Does a DIY Audio Crossover make sense? How passive filters work! || EB#41

[@McMasterProGenius]

FULL BRIDGE RECTIFIER... oh wrong channel :p

[@Rainbow__cookie]

Full bridge audio rectifier

[@caffeinatedinsanity2324]

Full bridge rectumfrier

[@blackturbine]

Beating a dead horse

[@draco5991rep]

@@caffeinatedinsanity2324 Where can I buy such a rectumfrier, Sir?

[@34Gw]

Mr. Boom lol

[@sobertillnoon]

-20dB per decade? Taking 10 years to turn down the volume seems kinda inefficient.

[@kendotek9414]

Lol.

[@SuperGrover]

Never heard decade used with audio before. Since when did octaves go decimal?

[@bassboy2947]

@@SuperGrover Bode plots are usually on the logramithic scale, in powers of 10.

[@BEdmonson85]

@@bassboy2947, exactly

[@henterpeter4806]

@@nyeleskettes Üdvözlet! Ön véletlenül nem a budapesti Radnóti Miklós Gimnáziumban (volt) fizikatanár?

[@tuttocrafting]

I've studied passive filter at school, 6 month of studying just that argument. Great Scott does it a couple of minutes video. I used to draw lot of Bode charts!

[@ElTwOJaY]

Yep, good old Laplace and transfer functions

[@mychevysparkevdidntcatchfi1489]

For stuff like this, one would work with lots of polynomial math than Bode plots. Chebychev was one sadistic bastard!

[@ElTwOJaY]

Haha he was, our professor would take points off if the plot did not show a slope of -20 db/dec

[@hunterjackson802]

Sry i speak peasant what degree was this for?

[@ElTwOJaY]

Electrical Engineering

[@Forkin]

Good one old TESLA capacitor :D

[@prochazkaml]

Just looked at the comment section to see if anyone else noticed XD

[@groowy]

I have some foil capacitors at home along with thick coils and even op amps from Tesla. "Good" old communism times in Czech Republic... XD

[@electrixhusky2379]

Tesla legendární značka :D

[@RegiiiCZ]

True, I love Tesla’s caps and any older Tesla tech =D

[@RegiiiCZ]

@@electrixhusky2379 jo no =D

[@MitsuZer0G]

Will you finish the walkie talkie project?

[@fahadahmed480]

im also looking for it

[@GnuReligion]

Did you see Andres's fix on the opamp? Had to do with unpredictable behavior of a wave as it peaks near the high rail. One resistor.

[@techman2471]

Check out Andrea Spiess' video on modifying the walkie-talkie to actually work.

[@shaunclarke04]

MitsuZer0G check the pinned commeny

[@EgonSorensen]

One small step size for man, one giant (but lightweight) rf-leaper - and lots of videos on KZbin 😁 I think up next content will be on the combination of the walki talkie and an audio crossover for audio output (currently directly driven by the atmel) - giving much better audio quality. HiFi even?

[@davidmarek6141]

0:40 WOOW, amazing, Capacitors made by TESLA from czechoslovakia!

[@dineshpanjwani885]

He's from Germany

[@beacrox_8888]

You can pop back the dustcap with a vacuum cleaner. EDIT:thanks for the likes

[@DasMrOSi]

and if that doesn't work: Use a small, long needle, poke a hole in it, and pry it out (using the same needle)

[@beacrox_8888]

Yeah

[@nvmyutube]

failing that again, toothpick and a very very tiny drop of glue, either hot or superglue, thats up to you and the material of the cap

[@greekgroover3165]

Sticky tape does the job as well as long as it’s not too sticky ;)

[@duroxkilo]

for rigid/metallic dust caps, hot glue + toothpick or lips-sucking techniques work... :}

[@architt4331]

ElectroBoom and GreatScott! Fans?

[@oskarpolanski3152]

Yeah ;P

[@kolbenlp7325]

Jea

[@Vesic]

Is this video recorded with a new Sony camera ? :)

[@greatscottlab]

It is :-)

[@JjMn1000]

Cool!

