Why to NOT use Online Crossover Calculators

[@jenniferwhitewolf3784]

In the days before computers, we ran impedance curves manually.. and used 8 digit calculators to do the math.. Then, measured each driver output alone ( B&K Acoustic meter) and adjusted the parts as necessary to establish desired -3 and -6 points, and pass band levels.. then finally running everything together and re-measuring output. Even that was slow and methodical, as outdoor non reflective, non reverberant space was needed for sweeps. Eventually TEF came along and speeded up inside measurements, and then an avalanche of computer aids became available. This IS absolutely correct advice in this video!! I have been doing speakers for over half a century since long before any computer aids .. Never expect much from 'quickie' solutions with no measurements.

[@ImpulseAudioSpeakers]

Jennifer WhiteWolf pinning this comment because it’s true and people think because these calculators have been around along time they must work. Reality is as you described it.

[@jenniferwhitewolf3784]

Impulse Audio Thanks!

[@edjackson4389]

As long as you have the actual impedance readings (I ge mine with DATS V3 and the manufactures frequency graph the online calculators are not the devil and can get you close to where you wanna be. Getting it the rest of the way thru experience, a trained ear, and some trial and error is 90% of the fun man. Most people aren't running a speaker company, they are amusing themselves. As long as it's a hobby there is no harm in using the calculators geez

[@keithmarlowe5569]

I did less than that on first build. I looked at manufacturer graphs, and crossed before cone break up and drop off on the woofer. Then I double checked with Tech Support at Parts Express. Now I'm watching videos on using calculators and spreadsheets, and I'm within 500 Hz. I do wonder if that would make a difference though. Both drivers are flat tween 2,000 and 2,500. I need to rebuild the boxes, so I'm curious if crossing at 2,000 vs 2,500 will matter.

[@150551100]

Hi Dude. Like you i am an avid audiophile and enjoy building stuff. Many years ago,i buid my own speakers using a DynaAudio tweeter and dome midrange plus 2 10 inches Vifa woofers. The box dimentions were calculated using the BASS BOX software,same for the xover.I used open core coils and Solen Capacitors. Xover frequecy cuts are 300hz and 3K,6 db roll off. Ibought all my components at SOLEN and they have a catalog of box desing (still selling) that they gathered all over the world.Mine was one of them. They cost me then about $900.00 per box,but nothing can touch those speakers under $10K. I was a subscriber of a magazine called then Speaker Builder.Over the years i learned a ton of information about speaker buiding.Many readers commented on their experience and things they found out that could improuve the sound,lower distortion,elemninate (or reduce) diffraction distortion. Some may sound strange but apparently,it does something,so all those little modifications put together makes your project sound better. Here it goes: 1) The inside of the speaker hole has to be rouded off.Looks like the cone of a speaker don't like square edges.It mouves better if it has more room to breathe ! 2)The worst box desing would be a square box,its has to be rectangular,it reduce standing waves effects. 2)Never install a speaker flush with your front baffle,it increse diffraction effects. 3)Glue a felt tissu on you front baffle then mount your speakers on top of it,it reduces diffration effects dramatically. Any serious audio entusiast tha wants to buid his speakers can build something outstanding at a fraction of the cost of a comercial unit. In order to do that,they are a few basic rules to observe. 1) Forget about $25 speakers,pick the best there is.A good tweeter between $100 and 200 is a start,same for the midrange. for a woofer,$200 is a start.I would not buid a speaker box without 2 10 inches woofers or at leas 1 12 inches woufer. (2 10 woufers per side or a 12 on each side,will dispense you to buy a subwoufer) 2)Xover parts in total per side should cost you around $200 (Canadian)(Open core coils and high quality paper capacitors.(No electrolitics). 3)The cabinet must be very sturdy,use 1 inche MDF wood with braces internally. 4) Finish or WAF (wife accepting factor) One set of speakers,i used stic on vener wood (it comes in a roll pregluded,you aply it to you box with a ironing iron to melt the glue,let it rest and use an oil base varnish.(I used a dye to match my hard wood floor) Another set of speakers i buid,i wanted a black piano finish.I called a piano manufacturer,asked what paint or paint process they use, They would not tell me that but the guy on the fone said to use oil base varnish,diluted 10%, sand between each cotes and after so many cotes,when you see what you like,you stop.After 5 cotes,i had my piano fnish. 5)Buy the BasBOX and xover sofware.Some parts retaylors have it and some will make the desing for you at no cost. Good luck to you and your followers. Bye. Lucien.

[@bootsarmstrong8421]

Thanks for all the great info. I got a piano finish the easy way. Paint with flat black paint. After it dries, paint on semigloss polyurethane. No sanding needed. It looks like a piano finish.

[@150551100]

@@bootsarmstrong8421 Thank you Mr Armstrong. That is exactly what i have done on my home cinema speakers.Paint rep advised i use floor matt black paint,polyurethane varnish (oil base)I used a fan to make it dry and harden before i could apply a second cote.Raised the room temperature to 80 deg,light sanding (2000 grit) between coats,.four coats was ok.

