I have just chat with my friend. He asked me about Active and Passive Crossover System. After gave him some brief explanation, I decided to write down about this topic on my blog. I think this is very interesting topic and most of people often ask about it (usually at car audio). I decide to split this topic to more than one article. The first (this one) should discuss about basic Crossover. Please check other links at Audio categories.
So, enjoy it. This is also my first online blog (I have some of them, but offline).
So now, “What is crossover or X-Over or XO?”
Simply to say, a crossover is used to divide the frequency spectrum. It will cross the frequency to certain range, depends on the needs of the designer.
A passive crossover sample from www.norh.com. The basic idea of a all type of passive crossover is capacitors, coils, and resistors. So, it’s a matter of how you design the crossover. Active crossover uses active component like opamp. We will discuss active crossover in different article.
Why does someone invent crossover?
Human’s ear theoritically can listen frequency from 20-20000 Hz. Some people who usually claim him/herself as and audiophile (or a golden ears) usually can listen beyond this normal range. Sadly, there’s no single driver speaker can produce this wide range perfectly. So, it is very common to see a box of speaker uses two or more drivers (one usually a big driver and the rest should be smaller size). By using more than one driver, each driver should produce only at certain frequency (at their optimum frequency). So, each driver has their own “working range”. Who is in charged to divide the frequency? The crossover!So, actually the crossover is invented because we are going to use more than one drivers in a box of speaker. Crossover will divide this frequency (from 20-20000 Hz) into 2 or more segment. Each segment should be send to each driver. This division will refer to x-way. So, if the crossover divides the frequency into two segment, we call it a 2-way configuration. If the division is 3, so we can call it a 3-way configuration and vice versa.
Common “jargon” in crossover is low pass filter, bandpass filter, and high pass filter.
It is easy to know the function from each filter.
* Low pass filter is used to pass the low frequency (and reject the high frequency).
* High pass filter is used to pass the high frequency (and reject the low frequency).
* Band pass filter is used to pass frequency from a certain range, let’s say from 1000 Hz to 5000 Hz. Frequency less than 1000 Hz and more than 5000 Hz will be rejected.
Every crossover has something called “cut-off frequency”. At this frequency, the volume should be -3 dB lower than before, and getting lower lower and lower after this frequency. Cut-off frequency often called F3 (I will let you know why, but later).
Every crossover also has something called “slope” or “order”. Slope will affect the roll-off frequency for crossover. Let’s say your slope is 6 dB. It means that every octave after the cut-off frequency, the volume will be reduced by -6 dB. Beside slope, we are used to use word “order” to inform the slope. First order crossover has 6 dB slope, second order has 12 dB, third order has 18 dB, fourth order has 24 dB, and vice versa.
Confuse? Yeah, I had same feelings when learning crossover at first time. But let me give you some “easy to understand” example.
You have designed a 2-way 1st order crossover (of course it will has 6 dB slope) with cut-off frequency at 3000 Hz. So, what is this all about?
* 2-way crossover means this crossover has low pass filter and high pass filter.
* 1st order crossover means the slope is 6 dB.
* 3000 Hz means the cut-off frequency.
For the high pass filter, the volume at 3000 Hz is 0 dB, at 1500 Hz (first octave) is -6 dB, at 750 Hz (second octave) is -12 dB, etc.
For the low pass filter, the volume at 3000 Hz is 0 dB, at 6000 Hz (first octave) is -6 dB, at 12000 Hz (second octave) is -12 dB, etc.
Actually, at 3000 Hz the both filter is not at 0 dB, but at -3 dB (so this is why we call the cut-off frequency as F3, it means that at this frequency the volume is -3 dB). But when the low and high pass filter is combined (the -3 dB summation of high and low pass filter) will make the frequency pulled back to 0 dB.
Here’s is a sample of low and high pass filter with cut-off frequency at 3000 Hz and 1st order (6 dB slope). The black line refers to the summation of both low and high pass filter. The line starts at 90 dB (both low and high pass filter). At 87 dB (the F3/cut-off), both lines meet, so we call this as Crossover cut-off or F3.