[@larrybrady5180]

i dont mean to be so offtopic but does anyone know of a trick to get back into an instagram account?? I somehow forgot my account password. I would appreciate any tips you can give me!

[@atticusalexander2101]

@Larry Brady Instablaster ;)

[@larrybrady5180]

@Atticus Alexander I really appreciate your reply. I found the site thru google and I'm in the hacking process atm. Takes quite some time so I will reply here later when my account password hopefully is recovered.

[@Elekron5556]

Tesla capacitors from czechoslovakia 👍

[@skyes9461]

Co ty tu :)

[@rklauco]

Wow, capacitor from Tesla!!! That's historic - originated in country that no longer exists ;)

[@ristomatti]

The developer of the software used on the video seems to have a proper sense of humor. Roughly translated from Finnish the name would be FuckedUpCAD.

[@greatscottlab]

Haha really?

[@ristomatti]

Yes. It's difficult to translate literally but it clearly comes from the spoken word saying "mennä vituiksi". Further "mennä" means "to go" and "vittu" being the Finnish equivalent to "fuck". In this context it's used to describe how something went, not as an action of doing something.

[@aaronvivian2252]

A huge part of the discussion missed when dealing with filtering components of capacitors and inductors is the phase change introduced. The phase change is visible on every one of your input/output examples on the oscilloscope. This is a critical consideration. For example, when a dual component (12dB/octave) crossover is used on both the high- and low-pass sections, the output for each will be 90 degrees out of phase totaling to 180 degrees, and the woofer and tweeter will be out of phase and cancel each other out at the crossover frequency. The correction for this situation would be to connect either the woofer or tweeter opposite to how they would normally be connected.

[@westelaudio943]

True, but not neccessarily. Sometimes the phase shift of the speaker drivers themselves offsets the crossover phase shift.

[@chazincaz]

Equalization is quantization 🧐

[@goodgoat3096]

You could improve the old cross over by simply replacing the capacitors which ALWAYS degrade over time. In order to start from scratch a crossover that gets the best performance out of the drivers or even to select a good set of drivers from scratch, you need the FRD and ZMA files or the ability to measure and derive them from the drivers themselves. Once you have the files, a free cross over design program like PCD (passive cross over designer) or WinPCD can be used to maximize the performance of the selected drivers. The enclosure type and size is also critical and there are several on line box programs. Parts Express provides the FRD and ZMA files for all of their Dayton Audio drivers and also offers several other brands of DIY drivers. The PE forum is a good source of information as is that of the DIY Sound group. The programs take care of all of the complex math.

[@goodgoat3096]

The Impulse Audio and 123Toid youtube channels have videos that go through the cross over design process using the free cross over design programs.

[@ericschulze5641]

Actually, all you need is a bunch of different coils and capacitors, and a ts chart so you know where to start, I've done it many many times, you don't need a computer at all, except to order the parts,

[@truthseeker6649]

the crossover can be complex because of 3 things : 1) to flatten the impedance of speaker drivers so they act more like resistors because otherwise your calculations using that formula won't make sense once you change that dummy load resistor with real speaker driver. 2) how steep are your crossover ? Bessel, Butterworth, LR, 1st or 2nd or 3rd order etc. 3) how well behaved the frequency response of your speaker drivers. hope that helps.

[@claritoresdiano1021]

needed lossless format (WAV/FLAC/ALAC)

[@bfx8185]

Nice:D TESLA rulez :) :) :)

[@pekkasaarinen2902]

That program name though. 😂 I guessed immediately that the author is Finnish.

