Total newbie, and want to build something on my own. Since I am going to screw up (as everyone has told me on here), I wanted to start with something cheap.
I am one of the believers that frequencies listed as ultrasonic and near ultrasonic make a bigger difference than most people think. The little Dayton ND16FA caught my eye on Zaph's site, but I want a speaker that can do for both music and theater, which means playing reasonably loud. These are listed as 89db sensitivity with RMS of 10 watts. These look like they were meant as a line array, so I can use more than 1, but run into comb filtering issues, and cheap ribbons start to look cheaper the more of these I'd need.
Take a look at the Shtick's.
These have 2 B3S (or other mids) with a ND16FA between them. Why use 2 mids, since I would think even 1 mid should be able to handle more power with less distortion than 1 ND16FA? This is not the only design I've seen like this.
How much does the crossover affect real world power handling of these ND16FA's? They have a Fs of 2,283Hz. I could cross them first order (electrical) at 4566Hz which I would imagine would distort or damage much easier than if I did a 4th order at 10 kHz. Which way did Dayton do the RMS spec? Or is RMS an arbitrary number I am paying too much attention to?
I am one of the believers that frequencies listed as ultrasonic and near ultrasonic make a bigger difference than most people think. The little Dayton ND16FA caught my eye on Zaph's site, but I want a speaker that can do for both music and theater, which means playing reasonably loud. These are listed as 89db sensitivity with RMS of 10 watts. These look like they were meant as a line array, so I can use more than 1, but run into comb filtering issues, and cheap ribbons start to look cheaper the more of these I'd need.
Take a look at the Shtick's.
These have 2 B3S (or other mids) with a ND16FA between them. Why use 2 mids, since I would think even 1 mid should be able to handle more power with less distortion than 1 ND16FA? This is not the only design I've seen like this.
How much does the crossover affect real world power handling of these ND16FA's? They have a Fs of 2,283Hz. I could cross them first order (electrical) at 4566Hz which I would imagine would distort or damage much easier than if I did a 4th order at 10 kHz. Which way did Dayton do the RMS spec? Or is RMS an arbitrary number I am paying too much attention to?
In this case, using a first order crossover is a bad idea. A first order crossover at 4.5kHz for a tweeter with a 2.2kHz resonance frequency will allow too much power to reach the tweeter, and there will not be enough attenuation at and below the resonance frequency. Don't use a first order crossover on a neo tweeter. In the old days when tweeters used a good sized magnet structure, the driver could stand the power that would reach them with just a blocking cap and would just heat up. The little neos can't take the power, and will likely be damaged.
The power spec in this case (10 Watts) is likely "real" power. Sometimes you see a tweeter quoted as being able to handle a couple of hundred watts of power. That's a misnomer really - that much broadband power would melt things quickly. What they mean is that, when used with a prudent crossover, the SYSTEM power could be up to a couple of hundred watts. Because the crossover limits the bandwidth coming to the tweeter, and high frequencies above a couple of hundred kHz do not contain much of the power in the music spectrum, the total power actually reaching the tweeter might be less than 10% of the total that is input to the crossover network from the power amp. The steeper the crossover, the better it is at rejecting out of band signals, so using steeper crossovers will mean less power will reach the tweeter.
A good way to judge where you can cross over a tweeter is to look at the distortion plots. This is often how Zaph comes up with his recommendations about where to cross over to the midrange. Zaph's mentions "Works well from 3500 Hz and up" but this would be only when used with a steep crossover, e.g. 4th order or higher in my book. Many drivers have increasing distortion as you come down in frequency towards the driver resonance, and this is one reason why you typically cross over at least one octave higher with a steep filter, or much higher with a shallow filter.
-Charlie
1. 10W is not aleatory (Dayton ND16FA).
2. Crossovers have low frequency components (lower than the crossover frequency) that can damage the tweeter.
3. You don't need to bother with that if not using pro/PA/studio power because the percentage of power going to a tweeter is very small. Let's say for 10W of power in the tweeter you need 100W of power woofer/sub/amp.
4. Usually, if you plan ahead, you never reach max. power.
5. You don't need to screw up. 😉
2. Crossovers have low frequency components (lower than the crossover frequency) that can damage the tweeter.
3. You don't need to bother with that if not using pro/PA/studio power because the percentage of power going to a tweeter is very small. Let's say for 10W of power in the tweeter you need 100W of power woofer/sub/amp.
4. Usually, if you plan ahead, you never reach max. power.
5. You don't need to screw up. 😉
You took my example the wrong way. I meant to post an example comparing the most conservative and least conservative crossover someone might consider, I was not suggesting building either. My point was that a more conservative crossover should give a higher RMS, or at least less distortion, and therefore the RMS number Dayton posted is only vaguely useful unless I know what crossover they meant to use it with.
I agree 10 watts is probably enough for most music and movies. Suppose I ran a test test tone of 20kHz at 20 watts (about 102db). Would the voice coil melt or the tweeter otherwise be permanently destroyed? If so, that is a serious limitation, even if I don't sit around listening to 102db 20kHz test tones all day.
I still want my previous questions answered, but people seem quick to make crossover suggestions, so here are a few other drivers. Right now, this is still a "thought experiment" and not a real system, so I won't actually build it, but want to start thinking about something I could build.
I would cross the Dayton ND16FA at 4.5k using a 2nd order Linkwitz Riley to a Fountek FR88-EX. The price has gone up dramatically since Zaph did his tests, but this is still within budget, and IMO the cheapest significant step up from the HiVi B3S/B3n that works with the Dayton ND16FA.