At 1500 Hz (first octave), the high pass filter should be decreased to around 84 dB (hey it’s a 1st order Crossover or 6 dB slope decrease per-octave), at 750 Hz (second octave) should be at 78 dB, etc.
At 6000 Hz (first octave), the low pass filter should be decreased to around 84 dB (hey it’s a 1st order Crossover or 6 dB slope decrease per-octave), at 12000 Hz (second octave) should be at 78 dB, etc.
Active and Passive Crossover
In this world, there are two types of Crossover. One is active crossover, and one is passive crossover. Active crossover uses active components (opamp/transistor), while passive crossover uses passive components (capacitor, coil, resistor). One important thing to be noted is active crossover’s position is before the amplifier, while passive crossover is after amplifier.
So, active crossover will work in line level (low signal). It divides the frequency spectrum then sends it to the amplifier to be amplified. Active crossover’s cut-off frequency also easy to be adjusted (via analog potentiometer or digitally adjusted). This makes active crossover is very popular among latest digital audio equipment.
Passive crossover work in high level signal. They are positioned after the amplifier. So, they will process the high level signal into two or more frequency band. Passive crossover is hard to be adjusted. Their value is based on the component’s value. So, if you want to change the cut-off frequency, you have to change the component’s value. Not and easy job, eh?
Why do people choose active crossover over passive one?
– Easy to adjust the cut-off frequency.
– Active circuit, produce no loss when processing the source signal.
– Almost unlimited flexibility for higher octave filter (more than 24 dB/octave).
Why don’t people choose active crossover over passive one?
– Requires one amplifier per crossover. If you want to have 2 way active crossover (low pass and high pass), then you will need two amplifiers per channel (four amplifiers for stereo). If you want to have 3 way active crossover (low pass, band pass, and high pass), then you will need three amplifiers (six amplifiers for stereo). Why? Because this type of crossover is positioned behind the amplifier. So, each amplifier should amplify their own band sent from each active crossover only (low/mid/high frequency). The principe is: Divide then amplify (so the amount of amplifier needed is as much as the divided band).
– Quite complicated to be built for you who don’t know electronics.
– People believe this type of crossover has more digital sound (too accurate?).
Why do people choose passive crossover over passive one?
– People believe this type of crossover has more analog sound.
– Easy to design (less components used).
– One amplifier per channel is sufficient (two for stereo configuration). Because this type of crossover is positioned after the amplifier, so you only need one amplifier. The crossover will separate the way (2 or 3 way) then sends it to the correct driver. The principe is: Amplify then divide (so you only need one amplifier to amplify the signal, then after that the passive will divide the frequency). Off course you can use one amplifier per band (like active crossover, so called bi-amp), but this is not a mandatory.
Why don’t people choose passive crossover over passive one?
– Passive circuit. The more components used, the more they will “eat” your source signal. Your 100 Watts amplifier will no longer rated at 100 Watts after passing the passive crossover. This should be no problem for high rated amplifier. But how if I use single ended tube amp (with power output rated at 1-5 Watts only)? The best solution is: Use less components (simple design, 1st or 2nd order only), so they will not cut your power in a big margin.
– Hard to adjust the cut-off frequency. You need to alter the component’s value manually (by replacing them one by one).
– Limited flexibility for higher order crossover. Higher order of passive crossover needs higher value of components, and higher value means higher price.
That is the simple Crossover Tutorial. I’ll continue later with more detailed and deep discussion. Hope this doesnt take your brain’s resources too much π
whiteboy
February 15, 2008 05:24thanks, you don’t know how easy you made it for me to grasp the theory!!! what i’m having a problem getting now is———–I know about capacitors and their jobs, it’s the inductors that is giving me fits…It seems to me that putting them connected between the leads they would short out the path!!! Also, i took apart a radio shack speaker someone had thrown away which had the pots for the tweeter and midrange and noticed the mid was wired reverse polarity???///??? I play guitar and used to make sure the PA speakers were wired right and knew how to wire a 4-12 speaker cabinet and when i saw that stereo speaker like that i felt like someone had made a bad mistake.ha! little do i know, huh. if there is anyway you could teach this old a new trick—
i sure would appreciate it, Thanks.
