I don't think that series filters are intrinsically different to parallel ones regarding to driver 's impedance influence on slopes and on everything in general, meaning that you cannot exactly know how this ... xover behaves untill you know what ... drivers you are using, because this ... choice matters both electrically and acoustically. That's life, no miracle, you have to make some choices and know what the consequences of your ... choices will be. This is the only universal xover law that i know about...
Last edited:
If you see that 3 parts are shared depending on the route your signal takes, you have a damped second order lowpass, ie- a coil and CR shunt. Then you have a C-LR shunt-C in the tweeter circuit, ie- a damped third order.
Being that this is a SXO, and the parts do conduct AC both directions for a perspective. It's the resistor and it's placement, as well as where the drivers are located in the circuit that make this very different than an average SXO.
Later,
Wolf
Being that this is a SXO, and the parts do conduct AC both directions for a perspective. It's the resistor and it's placement, as well as where the drivers are located in the circuit that make this very different than an average SXO.
Later,
Wolf
You do? It looks pretty damn close to a 6dB/octave roll off to me
Freq - woofer level, BSC woofer level = difference
300Hz - 89db, 86dB = -3dB
600Hz - 88db - 81.5dB = -6.5dB difference
1200Hz - 80db - 69dB = -11dB difference
2400Hz - 67db - 52.5dB = -14.5dB difference
Ideal 1st order butterworth
Fc = -3dB
2Fc = -6dB
3Fc = -12dB
4fc = -18dB
1.2 and 2.4k are probably off a bit due to guessing the position in between the lines on a logarithmic scale. Also the woofer impedance is starting to rise about 1kHz so you get slightly less than 6dB/octave from there up. What is important it that it adheres to a 1st order slope almost exactly for the first 2 octaves.
How else do you propose to implement baffle step compensation passively 😕
Last edited:
So, what the fc is? i am trying to figure out what drivers i will need.
Cheers
Sent from my HUAWEI MT7-TL10 using Tapatalk
That will depend on the drivers for sure. I can't wager an Fc without a full sim, and this would require Xsim to simulate it.
Later,
Wolf
Later,
Wolf
One thing with 2.5ways is the danger that the second woofer adds too much of a boost, obviously depending on drivers etc. Then the speakers end up about 2 meters from the wall and the wife has a fit.
Are there any 2.5 way designs where the midbass is sealed and 0.5 is BR? Or would that cause other problems
Are there any 2.5 way designs where the midbass is sealed and 0.5 is BR? Or would that cause other problems
Last edited:
Phase problems, see no full correlation anymore, as Bassreflex is a fourth-order filter, but Sealed only second-order.
I have a pair. Bass is very deep, but the midrange and high are not bright enough.
That is why i am thingking about a pair of new midbass and xover like this.
If the tuning on the BR is low enough compared to the F3 of the sealed midbass then it could work. If they are too close then you could get some cancellation before the sealed driver rolls off an adequate amount. The other consideration is that you no longer get the benefit of excursion reduction and therefore distortion reduction from porting both drivers.
Why not just use a shelving BR alignment (for both drivers in common chamber) though?
I think if spacing the speakers out from the wall resolves the problem then the issue might not be the bass alignment but simply the amount of baffle step compensation. You can reduce the amount of BSC in any design by bypassing the baffle step inductor with a resistor to raise the level of the midrange. Obviously it isn't something you can change easily because you have to raise the level of any mid/tweeter to match - that's why it's recommended to aim for 4dB of baffle step instead of a full 6dB if the speakers are going to be placed near a wall.
Last edited:
Adding a bsc means one of the key points of a 2.5 way is lost IMO. It's a tricky one..
The other possible solution is to go two chamber sealed and roll in the second woofer lower down to fill in below 100 hz. But then you'd probably lose too much output just below baffle step.
The other possible solution is to go two chamber sealed and roll in the second woofer lower down to fill in below 100 hz. But then you'd probably lose too much output just below baffle step.
Huh? Are you confusing '2.5way' and 'TMM'?
'.5-way' literally means adding a low-pass filter to half of the woofers in order to achieve baffle step compensation. One woofer only plays the frequencies below the baffle step frequency. The other plays the low frequencies as well as some frequency range above the baffle step frequency. Thus you get +6dB due to the extra woofer playing below the baffle step frequency which makes up for the 6dB loss caused by baffle step losses. Without this you don't have a 2.5way, you have a 2-way TMM or MTM.
The '2.5-way' nomenclature comes about because while you are feeding 3 speaker drivers (two woofers, one tweeter) 3 different signals, the woofers both play the bass therefore it is not a traditional '3-way' with a dedicated bass driver(s) but it is neither '2-way' because that would imply that both woofers were playing the same signal. Hence the '2.5-way' because it is somewhere in between being 2-way or 3-way.
In a similar fashion you can have a 3.5-way like so:
http://www.zaphaudio.com/ZDT3.5.html
Also be aware that you can apply BSC to a 2-way TMM without turning it into a 2.5way. This implies that both woofers are driven by the same signal, but a filter is put in place to create a 6dB shelf above the baffle step frequency.
Last edited:
Sorry it was early/half asleep -I thought you were talking about adding a correction circuit to a 2.5 way.. I'll re read.
Hi,
Nope, you can't analyse a series x/o like its two parallel paths.
It is 1st order and 2nd order, not 2nd order and 3rd order.
An externally hosted image should be here but it was not working when we last tested it.
Figure 1.1 - Series and Parallel 1st Order Filters (8R-1KHz)
According to your very misguided thinking 1st order series is
2nd order because the bass has a series inductor and parallel
capacitor, and the tweeter a series capacitor and parallel inductor.
Just plain wrong. The responses above are identical.
As are the voltages and currents for all components.
rgds, sreten.
Last edited:
Hi, Sreten, thou are right, that the series crossover thou posted is equivalent to the parallel one. But, if component values are changed (bigger C, smaller L), the result can become a cross-over of second order, because L and C are interacting now. If L and C sit in separate, parallel branches, they cannot interact.
The L and C are always interacting in a series crossover.
Of course to some degree it could be observed that one takes over from the other, ie that without the C, the woofer sees a parallel RL shelving filter and the C then attenuates the shelved region at 6dB/octave.
However different L and C values (same fresonance) change the damping at their resonance, higher or lower, and the C still takes over the top end but at a different frequency.
Hi,
Nope, the 1st order series x/o can never become second order,
and can never have offset x/o points like the parallel 1st order.
Some people just don't get how series actually works.
rgds, sreten.
Lots of info regarding 'zeta' and 'quasi-12dB' SXO circuits in reference to SXO 1st-order are using a resonance at the Fc of the xover between the 2 parts to steepen the knee of the transfer function. Being that I've never done a xover based on zeta, I cannot say whether this is accurate or not.
My hunch is that the Q is reduced at the Fc, and even if the initial slope is greater than 1st order, that it will in effect asymptote to 1st order electrical + acoustic slope further down the rolloff.
Now- you refer to having offset xover points on a parallel xover. If you have 2 drivers, you have one xover point- period. You can use differing Fn* marks to spread the response out and reduce the peaking/dipping at Fc, but there is still only one Fc. This is the same for all 2-way approaches. 2 drivers = one Fc; where the rolloffs of the drivers pass each other. (*n= -xdB)
This is commonly misunderstood by many newcomers to the loudspeaker hobby.
Later,
Wolf
Yes, you can manipulate a series XO so that initial roll-off is close to 2nd order, but if you go far enuff it is still 1st order.
dave
- Status
- Not open for further replies.