[@isenhertor]

i went back to see what you where talking about... lol!! he blended that very well so it sounds technical! hahaha

[@sleeptyper]

Positive thinking has always been a finnish virtue. 🤣

[@tomilaukkanen3095]

I caught that also immediately, I laughed my ass off. Genius guy. Maybe he did have a real life example with beta to name the software XD

[@MrCh0o]

Can you share with those who don't know Finnish, please? :D

[@tomilaukkanen3095]

@@MrCh0o Heh, sure the software name is slightly altered, meaning F*cked cad ( same as F'd up) :D

[@xgamerbih]

A little project/challenge for you, Scott: since you love bikes, you could create a video converting a motor (2 stroke/4 stroke) from a carburetor fuel delivery system, to an EFI (Electronic Fuel Injection) one, since it's 2019! Surely not a challenge for somebody like you. Also, it can be done with any microcontroller, Arduino too, since it's all about reading data, and doing actions in the given timing.

[@blackturbine]

So audio crossovers are literally glorified low pass and hi pass filter To be honest i expected some kind of integrated circuit for better response or to remove phase shift Edit: grammar

[@mauriciozavaleta1223]

There are digital crossovers but more kind of software type used in audio production , these are just ok for loud speakers

[@blackturbine]

@@mauriciozavaleta1223 i can respect that

[@blackturbine]

@@armanrgun47 hi

[@cekpi7]

You wont need to deal with phase shift or delay if you only have one speaker.

[@DasMrOSi]

crossovers even introduce phase shift ;-) we are talking about technology from the 50s, passive crossovers haven't changed since, and probably never really will. You really can't do much better with passive components. Active speakers nowadays are something different... you can do a lot of stuff with dsp based systems.

[@startobytes]

Cool Video, but can you make a big ESP8266 Video ?

[@greatscottlab]

I can put it on my to do list

[@startobytes]

@@greatscottlab cool, thanks

[@electron-1979]

How "big"?

[@startobytes]

@@electron-1979 Up to 5 Parts

[@caiarcosbotias1710]

Wow those TESLA capacitors are old! Wasn't it a state run company from communist Chezcoslovaquia?

[@ladaspk970]

Yes, this caps. Are from Czechoslovakia. CAPS. could be from 80s or 90s. Tesla Transistor are better then todays BC547 etc. They have less noise. :)

[@CorgonCorgon]

@@ladaspk970 No, you're wrong. These are Elon Musk's new supercapacitors.

[@ladaspk970]

@@CorgonCorgon :)

[@vitcenek8611]

@@CorgonCorgon Sorry, but American Tesla is garbage. Our Czechoslovak Tesla was founded in 20's. I collect their old stuff like tube radios, tube instruments like osciloscopes and others :D

[@verenigingvandemagogen4548]

@@CorgonCorgon Elon Musk is a fraud.

[@randomcrap763]

You just tested with a 4 ohm load, but you used a resistive load, speakers are an impedance load not resistive.

[@Marlenalynnann]

Ac resistive loads = impedance load

[@verenigingvandemagogen4548]

His speakers have an impedance of 4 ohms. Impedance is the combined result of all resistive, capacitive and inductive components in a circuit. Meaning there's no reactance, just like with resistive loads. However, speakers like electrostats or magnetostats have an steep impedance curve and is not a static impedance set on 4 or 8 ohms. Replacing those with a dummy resistive load won't do the job.

[@wi_zeus6798]

An impedance of 4 Ohm does not necessarily equal a purely resistive component of the same value. Basically it could be any combination of resistive, capacitive and inductive elements, that equals this impedance at a given frequency. Impedance is sort of "AC Ohms", but it is frequency dependant if there's not just a purely resistive component. A 4 Ohm impedance can have a significant phase shift between current and voltage.

[@verenigingvandemagogen4548]

@@wi_zeus6798 Not necessarily indeed. When impedance is given at a specific frequency it can differ on a different frequency of course. But a static figure on itself will stay the same. 4 is 4 and will always be 4, independent of the frequency, since it isn't specified depending on a frequency. The reason that in this case impedance is specified as a static resistance of 4 ohms is because the deviations caused by frequency response are insignificantly small. So basically its the nominal impedance which is specified by manufacturers of most standard dynamic loudspeakers. However, it does not apply to all types of speakers, especially electrostatic speakers often have their impedance curve specified in a graph, since those variations actually are noticeable, mostly because of their high nominal impedance, which needs to be matched with the amplifier by use of an audio transformer.