Whiteboy
Jimmy Auw
February 15, 2008 15:01Reverse polarity sometimes used in specific crossover design, let’s say 2nd order.
You have to read the phase for each filter, then you will understand why we should reverse the polarity.
Manuel
May 18, 2008 04:43Jimmy,
You are the man!
thanks,
MJ
Wolf
May 29, 2008 15:31Kudos Jimmy, very straight forward without losing the quality of the content (bit like passive verses active crossovers! :o) I have a question if you have the time?
I have a 1×15 4-Ohm bass cab that I am considering adding a light weight 2khz-20khz 400W bullet tweeter through a Passive High Pass 3Khz Crossover Filter to, to enhance the mid to high slope.
My question is how will adding these components affect the sum cabinet impedance that the amplifier sees?
Much appreciated, Wolf.
Jimmy Auw
May 29, 2008 16:28The amplifier will only see your woofer impedance…
Thank you.
Wolf
May 29, 2008 17:39Thanks sooo… much, I have been trawling the web for several hours trying to find a reference that qualifies this.
So does adding capacitive (series) decoupling to a speaker negates its physical impedance seen by the amplifier?
Many thanks, Wolf.
Jimmy Auw
May 29, 2008 21:34Actually because most likely amplifier will see the heaviest load… tweeter wont consume lot of power compared with your woofer…
Thanks.
Wolf
May 29, 2008 22:30Thanks for helping my fuzzy grey cells d:o)
Marko
September 1, 2008 13:51Well actually, if they are crossed over properly the amp will only recognise one signal going thru if they are the same impedance. Since effectively the signal will be put thru to whichever driver the crossover tells tells it to.Try to match the impedance of your tweeter to your woofer. Although yes a tweeter won’t draw a lot of current anyway. Hope that helps.
kabubi
January 1, 2010 22:30hello Jimmy,
thank you for making the concept more understandable to idiots like me. i have built horns around a tri-amp system and have a Peavey crossover to set the slope/cutoff points.
i want to go straight to the point and ask your recommendation for the slopes and cut-off points to set in my crossover. please please recommend your slope/frequency cutoffs for: 1) highs, 2) midrange and 3) lows.
thank you in advance and i wish you a wonderful 2010 year.
Jimmy Auw
January 4, 2010 22:49Hi,
Happy New Year!
I can’t suggest directly without knowing the driver. The best is to do the measurement first.
Sorry.
Thanks.
RenΓΒ© Skriver
April 25, 2012 04:43Thank you, good guide to make one understand the basics behind crossover designs.
proxytpb.nl
October 14, 2014 09:59X5 would not pick up any viruses in any respect,
which is a good thing if it is accurate, along with a bad thing when it is not accurate.
It runs very smoothly and quietly inside background of the Mac.
While the rally call of Autobot leader Optimus Prime set a dark tone
of an entire generation of toys and cartoons at the center
and late 1980s, finding the original “G1” toys of him and his awesome pals – and his awesome enemies – has become increasingly difficult.
The Pirate Bay
April 7, 2015 21:33Now that’s all fine and dandy whether or not this target cyber criminals, hackers, and true thieves.
The P2P software begins downloading the requested torrent in streams from multiple sources.
When the movies are available around the internet,
several thousands of people will get full access by paying membership fees.
Akbar NP
April 12, 2015 00:20hallo om Jimmy, maaf menghidupkan thread yang telah lama :’)
senang rasanya menghabiskan akhir pekan dengan jalan jalan di rumahnya om Jim π
ada beberapa pertanyaan om Jim, berkaitan dengan crossover
1. Saya berencana membangun crossover pasif 2way 2nd order, so it have 180 degree in phase, advice yang didapat adalah dengan membalikan polaritas yang akan kita tancapkan pada tweeter. kebetulan tweeter (murah) saya mempunyai sensitivitas berlebih sekitar 3db (dibanding woofer), sehingga saya berencana untuk ‘menemplokkan’ sebuah L-pad demi menemukan keselarasan sensitivitas antar tweeter dan woofer #tsaaaah. nah yang ingin saya tanyakan, bagaimana hubungan polaritas dengan L-pad tersebut, apakah bentuk L-pad perlu dibalik ? agak sulit apabila tidak menggunakan skema/gambar sih sebenernya saya coba deskripsikan ya om Jim
misal seharusnya (sebelum adanya L-pad)
kutub +xo menuju kutub -tw (dan sebaliknya).
