Kef 104ab with XT25 - looking for info/advice with crossovers

[Mike-Toronto]

Hello again forum
I am pretty new around here, and with speaker DIY so sorry if I am missing some of the obvious.

I am rebuilding/modifying my long loved Kef 104's. They served me well for about ten years and now have dead tweeters and what I fear to be a dying crossover. I decided to put in a pair of XT25's to replace the T27's.
I do not have any serious test equipment setup here. Just a microphone and the ability to meter test tone volumes at any given frequency. According to very basic sweeps the relationship between the two drivers is good. Regardless I can hear the transition. It sounds as if the HPF on the tweeter needs either to be lower or have a much more gentle filter.

I would get to work on a new crossover design but I am missing information on the drivers. I have do not have a impedance graph for the B200 and the other main thing is that the XT25 specs are incomplete on there site. There is no voice coil inductance value at all. Nor wattage for that matter.

Is the only way to get decent results in this situation to measure the drivers myself? I have taken a quick look at speaker workshop but have not gotten my head around it yet. Is there an easier freeware app that has the ability to get the basic values?

Sorry for the rambling post. The is what happens in my head after a week of google. Any assistance or advice would be greatly appreciated.

[AndrewT]

find out which b200 is fitted to your 104s.
search this Forum. The T/S data is available for about 4 or 5types of B200.

[Mike-Toronto]

Sorry I thought I had posted it what version the B200 was. It is the SP1039. As far as I know I have found all the data that can be found on this driver. Via these two documents. Even with this data it seems that 75% of what Speaker Workshop is looking for is missing. I current do not have the needed equipment to test the drivers myself so unfortunately that is not in the picture.

Thanks to www.hifiloudspakers.info and www.t-linespeakers.org/ for this fantastic documentation.



I am beginning to feel that it may be best to simply keep the B200 SP1039 with the current crossover and build the tweeter around it. I was hoping to get the crossover lower than the current 3k. I would be very happy to get down to around 2.5k - 2k.

Can anyone tell me by looking at that crossover how I might do that? I have just started into this area and quite simply are in over my head. I have been running things through "crossover calculators" on line and trying to make sence of it. It does not help that this crossover has three positions for the HF filter.

In regards to the XT25 I feel that I have moved a little bit further along by doing more searching around the web. I found this crossover on the zaph audio site.
http://www.zaphaudio.com/ZD5.html At the very least I can take the Zobel network from that and apply my own crossover frequency. I think that is accurate enough for me at this point.

I wish I had the same information for the B200 SP1039 so I could do my best to flatten it out with a Zobel if need be and apply a simple LPF to it. From there try and sort out phase by tweeter placement. Well thats all ive got for today.

[Inductor]

First of all your substitute Vifa XT25 has a Rvc of 3.0R and your original Kef T27 is 6.25/7.1R@2.3K. As you see you have to completely change your xover. For some more closed equiv.:
REPLACEMENT UNITS->KEF 104/AB wilmslow-audio.co.uk
https://secure.wilmslow-audio.co.uk/...9f0f470201e189
KEF T27 SP1032
Isophon KK8, KEF T27

[Mike-Toronto]

I did see that page before I bought the XT25's. I was feeling bold at the time and was willing to dig into a bit of a project. I am now wondering if I have made the right decision. If I can get the specs that Solen need then I will most likely just get them to build it.

[Inductor]

You can test this layout having in mind that it's a simulation and that you might have to adjust. Also your tweeter/speaker is working on 4 ohms for high frequencies.

[Inductor]

xover

[Iain McNeill]

I'd say the biggest problem is the sensitivity difference T27 at 80dB vs XT25 at 90+. Not sure if the KEF noise method is equivalent but 10dB is a lot.The impedance mismatch will shift the XO frequency.

I'm currently using XT25 ring radiators (not Tymphany) and I like how they sound. Excellent on-axis, smooth off-axis roll off, low distortion. With their low F0 I'd say they were an excellent replacement - just need to fix the XO.

Inductor has a good solution but I'd break it into 2 parts. Start with the level correction which is the 2.2ohm and 3.3ohm resistor (in Inductors schematic) divider right next to the tweeter.

You can make it adjustable by using an 8ohm L-pad as a potentiometer instead of the resistors. Play with the balance between the highs and lows but try to keep the L and R values similar. Some kind of calibrated scale would be useful especially when adjusting levels between XO circuit A/B testing.

You may find that balancing the levels is good enough.


for a while!!!

[Inductor]

Try this one, as the impedance is about 5/6 ohm with high frequencies.

If you want to lighten up (from 10k and up) things a little bit (your tweeter) from a relaxed configuration give a cap in parallel of 0.5uF/1uF to the 6R8 resistor.

[alan-1-b]

Mike,

where are you up to with your tweeter substitution ?

If you want a simple modification to try first so that you can listen to how the Vifa tweeter sounds with the old B200, then try this:-

Connect a 3.9 ohm, minimum 5 watt, preferably non-inductive, resistor in direct electrical Series with the tweeter
- between the tweeter and the old cross-over.
A 10 watt rated resistor would be better if you intend to play the speakers loudly.
4.7 ohms is close enough if you cannot easily get 3.9 to try.

Remove the 0.6uF capacitor, and the 3 ohm resistor that is in Parallel with the second 10uF capacitor, but leave that 10uF cap in circuit.

Listen to that, then next remove the 10 ohm resistor that is in Parallel with the 0.6mH inductor in the bass circuit.
Listen again, and decide whether you prefer with or with-out that 10 ohm resistor.

If you are happy with all this, then next you can substitute a new 5.1uF or 4.7 uF cap in place of the old 10uF's pair,
and a new 3.9uF, or 4uF, in place of the old 3.3uF.

But, if you are not happy do not buy the new capacitors
-and instead, try Zaph's Vifa option, or whatever else you can find.

I am NOT recommending you send the problem to Solen, because a Simulator produced circuit may not produce what you will actually like the sound from in practice.
Implementing cross-overs for good audible effect in loudspeakers is more complicated than simulators can achieve.

Also, as it has 3rd Order cross-over filter slopes for both drivers, it can be connected in either tweeter Polarity.
There will be audible differences, and you may prefer one option to the other,
thus listen with the tweeter in the Reverse Polarity, as in the original, and then swap the tweeter connections + and - around and listen with the tweeter thus in the same Polarity as the woofer.

Listen more-so on the woofer axis vertically, than on the tweeter axis,
as that will give closer to equal signal arrival times from both drivers,
and thus a more coherent sound.

Expect some of the other members of this Forum to contest some, or all, the suggestions I have made, however I recommend you listen and decide for yourself, and then you will at least understand more about cross-over behavior.