[@K2teknik.]

He even measured the original speakers impedance with his multi-meter, so why worry about resistive load vs not resistive load ?

[@FreeCircuitLab]

Can you make a video on how to make an Arduino VGA game consol? it's very easy project but there is no proper guide on KZbin. and if you make this I'm sure we will learn something new

[@TinkerManMick]

Now i wish i payed more attention in math 😂

[@TinkerManMick]

@@justgame5508 Wow that does sure go beyond me! I enjoy watching these videos still, even if a lot of it is beyond what i understand - i still always seem to learn something

[@MrAdi2500]

Well, i have ti correct you in one Point: Most speaker companys save Money with the Crossover. They use cheaper components for it. You can tune your Speakers by rebuilding the Crossover with high quality parts and wiring it in Air instead of the PCB. This applies also to high cost Hifi and high end speakers.

[@greatscottlab]

Thanks for the information :-)

[@nosafetyswitch9378]

I once wanted to put an LM35 inside the output filter inductor of my 3D printer to measure its temp. The high frequency switching induced a shitload of noise on the sensor's wires and would display nonsense. Thats where I remembered those RC filters... Needles to say that a low pass with fc around 500Hz did the job and the measurements were crystal clear afterwards. So dont neglect your filters people even in simple tasks!!!

[@KnightsWithoutATable]

Especially when reading a DC signal.

[@wumpusthehunted2628]

Another EE here (but haven't dealt with such filters since school). I can't help but think that the signal should be treated digitally (near-"brick walls rolloffs" can be trivially calculated using 1024 point FFTs (not an option when I learned this). Reasons to not do this include the cost of the DSP (probably just a raspberry pi for this crew, but hopefully you can break out the channels at the source of the audio), plus the need for more amplifying channels (although each can have lower power and be optimized for thier frequency range, like using class-d for the woofer). Passive crossovers are likely better between mid range and tweeters, and just split out the woofer (and subwoofer if needed) digitally (this may even reduce a channel needed. At least that's what I've heard of the phase issues (all rlc filters induce phase issues, but a good DSP method's won't be measurable), the higher frequency phase errors aren't nearly as audible to humans.

[@Henry680]

Yeah DSP is certainly the preferable method, and allows for far more processing than just filtering, but as you say... cost increase and requires a channel per band. An Rpi isn't really suitable for DSP purposes, but a dedicated DSP chip is not expensive, and could be controlled by an Rpi easily if a UI is desired.

[@GabrielMoscardiPauka]

Let’s just keep in mind that when the speaker is connected, the transfer function changes. Vout is not the same with load than it is without load.

[@willwaguespack8192]

wow this video did a better job explaining filters than my EE proffessor

[@Lasseu]

VituixCAD is finnish and means = FuckedCAD

[@donaldfilbert4832]

An outstanding - clear; concise; and comprehensive job covering the material !! Thanks Great Scott !!

[@andresd270]

I can already hear the phase problems

[@architt4331]

JLC PCB is like your new intro, in every video

[@greatscottlab]

Well, they keep the show going.

[@ashergregory2013]

Would you consider building an active crossover that would feed into a dual channel 50w D-class amplifier and BI-AMP a two way speaker (Woofer and Tweeter) and comparing the improvement over a passive crossover driven by a single 100w amp.

[@FreeCircuitLab]

In India we use rc filters at the input section and the filtered signals are separately amplified by multiple amplifier IC s,and for better low frequency response we use op-amp low pass filter at the input of the channel dedicated for subwoofer.then the stereo amplifier consist of three channels (2.1 system)

[@boy887766]

YES ! old czechoslovak TESLA electrolytic capacitors ..... haha

[@SAM-cj4my]

Mr.Scott, I make lot of soldering and assembling circuit board every day and I’ve huge headache when my solder iron get oxidised.Can you give me good solution to fix this issue.

