passive crossover |
Post Reply | Page <123> |
Author | |
Andy Kos
Old Croc Joined: 15 May 2007 Location: Southampton Status: Offline Points: 3035 |
Post Options
Thanks(0)
|
I would agree and disagree... it very much depends on what components you are using. If you happen to have a crossover point that's not near the resonant frequency of any of the drivers, and the drivers are relatively well behaved, the laws of physics are fairly stable and standard accepted maths still generally works - online calculators will get you the correct answers providing you put the correct information in them. They may not take every factor into consideration, but you will get a crossover that works. If you want to fix tonality issues, then you are making more than just a crossover, you are also creating a passive EQ - that's much more complicated and that does require more in-depth knowledge and ideally some measuring equipment - but there are online calculators to approximate this too, that will work out the components you need to achieve the desired EQ notch. I think it does very much depend on what you want to achieve, and it does help to have some understanding of what you are doing - blindly throwing numbers into an online calculator isn't the best solution, but using an online calculator with a certain level of knowledge is just fine.
Edited by Andy Kos - 11 February 2016 at 12:22pm |
|
just a guy with a warehouse and a few speakers... www.bluearan.co.uk
|
|
stevie
Registered User Joined: 16 March 2008 Location: Dorchester Status: Offline Points: 425 |
Post Options
Thanks(0)
|
I feel your pain.
There is someone in Australia doing kits for your ribbons and they've published a crossover for the tweeter: <http://www.stonessoundstudio.com.au/ston/jp_Neopro5i_ribbon_tweeter_crossover_1.8Khz_3.htm> It might give you a start and should prevent your tweeters from blowing up. You'll have to pad it down further to use it with the B&C. You might like to scour the web for designs using those drive units. The tweeter also goes under the Audaphon brand name, which might help. Here's a site in a language that I don't understand (use Google Translate) with a crossover for a system using the tweeter: <http://www.audio-kontakt.com/forum/viewtopic.php?p=101392> Brilliant HF unit, by the way. I'd cross it over around 1.8kHz at 24dB per octave, at it has a nasty resonance at 1kHz that needs to be suppressed. |
|
DMorison
Old Croc Joined: 14 March 2007 Location: Aberdeen Status: Offline Points: 1647 |
Post Options
Thanks(0)
|
This Link seems to suggest you need at least 24dB for a 1500Hz crossover, so unless you really know what you're doing (and then you wouldn't have started this thread ) I'd go along with the other suggestions to raise the crossover point a bit. If you're not going to invest in some measurement gear, then at least read through This Guide to get more of a feel for the various snags to watch out for. Also, I'd echo the other comment about not tuning quite so low - around a third of an octave below Fs might be do-able but I'd be reluctant to try going below that without being willing to sacrifice a driver to testing. Again, you might get away with it if you know what you're doing - taking advantage of boundary loading the box to extend LF a bit perhaps, or if you know it's far more sensitive than you need so you'd only ever use a fration of its power rating for example. HTH, David.
|
|
stevie
Registered User Joined: 16 March 2008 Location: Dorchester Status: Offline Points: 425 |
Post Options
Thanks(0)
|
This is a 12 and a horn system I've just been working on. Fairly standard Celestion components with a JBL Econowave horn in a trap box. X-over is 1.3kHz. It looks like this:
Now put a standard crossover on there (2nd Order L/R with a zobel on the LF and an L-pad on the HF) and you get this: which is typical of the kind of results you get using crossover calculators. Garbage. Now perhaps Andy could post details of a system that does work with standard values as calculated by these crossovers - because I have never seen one. Edited by stevie - 11 February 2016 at 3:35pm |
|
Andy Kos
Old Croc Joined: 15 May 2007 Location: Southampton Status: Offline Points: 3035 |
Post Options
Thanks(0)
|
What is a 'standard' value? The calculator is basically just doing the maths for you - what you get out of it is very much dependent on what you put into it. Different components will yield different results. Comparing your two graphs, it does look very much like you are applying EQ as well as a crossover. A standard online crossover calculator will not give you the passive EQ you need, but it will get close with regard to just 'being a crossover'. As I stated before, if you have two relatively well behaved components, a standard default crossover will get you close. I have done this dozens of times myself for simple boxes such as 12", 10" or 8" with tweeter, and the standard calcs get me close enough for many applications. If I want the crossover to try to help get a more flat response I will develop it as required. I don't have plots to hand, but I can probably dig them out, where I have used a standard Butterworth at 3Khz to crossover between a horn-loaded 8" and 1" Comp. The results were acceptable, I then adjusted the capacitor on the High Pass slightly, and this helped smooth out a slight lump around 2.8Khz caused by the 8" Horn. I realise simple solutions don't work in every situation, and in many cases they are not ideal, but at the same time it does not mean they are completely invalid just because you have a different agenda. There are people who will quite happily try to fix the response on your second plot using a graphic and/or parametric EQ. There are other people who might just use the bass and treble controls on their amp to give the sound they want and be quite happy with it. The fact that a frequency response is not perfectly flat does not make it wrong - some people would argue that it adds colour to the sound - but it's only wrong if your belief is that a perfectly flat response is the only correct solution. |
|
just a guy with a warehouse and a few speakers... www.bluearan.co.uk
|
|
stevie
Registered User Joined: 16 March 2008 Location: Dorchester Status: Offline Points: 425 |
Post Options
Thanks(0)
|
Of course I'm applying eq! That's what crossovers do. A passive crossover doesn't just filter the frequency spectrum into different parts for each driver. To quote Siegfried Linkwitz: "Crossover filters for a speaker usually incorporate frequency response corrections for the individual drivers to obtain a desired overall response."
