crossover guides

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Can anyone help me with some links to crossover guides that talk about why which 2nd order crossover should be used, say linkwitz-riley, butterworth, bessel, chebichev?
I want to attenuate my speaker's tweeter and pulled out my crossover to see where i would place the resistors, and looking at the values they must have chosen the values trying to avoid some bumps/dips in the resulting curve.
If the guide would discuss why which type of capacitor is used that would be even better, one cap is a electrolytic little cap and the other seems to be a square metalized film capacitor.
Also if it would explain why i would use a air core vs ferrit core inductor would be interesting to know.

If i can't learn why the capacitor values might have been chosen that way and change the values if it makes sense, i would at the least like to upgrade the quality of the capacitors used.

This way i can get those changed at the same time i attenuate the tweeters.

And finally, in calculators i use the nominal 8ohm of the drivers, and never doubted it until a friend asked if i measured the resistance of the drivers to make that calculation. I was under the impression that since the resistance of a driver changes over the frequency range, only the nominal resistance is used for calculations?

Thanks in advance for any light you can shed,

Daniel
 
diyAudio Member
Joined 2007
Actually I think you need the actual impedance at the desired cross-over frequency, Voice coil resistance is relatively constant, it increases with temperature but impedance is variable depending on lots of factors.

I think that's why I still prefer older Vifa drivers, the factory published performance graphs were/are close ( on average ) to real life.

I think you will find that in most cases impedance rises as frequency increases, and can be non-linear, that is the impedance (Z) rises more towards the end.

Vance Dickasons :" Loudspeaker Cookbook ": does have a rather involved section on cross-overs that you mat want to read
 
one thing i always wanted to do was to find out what effect a xover type had on sound.

I finally managed to do this with a bss omnidrive, the top one, 2/3 rack unit, not the nasty 1u ones with difficult to navigate menus.

it allowed you to have 2nd, 3rd, and 4th order slopes, with totally variable frequnecy, with LR, bessel and butterworth curves.

what I found was quite startling, and much to my surprise.

there was actually very little difference in tonal colour, one I had found a fairly safe xover freq.

I then varied slopes and types.

i found, perhpas predictably, that lr 4th order was the most 'uncoloured'

and the same thing happened when reducing slope as to changing filter Q or type, ie bessle and butterworth.

the sound became gradually 'sharper' ever so slightly brighter, butterworth was in between and bessel the 'brightest'

ditto when I reduced slope, had a similar effect, slightly more interesting and colour as you reduced

but it was pretty subtle, not massively changed in tonality or colour of instruments, it was a bit of a disappoinment, I thought all sorts of stereo effects and placement would result, that it was part of the secret of differences between sounds of speakers, but alas, no it wasn't the effect I was looking forl, basically to totally alter the sound of a speaker to simulate other speakers.

there is far more to a speaker than xovers, they are in fact from my experiment not critical if passives follow active xperiment.

I would say less than 10%, perhaps even 5% sound change, maybe not even that.

its much more drive unit and enclosure/loading based.

however, perhaps some of the changes in dcr of chokes and capacitor colourations and resistors, too in construction may tonally alter passive xovers, too.

re. caps, it is totally contentious, there's a massive amount of rubbish, myth, with little fact, purely personal opinion, but to a certain extent, quality of dielectric perfection influences it, ie in order,

polyester, polyprops, polystyrene? teflon (polytetrafluroethylene) in plastics is the pecking order, silver mica meant to be best?

but they have to have values suitable for your application.

electrolytics are frowned on, but being a newbe I guess and asking the most pertinent unbiased questions, what I can say is simply try all and decide for yourself.

chokes?

adding a magnetic core means less turns of wire to get the same henries ( analogous to farads in cap terms, but the opposite electrically)

but if the core isn't big enuff when the power is high, it will saturate giving 3rd harmonic distortion ( same as tube amp transformer core saturation) magnetic saturation, the core can't magnetise any more than the capacity.

so its a trade off, an air core will never saturate, but you need more turns, so either DC resistance will rise which is said to be undesirable, or use as thicker wire as you can get.

I never played with that.

a speaker has a number of resistances, basically there are 3 types, DC res and AC res. which are capacitive and inductive

add these together and you get other types of resistances, impedance, which is a combination of AC and DC resistance.

its not clearcut, a LSpeaker coil is basically an inductor, and in any component its not perfect, so you get both inductive resistance due to coil and DC res. due to wire.

as to changing values willy nilly, I have been studying for years and still I can't design a xover properly, and never will be able to like the pros.

they use computer progs. with expensive mics. to measure, the only proper way, otherwise its by ear, textbook equations don't take into account varying resistances like a speaker has, its all over the show, the cabinet also affects the impedance.

basically teh pros get a 'target' curve, linkwitz target should get the drivers in perfect phase with other things into account, driver spacing, baffle, etc., and then use the drivers natural rolloff, and the computer will adjust the values of components to match this target curve and hey presto, perfect xover.
 

I have a question regarding the FAQ from diy audio and video about the 3-way crossover design. It says, after the diagram
An externally hosted image should be here but it was not working when we last tested it.

"Although the diagrams in this document show each the high speakers being run through multiple high pass filters, this is not necessary. In the above diagram, the input for the second and third crossover could be directly tied to the main input instead of the high output from another crossover."

Is this correct or I miss something? If we connect the second crossover directly to input, then the midrange driver will be feed with all low frequencies, since the capacitor from the 1st crossover will no longer block the low frequencies. So the woofer and the midrange will reproduce the low frequencies at the same time. For example woofer from 20 Hz to 150 Hz and midrange from 20 Hz to 1500 Hz.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.