[@Justwantahover]

If you want a FR driver to sound even better than tweeters, do this: Make two short sticks out of a split toothpick (make extremely thin) and at 3mm lengths. Small. Press down the dust caps (only works with round dus tcaps, not phase plugs). Press down 2/3rds of the dust cap (carefully) with the blunt end of a pen or pencil or anything. Make the indentation neat and evenly round. Then carefully glue the stick onto the center of the depressed dust cap. Now make an aluminium "tizzer". Using coffee can seal foil, cut out 1/8" strips and cut into a diamond shape by cutting them at like 30 degree cuts along the 1/8" strips. The diamond shape should be about 1/2" long side. Then cut them exactly in half so you have two 1/4" isosceles triangles. Or make another if you can't get them the same size first time (and pick out the best two). And now cut the corners off the base of the triangles so you have a teardrop shape foil (two of them). And with a metal skewer or the end of a compass (or anything pricky) make a hole in each of the foils at the big end. This helps to position the foils onto the stick and helps the gluing. Using a toothpick, dab a bit of glue onto the sticks, but let the mounted sticks fully dry before you do this. And using a wet toothpick, you can handle the foils and place them onto the ends of the sticks so the points face the same direction and completely level with the plane of the cone (flat position). Let them dry and hear the BEST TREBLE you ever heard! The cymbals penetrate the bass like real cymbals do, and the treble extends way high and sounds like a tube amp. It's FINICKY! Extremely difficult the first time, but way worth the trouble. If you do this you will NEVER go back to tweeters. Even the cheapest $3 FR drivers sound amazing! Here is a bluetooth speaker comparison, listen on headphones (sorry about the tinny camera sound, the treble isn't really that prominent): kzbin.info/www/bejne/pqO4opJsr8qtesk

[@circlemover]

Cool. Most quick fixes delivered by the web are junk no matter the subject. I like your straight up advice, informative, to-the-point without any fancy video edits.

[@raykdreisatzgehtanders7239]

Those calculators are very useful for input stage crossovers with resistive loads. You only have to match the driver levels to eachother. Things like phase and baffle step rarely come to play in real life situations, you don't listen to a speaker on axis in an anechoic chamber. ElectroBOOM described it well in his video about sound cancellation.

[@ImpulseAudioSpeakers]

Rayk Dreisatz geht anders I’ll agree that they work for resistive loads and will tell you the transfer function. We don’t agree about phase and baffle step. If drivers are out of phase and cause a big hole in the response, it won’t sound as good. We don’t listen in anechoic chambers, but the frequency response in an anechoic environment has been shown over and over to correlate to good sound because the first arrival response dominates our brain’s aural sense.

[@raykdreisatzgehtanders7239]

@@ImpulseAudioSpeakers Of course it's better, but the question is if the advantage always comes into play. I once recapped a crossover in an old 60s speaker, however I forgot the correct polarity of the midtweeter. Turned out that inverting the polarity just shifted the frequency responce hole to another position. The best answer is using the sqauwker for the complete band where the ears are most sensitive. A good wideband driver cut off at ~400 and 6k with 12db/oct BW (inverted polarity) will do the job fine. Baffle step is cancelled or shifted if the speakers are placed next to furniture or in corners.

[@ImpulseAudioSpeakers]

Rayk Dreisatz geht anders sounds like that old speaker may not have been designed well to begin with to be honest. And I disagree with the common advice that proximity to walls fixes baffle step loss. All it does it contribute up to 6dB in some bands and create massive suck outs in other bands. It can sometimes sound a little more full but never correct. Seems you’re suggesting it’s ok to use these calculators if: the drivers are resistive loads, the sound quality doesn’t have to be great, and if the speaker is close to a wall. Do I have that right? I would just do it the proper way using advanced (free to use) software to get closer to what we know is optimum and tweak from there, referring back to the measurements and software to occasionally to make sure you’re still on track. This video was to point out the pitfalls of using these calculators as a lot of newbs aren’t aware of how bad they can be. If you have a handle on how to do it right, no problem.

[@raykdreisatzgehtanders7239]

@@ImpulseAudioSpeakers I was referring to crossovers with parallel load resistors (16-50 ohms) connected to the input stages of the amplifiers in bi- or tri-amped powered speakers. An unusual concept but it works suprisingly well. You just need a headphone amp to drive them. Sound quality is not high end, right, but I am satisfied. Drivers that are resistive loads have yet to be invented, although some tweeters come close. Using crossover calcs or prefabricated crossovers for traditional passive speakers is never very accurate. It's still better than what I've done as a teen: I just copied the 6.8uF/2.2uF network I found in my cheap Pioneer speakers and used it in my DIY ones with a 15" woofer (one of those "funky pup" things). The mid was a 4" sealed back, the tweeter a mylar dome. The cabinet was cheap particle board with a thin hardboard back. Sounded absolutely horrible! My friends were still impressed with the bass, though. 😂 Later on it started to rattle and I took the embarrasing things apart.