setelah dibumbui Lpad
kutub +xo masuk ke (+/-Lpadin ????), kutub +Lpadout menuju (+/-tw ????)
perhitungan dan skema L-pad saya buat atas jasa diyaudioandvideo[dot]com
2. kebanyakan orang/commonly, orang akan menetapkan titik crossover dengan cara menentukan titik tengah antara
titik high Freq Res woofer dan titik low FR tweeter. nah kebetulan drivers saya (ACR C610WH + ACR 9KH) punya rentang FR yang cukup lebar. C610WH 56Hz-10KHz dan 9KH 3-20KHz (dengan melihat data dan kurva FRnya)
sehingga apabila mengkuti orang pada umumnya, titik tengah dari FR keduanya di 6.5KHz (cukup tinggi untuk ukuran 2way yang tanpa mid driver) karena saya agak ragu dengan angka yang sebesar itu, saya coba drive keduanya (tanpa box dan crossover) dengan tape Sony jadul milik ayah saya (out kanan woofer, out kiri tweeter). test menggunakan cd David Foster – Best of Me (milik ayah saya juga) song : I Will – David Foster ft. Chris Kirkpatrick. waktu test tidak terlalu lama karena diawal saja saya cukup kaget dan khawatir. skip skip, kesimpulan yang saya dapat adalah vocal wanita terasa lebih jelas, lepas, dan detil di lontarkan oleh sang tweeter.
nah yang saya khawatirkan, apabila ikut mainstream dengan menaruh XO point pada 6500 maka mid terutama vocal akan justru terasa kurang lepas dan tertahan, karena akan direpro oleh woofer.
bagaimana baiknya menurut om Jimmy? apakah XO point tetap di 6500, karena itu nilai mediannya, dan suara vocal juga akan terbantu oleh keluaran suara yang berada di kisaran 6KHz-10KHz yang direpro oleh tweeter. sepengetahuan saya sih, vocal frequency range berada dibawah 6.5KHz bahkan paling hanya disekitar 2KHz, tapi saya tidak tahu apakah reproduksi suara diatas 6.5KHz juga membantu suara vocal, terlebih 2nd order memiliki slope hingga 12db/oktaf, sehingga saya kira output tweeter dibawah xo point tidak akan banyak membantu.
atau dibuat XO point dibawah titik mediannya, misal 4KHz, atau lebih rendah, sehingga mid frequency sepenuhnya dihajar oleh sang tweeter ? dan apa konsekuensinya apabila XO point tidak berada pada titik mediannya ?
sebelumnya terimakasih banyak om Jim, atas segala sharing dan ilmunya di web ini, arigatou π
Jimmy Auw
April 13, 2015 20:53Malam Pak Akbar,
1. Setau saya LPad tidak masalah dipasangnya gimana.
2. Setau saya tidak ada standar median berdasarkan respons seperti itu. Hal ini dikarenakan kita harus tahu juga gimana respons frekuensinya, distorsinya, impedansinya, etc. Jadi memotong berdasarkan median saja menurut saya sih terlalu “polos” hehehe.
Jujur saya desain XO sangat rumit dan bahkan di Indonesia mungkin terhitung jari yang bisa desain dengan benar (dan saya jelas BUKAN salah satunya hehehe).
Kalo cuman buat experiment, rasanya tweeter lokal kebanyakan punya distorsi yang cukup tinggi. Harus dipotong di atas 4K biasanya baru lumayan. Sayangnya ga banyak midbass yg bagus dan bisa naik setinggi itu… Pilihan rumit memang…
Thanks.