This does not get me low enough to cross to a subwoofer, so I have to make this a 3-way, and could use a Dayton RS225-8, which gets me all the way to full range. I could go to a smaller woofer if I had limited cabinet space. Either way, I don't see a reason to use a higher order crossover with the Fountek FR88-EX, so would cross with a 1st order, probably at 380Hz, which is 2 octaves above its FS of 95Hz.
This is starting to resemble a "poor-man" version of the Statements, and I could easily go to the Tang Band W4-1337SA mid if I wanted, but I like the lower distortion (and lower price) on the Fountek.
If people think this is a good design, I will model it in Spice as a next step. Otherwise, back to the drawing board.
I would cross the Dayton ND16FA at 4.5k using a 2nd order Linkwitz Riley to a Fountek FR88-EX. The price has gone up dramatically since Zaph did his tests, but this is still within budget, and IMO the cheapest significant step up from the HiVi B3S/B3n that works with the Dayton ND16FA.
This does not get me low enough to cross to a subwoofer, so I have to make this a 3-way, and could use a Dayton RS225-8, which gets me all the way to full range. I could go to a smaller woofer if I had limited cabinet space. Either way, I don't see a reason to use a higher order crossover with the Fountek FR88-EX, so would cross with a 1st order, probably at 380Hz, which is 2 octaves above its FS of 95Hz.
This is starting to resemble a "poor-man" version of the Statements, and I could easily go to the Tang Band W4-1337SA mid if I wanted, but I like the lower distortion (and lower price) on the Fountek.
If people think this is a good design, I will model it in Spice as a next step. Otherwise, back to the drawing board.
Let's see if I understand. (Forget the subs now for a moment) The driver you like most, as you are saying is (answer space for the driver name)______________, because of lower distortion, good quality vs "poor-man" Statements, voicing good, nice specs..., good frequency output, good looking, you tell me. (Name just one driver please, so that we/you can "make up your mind").
As I said, this is more a thought experiment than an actual build, but it probably resembles what I will end up with.
If I want a clean top octave and maybe even above 20,000Hz, I need a ribbon or a neo tweeter.
Unless I spend big bucks on a more expensive ribbon or other non-traditional driver, I need a to cross over to this top driver at a high (3kHz+) point. This means I can cross to a conventional tweeter or a 3" driver - I don't think anything larger will play high frequencies well enough to cross that high. A 3" driver will go lower and have a smoother mids and much of the tweeters bandwidth would be wasted, so a 3" is better.
I am sure there are lots of good 3" drivers, but of the ones with data on Zaph's site, the B3N, the B3S, the Fountek FR88-EX and the Tang Band W4-1337SA look the best as far as crossing over that high and being able to play well over a wide range of frequencies.
The B3N and B3S are inferior to the other 2, and all 4 are within budget. The Tang Band costs more, has a smoother response, but more distortion - so I like'd go with the FR88EX unless I was deliberately trying to do something that resembles the statements. The statements got good reviews, and also cross this high to a 3" mid that covers a wide range of frequencies..
A 'statement' with a neo tweeter instead of a ribbon, and only 1 mid and 1 woofer would probably sound surprisingly similar and cost less than half as much. Lots of people miss the larger woofer when writing about the mini-statements, monitors and centers; but I haven't seen a single post saying how much better 2 W4-1337SA's sound than only 1. Of course it won't be the same, but again, it costs well under half the price.
If I want a clean top octave and maybe even above 20,000Hz, I need a ribbon or a neo tweeter.
Unless I spend big bucks on a more expensive ribbon or other non-traditional driver, I need a to cross over to this top driver at a high (3kHz+) point. This means I can cross to a conventional tweeter or a 3" driver - I don't think anything larger will play high frequencies well enough to cross that high. A 3" driver will go lower and have a smoother mids and much of the tweeters bandwidth would be wasted, so a 3" is better.
I am sure there are lots of good 3" drivers, but of the ones with data on Zaph's site, the B3N, the B3S, the Fountek FR88-EX and the Tang Band W4-1337SA look the best as far as crossing over that high and being able to play well over a wide range of frequencies.
The B3N and B3S are inferior to the other 2, and all 4 are within budget. The Tang Band costs more, has a smoother response, but more distortion - so I like'd go with the FR88EX unless I was deliberately trying to do something that resembles the statements. The statements got good reviews, and also cross this high to a 3" mid that covers a wide range of frequencies..
A 'statement' with a neo tweeter instead of a ribbon, and only 1 mid and 1 woofer would probably sound surprisingly similar and cost less than half as much. Lots of people miss the larger woofer when writing about the mini-statements, monitors and centers; but I haven't seen a single post saying how much better 2 W4-1337SA's sound than only 1. Of course it won't be the same, but again, it costs well under half the price.
Home audio speakers should normally be capable to produce a spl of 100 dB at 1m. A small driver like the Fountek FR88EX reaches only 96 dB at the rated power handling limit. To get 100 dB, you need two of them or a single 4.5" midrange. That's why the Shtick's and other designs do not have a single 3" midrange driver.
Attachments
So I either have to accept the power limitation, or I use 2 drivers and have comb filtering issues. This is harder than I thought - except the original statements should have the comb filtering issue and the rest of the statement line has the power limitation, so maybe this can work.
- Status
- Not open for further replies.