[@Evildandalo]

surgent sabharishh Make sure you have plenty of ventilation on your work desk. A fan with a lot of airflow will help, but a good solder vent to suck all the fumes away would be the best option.

[@SAM-cj4my]

Evildandalo ok I’ll try this next time

[@hmn800]

Hello, I wanted to know how the laptop charger actually works. I have seen most of the chargers having 3 output terminal, one is positive other one ground and the third one is identification pin. What is the use of that third pin? and if we want to design our own laptop charger, how would we do that?

[@murugasenram7749]

do you know about dc12 v tracked metal gear chassics and how to use with audrino is it possible to create crawler

[@murugasenram7749]

you need ro make it

[@waterknot1]

Nice job. Would have also been nice to cover the phase relationship between inductors and capacitors.

[@GalacticSparrow]

I remember struggling with this in engineering school. This video.... um yeah.... doesn't everyone know that? Great video tho. Lots of meat!

[@electronic7979]

Very good video. I Like it 👍

[@TopAnimeRating]

Hmmm one question here, The ooutput of the speaker is pretty much the current that flows trought it right? so doesnt the phase shift of a coil or capacitor change it as the incoming signal is voltage modulated and not current modulated. Or is this change just too small to notice

[@arjunprajapati8352]

Please, Make video on DIY SMPS upto 50A current and explain on which factor output current is depends. Questions :- 1. What effects on Voltage & Current whike varying frequency of primary winding. 2. What happens if I modify 10A SMPS Secondary winding thickness. 3. Why 20A SMPS size is bigger than 5A SMPS if output current is only depends on Secondary winding thickness.

[@publicmail2]

Maybe I missed it, but explaining why you need a crossover vs letting each speaker do its own thing.

[@cocotug0]

to avoid freqs that sound poorly in each speaker.

[@kingSnake35]

Could you make a circuit breaker that stop charging your phone after full.

[@kingSnake35]

By mean completely cut off.

[@areg7182]

@@kingSnake35 why

[@joelrunyan1608]

Your "engineer penmanship" style is fricking awesome dude... I used to do calculations similarly in college... but I've lost it since I don't use it anymore

[@shaanraina3204]

Dudeeee congratulations on 1M Hope you reach 10M soon Love from India ❤️

[@DumbSkippy]

Even numbered crossovers (2nd order, 4th order) change the phase by 180 Degrees. The fix is reversing the tweeter polarity

[@staLkerhu]

Or the midrange driver, if we are speaking about a 3 way speaker.

[@cquiles386]

Would you be willing to make a video about routing traces on a custom PCB?

[@lifeselectronic]

Nice video 😋

[@Adi_vk]

Hey great scott happy friendship day

[@Thecando]

What was the performance of the original circuit unit? Seemed like it used high quality components.

[@SunilAruldas]

Good point!

[@eet2894]

You can always get into a new world of audio and go for active crossovers... :)

[@mauriciozavaleta1223]

These info is great , I would love to see a video of phase line filters for avoiding EMC noise with motors, is my headache working with micro controllers and big motors

[@mychevysparkevdidntcatchfi1489]

Probably can't cover Chebychev polynomial and such in channel like this, but I often wondered if they use Cauer filters with judiciously selected null point.

[@TeslaNick2]

MiniDSP modules are the answer. Forget passive filtering DIRAC is the way. FIR filters rock and the ability to correct phase is essential these days.

[@chrislambe400]

Don't throw out IIR filters just yet. It can do a fantastic job when combined with digital delay between drivers. Also FIR can significantly delay the audio resulting in poor lip sync if you are not careful

[@TeslaNick2]

@@chrislambe400 Indeed. I'm just playing about with an IIR module right now... to run the sub in my wife's van.

