Hello,
I've been reading this, which was interesting, especially the 0.5 speaker
http://www.diyaudio.com/forums/multi-way/38121-baffle-step-correction-circuit-how-design.html
However I have some slightly different questions
I'm currently using 2x low end drivers and a full range unit, crossed over at 500Hz with a 1st order crossover
The woofers present a 4ohm load and the mid/high an 8ohm load
If I wanted to integrate a BSC would I be looking at a seperate one for the high and low drivers?
I take it BSC comes before the crossover?
How would it affect phase as I will now have 2x inductors on one driver and and an inductor and capacitor on the other?
Off the top of my head the BSC f3 point is around 720Hz, would it make more sense to crossover higher up and just pad down the mid/high driver a bit?
Regards
Tom
I've been reading this, which was interesting, especially the 0.5 speaker
http://www.diyaudio.com/forums/multi-way/38121-baffle-step-correction-circuit-how-design.html
However I have some slightly different questions
I'm currently using 2x low end drivers and a full range unit, crossed over at 500Hz with a 1st order crossover
The woofers present a 4ohm load and the mid/high an 8ohm load
If I wanted to integrate a BSC would I be looking at a seperate one for the high and low drivers?
I take it BSC comes before the crossover?
How would it affect phase as I will now have 2x inductors on one driver and and an inductor and capacitor on the other?
Off the top of my head the BSC f3 point is around 720Hz, would it make more sense to crossover higher up and just pad down the mid/high driver a bit?
Regards
Tom
If you are crossing over at 500 and calculate your baffle step issue is at 720, why do you think you need "baffle step correction".
Proper design is to compare your raw woofer response to your ideal target. The difference between the two will indicate what order and filtering you require. Tackling baffle step as an individual item is an unnecessary complication that gives no benefits.
David S
Proper design is to compare your raw woofer response to your ideal target. The difference between the two will indicate what order and filtering you require. Tackling baffle step as an individual item is an unnecessary complication that gives no benefits.
David S
In my speaker?
I didn't, I used the 115/baffle width
I think I need it because I was disappointed with the low end response and I read this 'Unless compensation is applied for the baffle step, the speakers will sound subjectively thin.' And I thought yes they do sound a bit thin! (or maybe bright) I was consious that all I could hear was the tss tss tss of cymbals in tracks
With a little eq lowering freqs above the 700Hz mark they have a pretty impressive sound, very warm and a lot closer to what I wanted
If I'm honest I'm pretty 'baffled' because to me if I was using 2x low end drivers, that would be a doubling of power / surface area compared to the mid anyway, so I wouldn't have thought I would have needed it...
If you are creating a 2 way with a typical high crossover point of 3kHz you may find that the bass shelves down from the natural midrange level. This is because a typical woofer is fairly flat in half space loading. For upper frequencies the baffle is, near enough, half space but for lower frequencies it isn't and bass response shelves down. This is referred to as baffle step.
For most systems professional designers will just incorporate shelving into the crossover network in the form of enough inductance to pull the midrange down. I like using a 2nd order network with an RC in series across the woofer (damped 2nd order, not a conjugate as the elements are not optimized for that). That allows you to use a large series inductor to flatten the midrange and still get a sensible crossover corner.
Designing in a separate crossover section for the sake of shelving (and presumably creating a crossover with no amount of shelving) is simply wasteful of crossover elements and confers no benefit.
In your case of having a crossover in approximately the shelving region, the issue should disappear. If you set the separate midrange (full range) driver to the right level relative to the woofers bass level then you have inherently dealt with it.
"The woofer is your BSC"
David S.
Ok, I dont really understand this bit
I do understand this bit, however I do feel that 500Hz is higher than I wanted and wanted to cross future pairs down lower, maybe around 200Hz (I want to keep the full range unit as full as possible)
I also planned to stick to 1st order crossovers
So, if I theoretically wanted to go ahead with including power sapping inductors into my design pre crossover - would I be looking at two different ones, one for the 4Ohm low end and one for the 8ohm high end?
I have a sub next to me at the moment, the baffle width is around the 200Hz mark, I can see that that would be ideal, and possibly something for me to consider in future designs
For as much talk and explaining, we don't know what we are talking about or explaining (besides theory and not your practical case). So do us a favor and post full specifications for the drivers/model/specs you are using and the full crossover layout (and baffle dimensions you are mentioning), agreed?!. Best.
If you want to cross below where you need to compensate, you could use a high pass filter on the upper driver that has less damping, second order may be useful here, as an alternative to using a more sensitive mid and bringing most of it down.