[@ImpulseAudioSpeakers]

Rayk Dreisatz geht anders oh yes that can work. I thought you were referring to resistive loads with a zobel or some other form of impedance correction is used. They can get close enough to a resistive load that the calculators will work. I’ve been just as bad at designing speakers in my teen years as well. Some pretty bad ideas!! Oh well, I probably had more fun with the bad idea speakers back then than I do now ;)

[@dissonantiacognitiva7438]

I spend days looking for speakers and mapping their responses Then added them to xsim till I got the combination I wanted with the crossover I wanted It was days work and no physical parts yet. But now I can buy the drivers I want with some reassurance

[@Toid]

Hey you had a bad connection, I don't trust these results ;) In all seriousness, great video!

[@TheDecguy]

Thanks for the video. I agree wholeheartedly. If someone is going to spend the time and effort to build a good box then it only makes sense to take the steps necessary to make them sound good.

[@MrSemperfidelis225]

The free online XO calculators aren't 'crap' but rather the data we punch into them is crap. IF the impedance is dead flat (and measured so the value is correct), and the acoustic response is flat as well, then the calculators should give a decent ball park starting point. The problem is that the impedance and acoustic response is not flat. Also integrating more than one sound source (woofer and tweeter) is not accounted for. But the free online calculators don't know that. What they do is give the values for L and C using the given values of R and freq, nothing wrong or crappy with that. Anyways thanks for informative video, speaker design is way more complex than just smacking great drivers together and doing some awesome woodwork. (Which is why so far I've only managed full range drivers with ok results)

[@ImpulseAudioSpeakers]

Fair enough. They’re math is technically correct. But for our purposes, they’re worthless. Is that a better term?? 🤣

[@rokljhui864]

I found a ported-enclosure design spreadsheet online, and built a 2-way. The crossover is a single capacitor. It sounds, beautiful.

[@jimgardiner1558]

I agree that blindly using crossover calculations is a no no but taking the proper precautions they can still be useful as a starting point. I used woofer impedance compensation to compensate for the woofers impedance increase above resonance, tweeter attenuation compensation to match the driver efficiencies and I even used a notch filter to reduce the tweeter resonance peak that was within one octave of the crossover frequency. Even after that I had a large dip at the crossover point which was likely due to about a 180 deg phase mismatch ( I believe it was due primarily to using a mix of 1st order low pass for the woofer and a 2nd order high pass for the tweeter). The simple fix was to reverse the tweeter polarity. The end result was a nice flat response curve with no visible indication of where the cross-over was. I did do a little tweaking with the resistor in the tweeter attenuation circuit to reduce the brightness of the tweeter.

[@ImpulseAudioSpeakers]

Jim Gardiner what of phase had been off 70 degrees or 40 degrees or 25 degrees? And why go through all that trouble when you could just do it properly and probably have a better XO with fewer parts? Thanks for watching and sharing.

[@jimgardiner1558]

@@ImpulseAudioSpeakers Good point. I often wonder about phase issues and what to do about them. I believe it is especially important in the crossover area where the two drivers interact but what about phase distortion in general? Does it matter if the upper harmonics are not in phase with the fundamental frequencies? Can our ears/brain tell the difference. Certainly using less components in the cross-over circuit is attractive from the point of view of phase distortion as well as cost.

[@ImpulseAudioSpeakers]

Jim Gardiner personally I only care at the XO to make sure both drivers are summing nicely. Other than that I don’t believe we can hear phase. Most research points to this conclusion.

[@jimgardiner1558]

@@ImpulseAudioSpeakers I think I read somewhere that phase distortion can affect the sound stage. Makes it more difficult to pin point the location of instruments etc. Not sure how to minimize this problem. By-the-way apparently the depth of the null apparently is an indicator of the phase relationship. They even recommend doing a reverse polarity check on a crossover to test how well the drivers are in phase (the deeper the null the better the phase match).

[@ImpulseAudioSpeakers]

Jim Gardiner the deeper the null the better the phase match AT that frequency. So it’s not always a good indicator. But a quick and easy test. Also, the difference between a rock bottom null and a -20db null can be as little as a few degrees. So I don’t get hung up on how deep the null is. I’m more interested in how broad the null is. That will mean you have good overlap of the drivers.

[@krismichalsky]

Only dogs/animals are going to hear that spike at 17kHz....

[@kozkoz7776]

I almost fell for these sites until I saw your video thanks for that I will be more vigilant very good job!

[@yvesboutin5604]

Even worse than calculating a crossover is to buy some ready-made at an electronic store or on-line and expect it to work properly on any given speaker. Thank you for clearing up the fog on this matter. P.S. Could you make a video on the importance of quality components for speakers?

[@Depo196]

Cloudy screen copy, please check. we have to control impedance of course. More over, it shouldn't be less, below the allowable, as result of C and L contour

[@roblc11]

Watching this just changed my whole world. I ended up using minidsp 2x4 to get a flat response, but it's expensive. This seems like the way to go.

[@mariusloubeeka5810]

Crossovers - You're Doing It Right! Everyone who builds speakers should print out all the words you say here in bold and hang that over their bed, in their workshop, everywhere! BTW: I also have such a box with crossover parts but I assorted mine by values in zipper bags.