And because crossover calculators don't do this, they can't possibly produce a satisfactory result. Not only do these calculators assume that the frequency response of each driver is flat - which it may be with a dome tweeter but never is with anything else - but they don't take into account the phase response or the impedance response of the drivers. They are therefore bound to fail. Sure, you can get lucky and cobble something up that is "acceptable", but the OP is sitting on four very expensive drive units and I'm sure he expects them to sound great in a system. The dip in the middle of the frequency response of my second curve above has nothing to do with equalisation. The problem is that, at the crossover frequency, the drivers are 90 degrees out of phase and one of them is starting to roll off. So those people who would happily try to fix that response can fiddle all they like: they will never get that system sounding right. Curve no. 1 has an asymmetrical crossover and a shedload of eq on the HF unit. Edited by stevie - 22 June 2016 at 6:04pm |
|
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5483 |
Post Options
Thanks(0)
|
It entirely depends on your definition of satisfactory. Online calculators are just fine for someone constructing their first effort and who is learning. Once you get the idea of that then you are ready for more advanced test and measurement which can incorporate more advanced EQ - it is not a waste of time to start simple. Because for a start once you have bought a mic, measured your cab, and find that perhaps your first effort is lacking - you can just build a new xover board, pop it in a cab, and then compare before and after. This is all part of learning. Walk before you can run and all that. As long as you have not blown your drivers (somewhat unlikely) then you can always have another crack, and now you have some spare passive parts in your kitbox to boot.
|
|
stevie
Registered User Joined: 16 March 2008 Location: Dorchester Status: Offline Points: 425 |
Post Options
Thanks(0)
|
Fair enough. I've no objection to using crossover calculators per se. But let's be clear about what they can and can't do. I've no objection to using calculators to do my annual accounts, but a calculator is not going to tell me how much tax I need to pay.
|
|
Andy Kos
Old Croc Joined: 15 May 2007 Location: Southampton Status: Offline Points: 3035 |
Post Options
Thanks(0)
|
... there is a strong chance these units may sound great without the need for a massive amount of EQ, perhaps that is why they are expensive. I've seen lots of great sounding hi-fi speakers with very, very simple crossovers in them. Often in hi-fi SPL is sacrificed for performance, with the aim of a flatter response. In PA gear, SPL is often the primary concern, and in the quest to achieve maximum SPL, it's not uncommon to encounter other problems, often resulting in less than optimal frequency response. To some extent it depends on you interpretation of the intended purpose of a crossover. At a basic level, a crossover is a high pass filter and a low pass filter. In my mind that is the function of a crossover regardless of whether it is passive or active.
Fixing the frequency response (in my mind) is not part of the crossover, it is a separate function. You could do this externally with active processing via graphic or parametric equaliser. However in a passive crossover it is often convenient to add frequency response adjustments in the crossover circuit, which often gives a tidy solution. My point was purely that online calculators do the maths correctly for a basic filter, and yes I totally agree they are unable to do anything more than that, and often the ideal solution is somewhat more complicated and completely beyond anything that can be worked out in a basic calculator. Edited by Andy Kos - 11 February 2016 at 9:02pm |
|
just a guy with a warehouse and a few speakers... www.bluearan.co.uk
|
|
godge
Young Croc Joined: 07 July 2010 Location: Gloucestershire Status: Offline Points: 597 |
Post Options
Thanks(0)
|
This has made for some great and really interesting reading so far, definitely learning a lot about all the different aspects of building a crossover. I am going to persevere with the passive crossover as I really want to learn about it and put it into practice. I've spent good money on drivers so I might as well take the time to learn it properly
|
|
Andy Kos
Old Croc Joined: 15 May 2007 Location: Southampton Status: Offline Points: 3035 |
Post Options
Thanks(0)
|
Thats a good plan - definitely worth reading about to learn as much as you are willing/able to - it will help you reach the best result.
|
|
just a guy with a warehouse and a few speakers... www.bluearan.co.uk
|
|
odc04r
Old Croc Joined: 12 July 2006 Location: Sarfampton Status: Offline Points: 5483 |
Post Options
Thanks(0)
|
What will really pay massive dividends is if you get to the point where you understand how to express a sine wave as a complex number (it isn't actually that hard although it sounds awful if you dislike maths), then you can derive the transfer function of LCR circuits and in turn derive the bode plots (gain and phase) of the circuit you have just built. All in all this is a bit of work but once you get it then you will figure out so much more about sound theory in general.
|
|
Post Reply | Page <123> |
Tweet |
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |