what will be the best crossover point for this tweeter? in 2 way with 5.5 inch midbass driver?
its resonant freq is decent enough as Fs=589Hz so considering its lower Fs can it be crossed low? like 2Khz?
its resonant freq is decent enough as Fs=589Hz so considering its lower Fs can it be crossed low? like 2Khz?
Hi, mimimum 2KHz LR4 acoustic, 2.5KHz LR2 acoustic according to Zaph, rgds, sreten.
Even though Fs is low, the driver is nothing special at all below 2KHz distortion wise.
Even though Fs is low, the driver is nothing special at all below 2KHz distortion wise.
Agreed. Be advised you may also have to damp the Fs spike a bit so it doesn't sound nasal or buzzy.
Later,
Wolf
Later,
Wolf
This tweeter is used here on pages 5,6,7 in this design with 5.25 peerless poly or drop one compensation filter and use nomex(discontinued but still can be found)
http://www.diyaudio.com/forums/multi-way/97043-elsinore-project-thread-6.html
Tweeter is on bottom below listening level...
I made several pairs both poly and nomex.. Nomex is lighter and has more speed and bass has better pitch definition and more resolution overall..
More details near the end of the thread
Joel
http://www.diyaudio.com/forums/multi-way/97043-elsinore-project-thread-6.html
Tweeter is on bottom below listening level...
I made several pairs both poly and nomex.. Nomex is lighter and has more speed and bass has better pitch definition and more resolution overall..
More details near the end of the thread
Joel
That's a blanket statement. So if it was an electrically 4th order cross over around 3khz it would require Fs comp? I only think it would need that if the cross over is low order.
I used this tweeter in an MTM for a competition where all the contestants had to use the same woofer and the listeners were blind. I used a second order electrical crossover for an acoustic 2250hz LR4. It won the competition.
People often miss what zaph is saying. He's saying because he only has a cap and an LR filter, the filter won't be effective at Fs. You would be able to see this in the transfer function. So what happens in his speaker, is the output of the tweeter isn't very attenuated at Fs unless he clips that impedance peak. If its not attenuated enough, the tweeter makes a lot of distortion.
Bottom line. How low ANY tweeter can cross depends on:
Desired output.
Filter/transfer function.
Tweeter performance.
I can make a $5 some tweeter play fine down to 1500hz. I won't even put a filter on it. But the output will be limited to 70db 😉 So consider all the design criteria and then you'll have your answer.
I used this tweeter in an MTM for a competition where all the contestants had to use the same woofer and the listeners were blind. I used a second order electrical crossover for an acoustic 2250hz LR4. It won the competition.
People often miss what zaph is saying. He's saying because he only has a cap and an LR filter, the filter won't be effective at Fs. You would be able to see this in the transfer function. So what happens in his speaker, is the output of the tweeter isn't very attenuated at Fs unless he clips that impedance peak. If its not attenuated enough, the tweeter makes a lot of distortion.
Bottom line. How low ANY tweeter can cross depends on:
Desired output.
Filter/transfer function.
Tweeter performance.
I can make a $5 some tweeter play fine down to 1500hz. I won't even put a filter on it. But the output will be limited to 70db 😉 So consider all the design criteria and then you'll have your answer.
No one's mentioned the use of waveguides which allow lower xover points by raising tweeter's efficiency.
Visit PETT dpellegrini's waveguide threads for more info.
Visit PETT dpellegrini's waveguide threads for more info.
XT25 got some troubles in sub-4K region. Not that bad, yet...
For high-end no-compromise thing - 4k is fine.
For compromised design - 2.5k.
Waveguide helps here to tame the distortion rise under 4k - you pump-up the efficiency by the waveguide in the 1-4k range, and then flatten it out by filter = less cone movement, less HD. This way it could go down to 1k 🙂
I've used them almost fullrange (low volume) and they were fine from 250Hz. They have some tiny subwoofers inside 🙂
For high-end no-compromise thing - 4k is fine.
For compromised design - 2.5k.
Waveguide helps here to tame the distortion rise under 4k - you pump-up the efficiency by the waveguide in the 1-4k range, and then flatten it out by filter = less cone movement, less HD. This way it could go down to 1k 🙂
I've used them almost fullrange (low volume) and they were fine from 250Hz. They have some tiny subwoofers inside 🙂
Just thinking aloud here, really. We had someone a while back who really rated the double-magnet version of the Vifa XT25. 🙂
FWIW, Troels Gravesen never crosses Vifa tweeters very low. Around 3.5kHz. It's actually a bit better behaved than the regular XT25TG30-04 version which Troels uses in the Paper woofer Ellam-XT and Polycone woofer VifaPL14WJ+XT25TG designs. But is 2dB more efficient, which is fairly easy to allow for I'd reckon.
