crossover filter

[tom_s]

The TCP115 can sound surprisingly good when matched with a budget tweeter and the appropriate crossover. But that crossover will be a little more than just a cap and resistor. I understand money being a major issue, but using completely different drivers on each side goes against the whole concept of stereo reproduction. I would rather have a good sounding mono speaker than a bad sounding stereo pair.

[SentinelAeon]

I forgot to add, woofer will be on left channel, full range on right. I normaly do crossover at 4K like this ... and it works ok except some strange passes in some stereo songs. The thing is ... to get around this ... i want woofer and full range to play most of the same frequencies to reduce this. Its not optimal and in this cases i rather just mix the signal into mono so both channels play the same thing.

I will have to test it and see. The reason is, this is the first person that asked me to make them a speaker that will actualy listen to classical music. My extra budget speakers with cheap parts play dance music just fine. But classical music is different and this is where i am limited by cheap parts. Dayton audio TCP115-4 and Visaton FRS 8 M are both ultra cheap parts. But as it is i am making the speakers basicaly for free. If i invest in high quality speakers i will actualy be losing money which is not my intent.

[Chris Roemer]

To expound - IF both drivers are nearly (perfectly) "in-phase" they'll actually "sum" +6dB from roughly 150-4000Hz.
If the drivers are 90* out-of-phase (same as 90* in phase), as in a Butterworth alignment, they'll sum +3dB over that overlapping range.

With NO filters (and 8ohm drivers), you'll end up w/mostly a 4ohm load.
With simple 1st order filters (cap on FR, and series coil on woofer) the load will be more like 8ohms (if these are 8ohm drivers), and the response will be approx. 90dB (IF you start w/90dB drivers) from about 60Hz up to almost 20kHz (+/- 2dB maybe - depending ...).

[LOUT]

Woofer: covers 50-4000Hz
Full Range: covers 150-18.000Hz

1) Does this mean the frequency at 150-4000Hz that overlap are amplified ?
2) Does this mean that the frequencies at 150-4000Hz cancel eachother and actualy it plays less loud ?
3) How does it affect the quality?

In that 150-4000hz range where both speakers are overlapping, they'll add together (play louder in there) as long as the speakers are playing in-phase with each other OR they'll cancel out (less loud than just one or the other) if they're out-of-phase with each other.
You'll be able to flip a speaker's phase 180degrees by flipping the -/+ terminals around...so this can be a decent way to get them in or out of phase depending on some other things that sadly make it more complicated.
A speaker's phase can be different at different frequencies, two different style speakers can have different phase even at the same frequency, and crossover parts will basically always change the phase compared to playing the speaker without any crossover parts...so there's a decent chance the speakers might add-together to play louder at some frequencies and cancel-out quieter for others within that 150-4000hz overlap.
If you move around or sit in a different place while listening, this can also make it so one speaker might be a little closer and the other a little farther from you compared to other spots you might listen from...this will also change how much they add or cancel, particularly at higher frequencies where they're overlapping (higher frequencies are shorter waves which makes them more sensitive to smaller changes in distance between two speakers playing the same overlapping high notes).

These changes can result in a less smooth/flat frequency response where some notes will be lounder and others softer when they should be more equal....and it can give a strange "phasing" or "comb-filtering" sound (like a kind of sci-fi swooshing) when you move a little and overlapping highs change from adding to cancelling or the other way around in little bits at a time.


It general;
overlapping is a LOT safer for low-ish mids and bass notes, more problematic for high-mids and highs,
it's safer when you're doing it on purpose with two of the same speakers using the same crossover parts (or carefully planned different parts that keep the speakers in-phase during their overlap) which will keep phasing simpler and together where you want it,
it's also safer when you can have the overlapping speakers as close to eachother as possible (often touching edges) this makes moving have less affect on the speak-to-speaker distances,
and it's better when they're above/below each other instead of side-by-side because a sitting or standing person is more likely to move their head side-to-side but not up/down significantly...and a multi-person audience will usually be spread out toward the sides rather than stacked on top of eachother....a top/bottom overlap will keep the same speaker-to-speaker distance as you move side-to-side and only change a small amount when you move between sitting or standing as long as you're a decent distance away.

[SentinelAeon]

Instead of making a new topic, i will post some questions here. I didn't have the chance to test the stereo mixing and/or button placement yet.

We usualy make a crossover for tweeter and woofer so each cover their own part of frequency graph. I would like to understand what happens if we do not do a proper crossover and woofer/full range both cover certain part of frequency. Just for the sake of simplicity lets asume that woofer and full range both play at same loudness. Meaning both playing vs only 1 playing adds +3dB (~25% increase). I will give you a simple example:

Woofer: covers 50-4000Hz
Full Range: covers 150-18.000Hz

So my question is, what is the effect of this ?

1) Does this mean the frequency at 150-4000Hz that overlap are amplified ?
2) Does this mean that the frequencies at 150-4000Hz cancel eachother and actualy it plays less loud ?
3) How does it affect the quality.

I hope you can help me understand how this works to better understand the reasons behind the crossovers we all use.

