6 R, no matter where they occur. I gather the the slope is high enough so it does not matter whether the residual peak occurs at 950 or 550 Hz. So with nothing changed, I would expect generally 0.5 dB less tweeter sensitivity and a 1-2 dB additional loss between 3 and 8 kHz because of the lower load. If I really wanted to get the same response, I could probably replace the 10R resitor with an RCL circuit tuned to the resonance of the XT25.
Anyway, I replaced the tweeter on one side only, and I cannot detect any changes in tonality or balance. Also, I cannot yet say that the sound is fundamentally different, in spite of all the difference between the tweeters (dome vs. ring, ferro vs. no ferro, standard pole piece vs. completely copper-covered pole piece. I will replace the other tweeter also and listen more carefully...
Eric
Well, I changed the second tweeter also. My first impression was that nothing had changed (no surprise, because the frequency response of both tweeters under 0, 30, 60° is almost identical, and from the XO loading, nothing was to be expected except for 1-2 dB less between 3 and 8 kHz).
On closer listening, it became very clear to me that something had changed fundamentally.
Tonality had not really changed. If anything, highs seemed brighter, which is strange for my simulations predict a slight loss of high frequency level.
Voices and instruments stood out in greater detail and were more realistic/lifelike. I would not say that this is about soundstage or spatial imaging. I perceived the change to be more in the time domain. It was like the way in which each single sound ended was now more natural.
Also, on recordings that I know very well, I started noticing details I hadn't noticed before, like a subtle hi-hat tapping that before the voice and the lead instruments were masking. I must say that I had a similar experience when I first listened to a Sonus Faber Cremona (which uses the XT25) and an AMT-1 design. This was what had originally prompted me to into my ambitous project of building something to replace the T+As (before I had only built smaller projects for friends but considered the T+A level unreachable by home efforts).
The new sound is still a little unfamiliar and strange. This may be because I have spent years and years with the old configuration. I will change the impedance compensation which would give me exactly (to 0.5 dB) the old levels at old frequencies. I suspect that I got used to the halo or trail of artificial sounds that each major sound generated with the old tweeters. Now this is missing, which makes the sounds stand out more clearly (even if several instruments are playing together) but leaves a void.
I wonder what the significant differences between the tweeters are:
- the magnet size is the same
- while the nominal diameter of the XT25 is bigger, I suspect that the actually radiating areas are similar in both diameter and surface area
- the XT25 has a heavily copper-plated pole piece which should significantly reduce intermodulation distortion and time-smearing effects like domain-jumping in the iron and the ferrite magnet - if this is so important, the DX25 should already have an edge on other Vifa tweeters - on the other hand, the ring radiator concept seems to generate substantial 2nd harmonic (why?)
- the XT25 has no ferrofluid, but I suspect time smearing due to currents in the fluid is highly unlikely
- the diaphragm construction is different
On closer listening, it became very clear to me that something had changed fundamentally.
Tonality had not really changed. If anything, highs seemed brighter, which is strange for my simulations predict a slight loss of high frequency level.
Voices and instruments stood out in greater detail and were more realistic/lifelike. I would not say that this is about soundstage or spatial imaging. I perceived the change to be more in the time domain. It was like the way in which each single sound ended was now more natural.
Also, on recordings that I know very well, I started noticing details I hadn't noticed before, like a subtle hi-hat tapping that before the voice and the lead instruments were masking. I must say that I had a similar experience when I first listened to a Sonus Faber Cremona (which uses the XT25) and an AMT-1 design. This was what had originally prompted me to into my ambitous project of building something to replace the T+As (before I had only built smaller projects for friends but considered the T+A level unreachable by home efforts).
The new sound is still a little unfamiliar and strange. This may be because I have spent years and years with the old configuration. I will change the impedance compensation which would give me exactly (to 0.5 dB) the old levels at old frequencies. I suspect that I got used to the halo or trail of artificial sounds that each major sound generated with the old tweeters. Now this is missing, which makes the sounds stand out more clearly (even if several instruments are playing together) but leaves a void.
I wonder what the significant differences between the tweeters are:
- the magnet size is the same
- while the nominal diameter of the XT25 is bigger, I suspect that the actually radiating areas are similar in both diameter and surface area
- the XT25 has a heavily copper-plated pole piece which should significantly reduce intermodulation distortion and time-smearing effects like domain-jumping in the iron and the ferrite magnet - if this is so important, the DX25 should already have an edge on other Vifa tweeters - on the other hand, the ring radiator concept seems to generate substantial 2nd harmonic (why?)
- the XT25 has no ferrofluid, but I suspect time smearing due to currents in the fluid is highly unlikely
- the diaphragm construction is different
capslok, thanks for the detailed review.
I have one question, as I have very limited knowledge of electronics, I cannot work this out for myself.
I have a pair of TLb speakers form t-linespeakers.org
They use the vifa D27TG-45-06 in a push-push 2 tweeter configuration for each speaker, 2 Vifa P13WH-00-08 also in a push push configuation. The crossover is a simple 10uf capacitor on the tweeter only. All drivers wired in parallel.
I would like to change the front tweeter only to the Vifa X25.
I would like some help with the crossover, as I have no idea on the changes I may need to make. I would like to keep it as simple as possible.
Thanks Oz
PS. If anyone can help me, I would be greatfull.
I have one question, as I have very limited knowledge of electronics, I cannot work this out for myself.
I have a pair of TLb speakers form t-linespeakers.org
They use the vifa D27TG-45-06 in a push-push 2 tweeter configuration for each speaker, 2 Vifa P13WH-00-08 also in a push push configuation. The crossover is a simple 10uf capacitor on the tweeter only. All drivers wired in parallel.
I would like to change the front tweeter only to the Vifa X25.
I would like some help with the crossover, as I have no idea on the changes I may need to make. I would like to keep it as simple as possible.
Thanks Oz
PS. If anyone can help me, I would be greatfull.
Capslock: curious if you're still using the XT25 tweeters? Long term impressions?
Oh, and thank you for your detailed subjective comparison between the two tweeters.
Oh, and thank you for your detailed subjective comparison between the two tweeters.
- Status
- Not open for further replies.