[@ImpulseAudioSpeakers]

Marius Loubeeka I had them organized once! Haha. I gotta organize them again but I don't take them time to do it :o Thanks!

[@juststuff5216]

Problem is, a good crossover design doesn't lead to a good sounding speaker, as even a bad speaker can have a flat response! I feel the same way about free speaker enclosure design software. I found the results to be a mixed bag, even when the graph was telling me otherwise.

[@ImpulseAudioSpeakers]

TransmissionLineDesigns this is true, but a bad crossover design always means a bad sounding speaker. Best to design a good crossover (and that doesnt always mean ruler flat) and listen and redesign and listen over and over until it is right. Every once in a while it is necessary to change a woofer or tweeter because theyre just not good.

[@juststuff5216]

"Best to design a good crossover (and that doesn't always mean ruler flat) and listen and redesign and listen over and over until it is right." Spot on! For those are aren't yet ready to delve into the full mechanics of speaker design, or who simply want one pair to get them started on the path, buying a speaker kit is a great way to experience the advantage of the enthusiast world (as you state in the video). I truly believe that enthusiasts are leaving the commercial market in their wake. Far too many companies are using generic components built to a price with little real consideration, or are charging far too much for a very basic coupling of drivers/crossover/enclosure. There are some very clever people out there (as the above video show), and they don't all come home with a predetermined pay check at the end of the month!!! Love your stuff (channel) :)

[@dannyr6485]

Hi there, good video. You mention you hooked up your impedence rig? What are you using to measure and graph your speaker impedence?

[@linandy1]

Great video ! Thanks for saving us from a big mistake .

[@NicB-Creations]

What I do is take the impedance at the frequency that I want to crossover, put that in an online callculator to get a ballpark crossover. Then input everything in Xsim and tweak from there. So far results sound good but I'm looking to do some measuring in the near future. Currently working on a 1.5way system with 2 full range drivers. 1 driver just acting as baffle step. Speaker building is addictive :)

[@Justwantahover]

If you want a FR driver to sound even better than tweeters, do this: Make two short sticks out of a split toothpick (make extremely thin) and at 3mm lengths. Small. Press down the dust caps (only works with round dus tcaps, not phase plugs). Press down 2/3rds of the dust cap (carefully) with the blunt end of a pen or pencil or anything. Make the indentation neat and evenly round. Then carefully glue the stick onto the center of the depressed dust cap. Now make an aluminium "tizzer". Using coffee can seal foil, cut out 1/8" strips and cut into a diamond shape by cutting them at like 30 degree cuts along the 1/8" strips. The diamond shape should be about 1/2" long side. Then cut them exactly in half so you have two 1/4" isosceles triangles. Or make another if you can't get them the same size first time (and pick out the best two). And now cut the corners off the base of the triangles so you have a teardrop shape foil (two of them). And with a metal skewer or the end of a compass (or anything pricky) make a hole in each of the foils at the big end. This helps to position the foils onto the stick and helps the gluing. Using a toothpick, dab a bit of glue onto the sticks, but let the mounted sticks fully dry before you do this. And using a wet toothpick, you can handle the foils and place them onto the ends of the sticks so the points face the same direction and completely level with the plane of the cone (flat position). Let them dry and hear the BEST TREBLE you ever heard! The cymbals penetrate the bass like real cymbals do, and the treble extends way high and sounds like a tube amp. It's FINICKY! Extremely difficult the first time, but way worth the trouble. I posted this cos you are a FR nut. If you do this you will NEVER go back to tweeters. Even the cheapest $3 FR drivers sound amazing! Here is a bluetooth speaker comparison, listen on headphones (sorry about the tinny camera sound, the treble isn't really that prominent): kzbin.info/www/bejne/pqO4opJsr8qtesk

[@stevegorkowski3246]

So before simulators we had no speakers never crossed over properly? It just took you a bit longer. Oh and the 2 drivers woofer 1500 crossover and the tweeter higher than 1600 . First red flag on a home diy project. You would be amazed what a signal generator can do and a db meter can do. Next you are going to say I can't wind my own inductors.

[@DjRjSolarStar]

If you were to input the impedance of the driver at your specific crossover frequency, would the results be better?

[@ImpulseAudioSpeakers]

DjKinetec better only slightly because impedance changes. A higher quality driver will have less inductance and impedance will change less. But other issues still remain like phase and frequency response. These problem really require measurements and proper XO design be done, every time.

[@Bucefal76]

No because phase is almost never 0 deg for all ferq. sweep when speaker is in housing. So you have to measure impedance, spl an phase. Then you can simulate your crossover with e.g. vituxcad. Additionally the phase need to be correct at listening axis (tweeter center) so other speakers have additional phase shift. Summary only measurement data as we see here plus simulation plus prototype measurement and real life adjustment can give correct crossover.

[@12AX7TFDia]

Impedance isn't a brick wall, as you say. As LF driver's Z rises your HF driver will start coming in, who's Z is going to equal a fairly flat Impedance line. The vast majority of xover calculators avail online are correct.

[@ImpulseAudioSpeakers]

I disagree completely. There are also frequency response variations to consider. And phase. And your assessment of impedance does not work that way. Try it, measure it, you’ll see.