[Chris Roemer]

A single (series) "cap" doesn't really do much to protect (roll off) the bottom end of a woofer, reason being that "moving coil" drivers have an impedance peak at their resonant freq. (Fs), which is already on the bottom end - just before rolloff. The Fs peak is typically 2x-4x a tweeter's "nominal" impedance, but woofers can have a peak MANY times nominal - like 10x, or MORE.

[SentinelAeon]

Well i used cheap capacitor to prevent it to play to low and get damaged. Some said that it would be good to also prevent woofer from going to high. Though if this things are more expensive than say 3$ a piece, i will not do it.

[Kornbread]

Most of the people here will use an air coil but when sizes start to become large many may switch to a cored inductor mainly because of price and size.

As an example ,,, just some larg'ish random size ... A 7mh air core can range from $20 (18awg, 1.6ohm, 2.5x1") to $50 (15awg, .98ohm, 32.5") while a 7mh cored inductor can be had for $8 (18awg, .61ohm, 1.25x1.75") to $30 (16awg., .297ohm, 4.5x1.6")

[Chris Roemer]

A coil has 2 (or 3) "main" specs: its inductance (mH, which generates most of the rolloff characteristics), its DCR (basically the resistance of the wire in the coil - as if it were in a straight run, i.e. NOT coiled) which has an attenuation effect on the driver's output, (and the coil's power rating - before saturation).

A coil w/a DCR of 1.0ohm MIGHT cut a woofer's output by -1dB (for instance). A general ROT is that a series coil's DCR should not exceed 5% of a driver's "nominal" impedance - but...
on the passband for a midrange (for example), if the driver needs padding down anyway, it's cheaper to use a higher DCR coil (higher gauge #) which makes more sense than buying an expensive coil then adding a series resistor in circuit.

Iron cored coils get some of their inductance from being wrapped around a ferrous "core" (could be solid, or laminated "plates", or basically a compressed ferrous "dust"). They're useful for achieving higher Le values at a lower cost and lower DCR (they use smaller gauge wire), but - they tend to saturate at lower power levels than air-core coils.

So, it depends on how you're driver model goes, and how it and your XO design interact w/other drivers (in a multi-way system).

By "cap" your woofer, do you mean to roll off its top end? ?

[SentinelAeon]

Since this is a crossover topic i will ask 1 more thing. What kind of inductor i need to cap my woofer ? I know the value but does it needs to be some special kind of inductor or i can just go to ebay and order coil of correct value ?

[Chris Roemer]

A box size that the NS3 and the PC68-4, as well as the TCP-4 CAN work well enough in is 0.10cf.
The PC can only do 60Hz but can get by w/only a 1"id port (that's 4" long), but it's limited (2mm) Xmax tops it out at only 3w RMS (near 87dB).
The larger 3" & 4" drivers need a 1-1/4" port, but with more Xmax, they can take higher power. The NS3 can take 6w down to 50Hz (at close to 90dB); whereas the TCP (which is the largest AND the greatest Xmax) can handle 30w RMS down to 50Hz, achieving 97dB. They'd both need a vent about 10" long (or PRs) in a 0.10cf box though.

[Kornbread]

Not familiar with that software. Maybe if there is an issue with port velocity it automatically steps the port od/id up in size?

[SentinelAeon]

I have a question. QSpeaker program which i like, always seems to give me certain port diameter. I can manualy change its diameter to bigger or smaller and it adjusts length accordingly but if i change enclosure size or resonant frequency, it will change to this "default" port diameter it likes. Its usualy 3.20cm. So my question is, what is special about this 3.2cm diameter ? Do i lose something if instead i have port of 2cm diameter? Because that would be a LOT better for me since port will be shorter and will take less space so there is more space for other components.

[SentinelAeon]

Well i like the speaker but it has a very very bad sensitivity, the daytons are much better in this regard. i might have a problem with my small amps where i need battery life.

[LOUT]

Has anyone compared the PicoNeo's NS3 (older or newer) to the PC68 for fullrange in a small, ported or PR box?

Goes kind of OT
[SPOILER]
I read the NS3's aluminum cone had some less favorable qualities, and I'm assuming its 3inch size means its highs above 6khz fall pretty hard...the PC68's smaller 2.5 size lets it stretch a little higher to around 8khz before the highs start falling badly and it uses a different cone material (though I have no idea if this is a positive or negative in reality). The 68's smaller size also lets it reach a tiny bit deeper in a slightly smaller box while its high-ish sensitivity lets it keep up (at least on paper) on a watt/per basis.
Although they're both about equally sensitive, I'm guessing the NS3's larger size and Xmax can reach louder max volumes. But assuming they're 8ohm drivers being used in a battery-powered wireless portable I think you'd need a 40/[email protected] portable amp along with a larger-than-average battery bank for that to matter.
I think the PC68 is also a little cheaper if you're pinching pennies.

I haven't heard the NS3's but I bought a few of the 8ohm versions of the 68's and found out I like them a lot more than their 3inch sisters (the PC83's that came with a KAB bluetooth amp kit) at least for full-range duty. Despite the 83's flatter response (which looks a lot like the NS3's spec'ed frequency response), I found it a bit too harsh and directional, while the 68's have been a lot of fun with a simple L/(R+C) BSC each in a 2Liter, ported box. [/SPOILER]