If you cross above this point you can use a low pass filter on the woofer that rolls off more slowly toward the top. A damped second order filter can be made to act like first order at the start of the roll off, moving to second order later.
If you cross above this point you can use a low pass filter on the woofer that rolls off more slowly toward the top. A damped second order filter can be made to act like first order at the start of the roll off, moving to second order later.
Last edited:
Oh, I see. Its a first order thing.
Th problem with first order is that you largely give up your ability to EQ with the network. Then when the results are unsatisfactory you can add back secondary circuits, such as "baffle step correction" and get what you needed in the first place: a sufficiently complex network to do the job.
As to "power robbing inductors", they would be used for shaping of the woofer only. Unfortunately, if you need to flatten a system with a passive network, then power robbing is a given.
David S
Why would the inductors shape the woofer only?
My personal views are that if I was using the same impedance drivers then I could feed them both through the same 'baffle step', and then cross them over, so the signal would be affected as a whole and then split for the drivers
As they are different impedances then that makes things a little more complicated for me, I think (and quite possibly something to learn from for future, but to me thats what this whole building things is about, I'm not good at learning from theory)
However I dont know if this is correct or not
For an idea of what I'm working with these are 2x HiVi M4Ns + 1x B3N, crossed over at 500Hz (I can't remember the exact values) mounted in a TL tuned to 45Hz and a baffle width of 155mm
If you were to do baffle step compensation at line level, you could have your cake and eat it too. That is, there is no problem with different impedance of the twin low-range drivers, and the full-range, and the crossover network can remain first order. Well, this would be "institutionalizing" what you did with your equalizer.
Here is one design of line level compensation, probably very familiar to members of this forum:
Baffle Step Compensation
Regards,
cT
Here is one design of line level compensation, probably very familiar to members of this forum:
Baffle Step Compensation
Regards,
cT
That suggests that you may have too much woofer for the FR, and an XO between 525 and 750 Hz would be appropriate. I have found that this low, series XOs work well. If you have a parallel XO, try rewiring the same parts in a series XO.
dave
Baffle step is a transition from response at at one level to response at another. With the usual 2 way this occurs in the woofer's range so all the compensation is applied to the woofer only. What we keep telling you is that your lower crossover point is more or less at the baffle step frequency. That means that in your case no particular compensation is needed, you will use the woofer entirely below the baffle step frequency and the full ranges above it. All you have to do is match their levels and you will have flat response free from b.s.😀
David
Hi,
The B3N is around 81dB, the M4N around 82dB. Two in parallel
will be 88dB. If crossed over 1st order (preferably a series type)
in the middle of the baffle step transition you will automatically
get full baffle step compensation, due to the given sensitivities.
500Hz should be very near correct for that baffle width.
rgds, sreten.
The B3N is around 81dB, the M4N around 82dB. Two in parallel
will be 88dB. If crossed over 1st order (preferably a series type)
in the middle of the baffle step transition you will automatically
get full baffle step compensation, due to the given sensitivities.
500Hz should be very near correct for that baffle width.
rgds, sreten.
Last edited:
Unless you're using some passive parts for EQ / breakup management, I think you're probably hearing the response of the B3N, rather than an LF drop-off:
http://www.parts-express.com/pdf/297-428g.pdf
Breakup on the M4N might also need to be dealt with for 1st order use, even at 500Hz, as it's pretty rough:
http://www.parts-express.com/pdf/297-434g.pdf
P.S. PE has response and impedance data posted, simulation could be done.
http://www.parts-express.com/pdf/297-428g.pdf
Breakup on the M4N might also need to be dealt with for 1st order use, even at 500Hz, as it's pretty rough:
http://www.parts-express.com/pdf/297-434g.pdf
P.S. PE has response and impedance data posted, simulation could be done.
Last edited:
Aye I have read that, nice idea, along with active crossovers or bi amping, something for future designs maybe..
Yes, I kinda would have thought so too. I wouldnt mind trying that, however, I've glued the sides on :-/ ah well, live and learn!
I know! tbh I was expecting a slightly bass heavy speaker if anything, was v suprised when that didn't happen
This is pretty shocking, maybe I am just listening to the woofer. I thought these were quite nice speakers, they have a very nice sounding low end response, but maybe I was wrong on my upper frequencies.
I did say I wasnt going to use them again, but mainly due to their poor power handling, and I am planning on avoiding metal drivers in general due to their spikey responses!
My plan now is an L Pad on the mid/high and see how that affects the sound 🙂
Hi,
Your bet bet is a series trap for the 8KHz peak, if B3N = B3S.
Its unlikely a L pad will improve matters, given the numbers.
rgds, sreten.