[@12AX7TFDia]

@@ImpulseAudioSpeakers Thanks for your reply. You could input inductive values into a simulator as many driver manufacturers will put this in the driver's specs. Then you would have to measure it again with the speaker in the box because mechanical loading will alter impedance again. Unless you stick a big resistor in parallel with you driver you are always going to see the speaker behave according to impedance, that's honestly not the fault of the calculator.

[@weeverob]

What do you do if you pick a speaker that does not have real data?

[@ImpulseAudioSpeakers]

Have to measure it I guess.

[@greggiorgio1846]

Awesome video! This answered so many questions for me. I have a couple of questions maybe you can help with. The first one is if the resistance of each driver is different at different frequencies is that one of the reasons a speaker may sound great on one amp but not another? Is the difference in the way the amp responds to the different impedances? The other question I have is if I decide to go the DSP crossover route would all of this be a non issue, or is there something I am missing?

[@ped-away-g1396]

because you're not supposed to choose the frequency to calculate the component values. because electronic components have fixed values called preferred number and most calculators don't take that into account so you're not gonna be able to find that strange value component and if you just use something close and call it a day, you're gonna get a little off response. don't be lazy, recalculate it, this time with fixed component values instead of fixed frequency and keep stepping the values around based on preferred number until you get the closest frequency you want. a precalculated table would be a good idea.

[@ImpulseAudioSpeakers]

Ped-Away-G are you referring to when the calculator spits out a value like 7.384 uF and you use a 7.2uF instead? Cause that’s not really the problem. The calculator is missing much much more than just some value rounding. It isn’t considering impedance, frequency response, phase, etc.

[@ped-away-g1396]

@@ImpulseAudioSpeakers what you mentioned is more of a design problem which is not about calculation anymore. btw i wonder why you still use passive crossover, designing one is a nightmare and they're not very flexible too.

[@ImpulseAudioSpeakers]

Ped-Away-G well the point of the video was to say don’t use these online calculators for design because they can’t design crossovers properly. So I’m not sure what you mean my point is a design issue. I use both active and passive. They each have their place. Active is usually a higher performance way of designing a XO but passive is very convenient. The design is about equal in terms of difficulty. Both require measurements, careful use of the drivers, listening, tweaking, etc. Active has a little more flexibility and power but the process of designing the speaker is the same. Most people are interested in passive crossovers so that’s what I generally talk about on this channel.

[@ped-away-g1396]

@@ImpulseAudioSpeakers i mean it's more about filter topology and driver selection which doesn't involve calculation. but you're right, online calculators are useless, no objections.

[@Tbonyandsteak]

pick the resistance at the crossover point, cant simulate all speakers

[@ImpulseAudioSpeakers]

Tbonyandsteak what about an octave above or below? XOers aren’t a brick wall. What about phase? What about frequency response? Why can’t all speakers be measured and simulated? Maybe some are very large and difficult to measure, but generally all speakers can be measured, simulated, tested, etc.

[@Tbonyandsteak]

@@ImpulseAudioSpeakers Not every one have measure equipment. Actually had that problem. Well no matter what crossover you choose, it will always be a problem, with that increasing resistance. Unless you straightening it out. But then again the more filter and resistance you put into it, the lesser transparent it becomes. To me it sounds best at 1. oder filter no adjustment if that can be managed. Since filters stores energy. I could do that since the woofer was a full tone, so I could place the crossover very high and that worked. I did use The MiniDsp to test the different setting and phases, also tried L pad though. And it took lots of listening hours. The result was not using l pad and a crossover at high frequencies.

[@ImpulseAudioSpeakers]

Tbonyandsteak I understand that. In that case I would recommend using drivers that measurement files can be obtained, like the ones I provide. Or build a kit. Or roll the dice and hope for the best. Measurement equipment is pretty cheap these days. I think the biggest problem for people is knowing how to use it. I hope to cover that in my next video, but I’ve been seriously lacking time lately.

[@Tbonyandsteak]

@@ImpulseAudioSpeakers Btw it was quite provoking speaker project, A corner speaker which give 6db plus It is wild beast to tame, it is the same as 3 inches 95 db transmission line, this the first layout. Nothing layed back in the soundstage, it allstraight to your face music. instagram.com/p/B5bOdoNJusN/? Turns out woofer needed to be spaced out 6 mm in order to coorperate with the tweeter

[@polocash11]

What software was this? Please provide links to this info. Also, I have a hard time fond apps for Mac.

[@ImpulseAudioSpeakers]

I use Xsim, SoundEasy, and WT3 (DATS). Finding software for Mac is hard especially for speaker design. I’m not sure what to recommend.

[@faridjamal6837]

Hello Ryan I am trying to build a cross over for Cerwin Vega components; I am using V465C for mid range and treble and Vega V84D for bass. I want to build a quality cross over, can you please offer some guidance?

[@josephanselmo6863]

Hello Ryan. When it comes to inductors. You know that when bobing the indctors, they have start point and an end poitn that should be + and - . When placing them in the circuit board of the crossover, do we have to place them as the way they were bobing, as they are + and -? Would affect the direction of the bobing placed in de circuit board the performance of the croossover and of the speakers as well?