Lack of ferrofluid DOES make for a slightly alarming impedance bump at Fs and the low impedance of 3 ohms DC needs correction. The other fix is of course a shunt resistor around 10 ohms to flatten impedance too. Hope it helps anyway.
FWIW, Troels Gravesen never crosses Vifa tweeters very low. Around 3.5kHz. It's actually a bit better behaved than the regular XT25TG30-04 version which Troels uses in the Paper woofer Ellam-XT and Polycone woofer VifaPL14WJ+XT25TG designs. But is 2dB more efficient, which is fairly easy to allow for I'd reckon.
Lack of ferrofluid DOES make for a slightly alarming impedance bump at Fs and the low impedance of 3 ohms DC needs correction. The other fix is of course a shunt resistor around 10 ohms to flatten impedance too. Hope it helps anyway.
Attachments
I'm planning a speaker at the moment using the XT25TG in a waveguide (I already have them but later on I may end up buying the BG for the extra efficiency).
The challenge comes if you are wanting to use one of these tweeters with a passive crossover, you may end up with an expensive set of components for the crossover.
I haven't tried it, yet, but I believe that at least some of the problems that others have had with these tweeters is because they are (understandably) unwilling to build (and spend time designing) an expensive and complex crossover when they could simply buy a better tweeter!
Since I'm using active crossover and already have the tweeters from an earlier abandoned speaker build I will probably use these for my next speaker project. The idea I have at the moment is to use two notches, one at driver resonance and another broader notch to get the response flat after horn loading and push the distortion lower between 1k and ~5k
I've found 3 sets of data from people who've been down this path (though to my knowledge all using passive crossovers) before:
Some measurements that Zaph did on the XT19 - Zaph|Audio
xt25tg in monacor wg300 - Wave guides & horns and
Troels' experiments here with a few SS tweeters including the xt25tg - ScanSpeak-Waveguides
The challenge comes if you are wanting to use one of these tweeters with a passive crossover, you may end up with an expensive set of components for the crossover.
I haven't tried it, yet, but I believe that at least some of the problems that others have had with these tweeters is because they are (understandably) unwilling to build (and spend time designing) an expensive and complex crossover when they could simply buy a better tweeter!
Since I'm using active crossover and already have the tweeters from an earlier abandoned speaker build I will probably use these for my next speaker project. The idea I have at the moment is to use two notches, one at driver resonance and another broader notch to get the response flat after horn loading and push the distortion lower between 1k and ~5k
I've found 3 sets of data from people who've been down this path (though to my knowledge all using passive crossovers) before:
Some measurements that Zaph did on the XT19 - Zaph|Audio
xt25tg in monacor wg300 - Wave guides & horns and
Troels' experiments here with a few SS tweeters including the xt25tg - ScanSpeak-Waveguides
Almost every statements are blanket statements (same with "very high" frequency). What required is for the reader to read between the lines. And nobody is extending what Zaph is saying. As was clear from the "ime".
The tweeter is very different than regular tweeters. That is the basic of the blanket statement. I don't know the theory behind the double coils, but it seems the bigger coil, especially without ferro-fluid, behaves differently.
I've forgotten about the tweeter (been playing with amps lately). But I'm curious how you could achieve fourth order at 2k25 with only 2nd order electrical filter? Was that a standard approach? (I'm not using my own computer, so no software and data). I roughly guess you have to create some kind of bainter notch, but at critical frequency that you should better avoid?
For me the criteria is simply the best sound. It means, very unlikely I don't need the notch at Fs (even if software based analysis doesn't agree) 🙂
Jay, your reaction to my post may be because I wasn't clear that only the very first sentence of my post was loosely attributed to your post. The rest was for the OP and not directed at you.
You're right about my LR4 cross over. It was likely more like 3rd order below the cross over knee. The impedance is moving towards a trough so the filter becomes stronger towards the trough. Then weaker moving towards Fs. My knee was approaching that trough iirc. But I didn't use any notches. I didn't have a problem. But there were budget restrictions on the competition. I listened hard and long and eventually decided the LCR wasn't required, as tempted as I was only cause of all the buzz around this tweeter needing one.
But yes, in your experience an LCR was required, as it was in many peoples experience. I'm not discounting that. Just wanted to extend some other thoughts. As system said, zaph also mentions the use of a shunt resistor that can work.
You're right about my LR4 cross over. It was likely more like 3rd order below the cross over knee. The impedance is moving towards a trough so the filter becomes stronger towards the trough. Then weaker moving towards Fs. My knee was approaching that trough iirc. But I didn't use any notches. I didn't have a problem. But there were budget restrictions on the competition. I listened hard and long and eventually decided the LCR wasn't required, as tempted as I was only cause of all the buzz around this tweeter needing one.
But yes, in your experience an LCR was required, as it was in many peoples experience. I'm not discounting that. Just wanted to extend some other thoughts. As system said, zaph also mentions the use of a shunt resistor that can work.
- Status
- Not open for further replies.