[@ImpulseAudioSpeakers]

Joseph Anselmo no the inductor can be placed either way. Speakers are alternating current devices so the electrical path is in both directions. All parts can be wired in any direction, which is why only non-polarized caps can be used in a cross over. Good luck :)

[@josephanselmo6863]

Thank bro. Now, U know when we we need a non polarize cap but that is dificult to find what we can do is just use two polarized cap placed or conected by their negative pole in order to get rigth value. What happen if we place two caps of diferent values by their negative pole?

[@2548jov]

HI sir im planning to use the software you presented re crossover design but there are speaker (maker) doesn''t provide fmr/zda data especially here in the Republic of the Philippines. Any free software to use to determine fmr/zda of a speaker.

[@ImpulseAudioSpeakers]

jose valentino midminap there is free software to measure the drivers and get the files, but you will need a microphone and impedance jig. You can use the files I make in my test videos if those are drivers you want to use.

[@joppepeelen]

the sites are not garbage , the calculation for the crossovers are perfectly fine, they ask impedance. and with a dynamic driver that is indeed rising. still the calculation of the website is valid. cant blame them. i agree far from ideal but you cant be so hard on sites that gives basic good information about crossovers. i for one use allot of planar magnetic speakers. witch have a flat impedance curve. so all there calculators where spot on.......

[@ImpulseAudioSpeakers]

joppe peelen your planar has flat impedance, frequency response, AND phase?!? Wow. Yes the calculation is mathematically correct, but it’s useless in the real world.

[@ericschulze5641]

Every time I build them, by trying and testing different coils and capacitors, just as these guys do after they screw around with their computer stuff, I have found EVERY time that my end result is the same as the crossover calculators, so he's wrong they work just fine,

[@lachieouthred781]

Hey there. Annoying DIY guy here 🙂 I'm making some speakers from repurposed drivers... I've searched high and low and can't find specs sheets. Where do you start in this scenario?

[@ImpulseAudioSpeakers]

You’ll need to get frequency response measurements of the drivers. If you watch some of my driver test videos you’ll see the best way to do this. Keep the driver far from any boundaries and use REW or HolmImpulse to get the frequency response. It’s a fair bit of work.

[@MasterAudio56]

u have to use soundeasy v24 or Lspcad 6 its a i easy design to crossover and plz upload this software based tutorials , thanks to impulse audio

[@edwardcase]

I use the ERSE calc

[@prime_time2253]

Hey! Awesome video, i liked it a lot and even subbed for future videos:) I do have a question though, I have several 10W, * ohm spekers with range 63-4000 Hz. I have to make the crossovers for them. 1. Can 4 speakers be connected to 1 crossover or do I need 1 crossover per speaker? 2. Can the crossover allow the speakers to play the sounds only between 63-4000 Hz? Or is it only up to 4000Hz and from there the tweeter takes over? I'm sorry for asking, i have been reading articles and watching tutorials on the matter, but can't get a straight answer.

[@westelaudio943]

Using passive crossovers in the bass range is not the way to go. The impedance fluctuates too much and the inductors need to be very big.

[@CorradoVR6z]

Great video. You should talk about how to accurately measure phase next.

[@ImpulseAudioSpeakers]

Corrado VRz I'd like to get into phase eventually. It's a very challenging subject. Also difficult to explain in a video. I may just forget all the theory part and tell people what to do.

[@CorradoVR6z]

Yeah, sometimes it can be tricky to get a clean phase measurement indoor, which is my current situation.

[@Bucefal76]

@@ImpulseAudioSpeakers yes please try this subject, i'm glad to help you with holmimpulse....if you need second opinion...

[@morganoox3838]

What software do you use?

[@ImpulseAudioSpeakers]

I use SoundEasy but I don’t recommend it. Try Room EQ Wizard or HolmImpulse. I much prefer those.

[@Tsinepanos]

Where you get the data from speaker that you cant find on Forums or Popular Site threads ?

[@ImpulseAudioSpeakers]

Kirkinezis Panagiotis the best is to measure it using a microphone and software such as REW or HolmImpulse. You may also trace the frequency and impedance curves provided by the manufacturer, but this is time consuming.

[@josephanselmo6863]

Hello Ryan. I following your advice in not to use online crossover calculators but I still have a doubt. When it comes to building coils of any measure, these pages are or aren't good? For instance, there is a web page that is called "DIY" that offers this possibility. Have u made the test? I'm asking u cause in building my coils. Nothing more to mention; Thanks for ur video and waiting for ur answers.

[@ImpulseAudioSpeakers]

Do you mean physically building your own inductors? If that's what you mean I would wind them using a chart or online calculator and measure them to make sure it worked. Then wind more or take some winds off to get it to measure right. Measure before you cut them to final length in case you need to add more windings.

[@MarlonDeJesusph]

Great video! I am wondering what software and hardware are you using to measure your speaker as shown on your video. :)

[@ImpulseAudioSpeakers]

Marlon De Jesus I'm using a combination of SoundEasy, Holm Impulse, Xsim, and Winisd for all my software needs. As for hardware I use a minidsp UMIK calibrated by Cross Spectrum Labs on a win7 PC. The amplifier is an old cheap Kenwood about 90 watts and fairly accurate. I also use a Dayton WT3 for impedance measurements. I've been debating a "shop tour" video covering all my speaker building tools and measuring gear. I could cover this subject more thoroughly.

[@MarlonDeJesusph]

Impulse Audio thanks! Yeah a shop tour would really be awesome specially for a noob speaker builder like me. :)

[@Bucefal76]

Holmimpulse, dats and vituixcad. Micro from behringer ecm 8000 and audio interface uforia with phantom input.

[@Borisvanderoost]

Thanks for showing me to the world of XSim! Any chance reversing the polarity of the tweeter may help prevent that suckout at the crossover frequency there? I feel like it's a phase issue.

[@ImpulseAudioSpeakers]

Boris de Straatfotograaf yes absolutely it would. But without the measurements how would you know. The calculator didn’t tell me to do that. And what if it was only 90 degrees out of phase, then reversing wouldn’t help. But yes good observation that it would help in this case.

[@Borisvanderoost]

@@ImpulseAudioSpeakers Really surprised that almost no online calculator says anything about reversing the polarity in second-order crossovers although it appears to be not an uncommon thing to do.

[@ImpulseAudioSpeakers]

Boris de Straatfotograaf even if they did, they’d still be useless because they don’t take into account so many other variables.

[@berockization]

Have you bought an iron yet?

[@ImpulseAudioSpeakers]

I’m confused by this. Haven’t watched my own video in a long time. Did I mention I need an iron?

[@berockization]

@@ImpulseAudioSpeakers Just ribbing about your wrinkled shirt. Good info btw... There are a lot of videos about online calculators, and I was about to go down a dark path.

[@ImpulseAudioSpeakers]

@@berockization lol. I was probably fresh out of the work shop.

[@edwardcase]

The problem with crossovers is real-world components aren't going to be exact. Some intuitive changes must be made

[@Bucefal76]

Who gives thumb down???? This is great video!

[@geokrilov]

Those who use and create online calculators.

[@maz4bz592]

Excellent video again, thanks Ryan.

[@josephanselmo6863]

Bro. I was making this mistake. I'm using online crossover calculators to make my owns but I still feel insecure. I don't trust them. They're like robots or something. I'm making a 3-way system; I have the program u use but I feel insecure about what right parameters on the frequency response are in order to pick the right elements. I NEED UR HELP!

[@sugardove6075]

So, As a newbie I purchased speaker boxes and then from Parts-Express bought drivers..Tang Band full range 6.5" woofers. Then the Tech help line guided me to use and purchase Inductors & capacitors.... these are VERY high end and after believing all would be good I reviewed multiple KZbin videos saying that the online crossover software on this and other sites S U C K S... feel like they got there sale at my stupidity and ignorance.

[@jackmontreal1]

call Solen he is expert for that. You pay and he will design pro xover.

[@itamarkas]

Can you recommend on DIY forums? Thanks

[@ImpulseAudioSpeakers]

itoosh hmm, I like diyaudio.com and AVS forum. But I havent spent much time on forums lately because I've been making videos :)

[@itamarkas]

Totally the best forums out there. Keep doing fantastic videos! Thanks.

[@Toid]

By the way, I love the intro! very nice!

[@ImpulseAudioSpeakers]

123Toid Ya its about time I did that. I'll keep refining these little things. I kind of enjoy the video editing as much as the speaker projects. Im still on an uphill learning curve though, not a pro like you yet ;)

[@Toid]

Haha, I wish I was a pro. I'm glad you like video editing. I wish I did!

[@eksine]

You bastard I was just about to use the damn calculator. You better have videos on your channel explaining how to do it right than, or else

[@ImpulseAudioSpeakers]

eksine hahaha. Awesome. I don’t have the best video for that, but working on one that explains the whole process as simply as possible.

[@eksine]

@@ImpulseAudioSpeakers thanks, I'm looking forward to it. I'm watching your other videos in the mean time. keep it up

[@kencohagen4967]

Why is the bass response not accurate. There is a huge difference in the two graphs. Bass response is just as important as mods and hit frequency response. So why don't these charts mean anything in the low frequency range. It's my beliefs, after listening to what other people think are good sounding systems and correlating as much data as possible, sometimes measured, other times just using my ears, that the difference seen here might not have been given the credit it deserves. For example my wife loaned her brother her car years ago when he was working in Washington State and we were in AZ. When we got the car back I was the first one to turn on the stereo. He had cranked up the bass, dialed down the treble all the way and set the faders to give full volume to the driver's side rear speaker. All it did was sound like a single channel cheap ass boom box. No hits, no minds, no detail at all, just mushy overbearing booming mid-bass! So what is the problem with these two graphs? How do we understand the difference in response between the two?

[@ImpulseAudioSpeakers]

ken cohagen I’m not sure I understand you properly. I absolutely agree bass is just as important. Measurements taken this way are not accurate down low due to the sample rate of the measurements system so in this case everything below 200hz or so should be ignored. And the big gap above 200hz could be a matter of the XO and how the driver responds to it. I did this a while ago so I can’t recall all the details. But yes absolutely bass is just as important as mids and highs. No disagreement there. So don’t use junk calculators, do the XO properly.

[@BostonMike68]

I don't know I just designed and built my own speakers and they sound amazing for under $200 so maybe I got lucky and it's not just me that think sound amazing

[@utubecomment21]

Crossover design is as personal as listening to the speakers themselves. What one person will find good, another will scoff at. Example: even brands like Wilson Audio speakers and their costly speakers come under scrutiny - www.diyaudio.com/forums/multi-way/69687-wilson-audio-aredoing-18.html Graphs are good, but I listen to speakers with my ears, not graphs.

[@ImpulseAudioSpeakers]

utubecomment21 I listen to my speakers with my ears also. I use graphs to assist me with design. It's been scientifically verified that people tend to prefer speakers with certain measured properties such as flat on axis response. To reject this and suggest it has nothing to do with how good a speaker sounds wouldn't be fair.

[@utubecomment21]

Nothing has been scientifically verified.... such studies have been published and pushed by certain brands/manufacturers but rarely does any such studies cover every base, and there are obvious flaws and unasked questions let alone unanswered questions. Example: kzbin.info/www/bejne/fnPIdZirYs9mntE And I've yet to read or see one report that is the be-all and end-all of acoustics!! As for your example of 'flat response' being preferable by the masses, in my 40+ years I've heard many (for example) flat response audio products in many arena's, home, car, studio, personal audio and flat response is far from being the I Ching of end user acoustics. For some reason, some people aim for flat response, but as there is no one solution that covers every end user requirement, so there is no one response curve that is going to satisfy the masses. To suggest so ... and i know there are some engineers working for certain manufacturers that push this view, ... is a nonsense. I know the likes of Sean Olive and Flyod Toole seem credible, but again I've yet to meet anyone (including myself) who has the whole picture and has covered all bases - kzbin.info/www/bejne/a2fbgYCnj9KYoLc kzbin.info?search_query=floyd+toole As Peter Thomas illuded to, every engineer has a view, but there is no one path I'm afraid. One has to pick 'a' path and take any design and factor in advantages and also the disadvantage that comes with every decision. There is no free lunch with any design, and in every design there is going to be some sort of compromise whether one is aware if it or not! kzbin.info/www/bejne/jquXZnyPj7aAnJI Loving your videos though.

[@ImpulseAudioSpeakers]

I'm glad you're enjoying the videos. In any case, I think we'll need to agree to disagree here. Stick around and hopefully a future video will change your mind as I get more and more into things.

[@jimmypete6444]

Excellent channel!!!

[@correia707]

Wow, very cool information!!!

[@giddlegump]

I always take advice from a guy who has a huge box of junk parts generated by failed projects.

[@ImpulseAudioSpeakers]

Says I’m not going to accept mediocrity 😁

[@victorjohnson7512]

I'm using them just to piss you off...

[@ImpulseAudioSpeakers]

Lol that’s fine 😄 I use them too.

[@sanchezramirez4280]

It shows that "youtbe, google expert" are no real experts. It seems them their are experts, because they even don't know what and how much they don't know.

[@ImpulseAudioSpeakers]

Sanchez Ramirez referring to me or people who use the calculators?

[@avesbilal]

Thanks for the great video, if possible for you could you please make a detailed video/s on designing crossover using XSIM btw I have already watched your XSIM video but it was brief.

[@ImpulseAudioSpeakers]

Aves Bilal Thanks for the feedback. Yes I will do that. Its a very big subject but I will try to bring everything together in an upcoming project. Hopefully it can make GOOD crossover design possible for newcomers to this hobby :)

[@MaxNakfoor]

That’s what zobel circuits are for…

[@ImpulseAudioSpeakers]

But why use a zobel when you can just design it properly and avoid that. Also, a zobel only fixes the impedance. You still have the issue of frequency response and phase.

[@MaxNakfoor]

@@ImpulseAudioSpeakers well like you said the impedance rises with frequency, usually you want a nice smooth filter slope. Kind of hard to do if you have say 8ohms at an xover point of 2.5k and 12ohms at 3k. Also confused why you showed the plot with the speakers out of phase, should have tried reversing polarity of the tweeter to see if you could have gotten the sound waves to cooperate. I do agree the xover point needs to be based on actual frequency response and not just the rated range.

[@ImpulseAudioSpeakers]

@@MaxNakfoor I showed them out of phase because that’s what the calculator outputs. Only with measurements could I see the drivers were out of phase. So if someone uses the calculator blindly they will likely have bad phase alignment.

[@MaxNakfoor]

@@ImpulseAudioSpeakers well sure but that’s easy to test by ear, just play a tone at the crossover point and check which way outputs the loudest

[@ImpulseAudioSpeakers]

@@MaxNakfoor harder than you think. Do the test and hear for yourself. Plus, what if it’s not 180 degrees. Maybe it’s 30 degrees and will be only slight.