[@vitcenek8611]

Where did you get our Czechoslovak Tesla capacitors? :D

[@rajveersingh2056]

In Germany

[@vojtab.6014]

Taky sem se říkal, Tesla , to mi něco říká 😄

[@clonkex]

Just so you know, in English, quotation marks stay at the top of the line on both ends. So like "this".

[@ItsDextrin]

XD

[@jiltomathew6589]

I dont understand anything still i am watching

[@blackwingmaster141]

Does DIY passive audio crossover make sense? YES, but only if you know EXACTLY what you're doing. In any other case, bi-amping speakers is far easier (and sometimes even cheaper!) nowdays

[@harrison00xXx]

Nikola Jambrović bi amping is only cheaper if you are using AVRs with bi amping. If you are using stereo equipment, bi amping is going to be expensive

[@blackwingmaster141]

​ Kokainarienv0gel I was thinking more in direction of making active speakers. Then bi-amping is cheaper, since TDA7293 (and similar ICs) are cheaper than coils needed for crossover. If you're using existing amplifiers then yes, bi-amping is quite a bit more expensive

[@harrison00xXx]

Nikola Jambrović once i had a bi amping setup with a denon 4311 and KEF IQ90 front, but as i got 9 speakers, i couldnt use it anymore since i lost with that 2 channels. Active Speakers are.... idk, i prefer passive speakers, i had enough subwoofers with electronic failures... that i see the benefits of passive speakers at all. this subwoofers were even hi quality ones like KEF and Klipsch. Electronics inside the speakers is always a bad choice, if we think about the fact, that these parts will fail first and the speaker alone is made for „forever“

[@Dudleymiddleton]

Your diagram drawing is very therapeutic! :)

[@kyoudaiken]

The best setup is a class T amplifier for the woofer and a class C one for the mid and high range. The signal is being processed by the amplifiers itself. This brings you the most efficient speaker with the best sound.

[@pixelflow]

The pwm filtering is really useful too. Doing Arduino to CV for music!

[@Stephan.Cardinaels]

Amazing video, possibly you can try to explain the phase as well? 😀

[@verenigingvandemagogen4548]

7:03 I think you confuse damping and gain. -3dB gain means 3 dB damping. The higher the damping, lower the power.

[@addy.is.live1]

This is the best video for sleeping at night lol i literally slept while watching the video

[@josephmojo123]

It's much better to crossover at the small signal level and amplify each frequency band individually More expensive, but less losses and often higher quality

[@verenigingvandemagogen4548]

Why is that better? Amplification directly on the input seems better to me, when it comes to S/N ratio.

[@josephmojo123]

@@verenigingvandemagogen4548 Because there is a huge amount of inefficiency in the passive components, and they are far more expensive, and often not available at high powers due to the size of components required. The example Great Scott uses is particularly low power so his solution is fine, but when you get up to even entry level sound systems for venues, you can be looking at a good few kilowatts

[@verenigingvandemagogen4548]

@@josephmojo123 Ah right, you mean to install the crossover into the line level signal instead of crossing over after the power amplifier? It would only make sense if using stand-alone woofers/tweeters instead of 2 or 3 way full range speakers for example. For example: I have a pair of 2-way full range Yamaha CBR12 speakers which have an integrated crossover. They run at 350 watt true RMS connected to a 2k watt power amplifier. That's not extremely high power, but still very suitable for small business applications. The built-in crossover is passive and works well. So it is possible. But if I would buy a stand-alone crossover, I would use it on the input of two dedicated power amps (one PA to drive fullrange speakers and one PA to drive the woofers for example). I really doubt if topology would make a big difference. I think it has more to do with the fact that integrated devices often are focused on consumer market instead of pro/businesses, making the average integrated devices of less quality than the average stand alone device.

[@chrislambe400]

@@josephmojo123 Exactly good luck making a LR4 LP at 120 Hz good for 300 Watts

[@staLkerhu]

"and often not available at high powers due to the size of components required" Dude, what kind of power level are we talkin' about? Like, PA systems can be demanding, but if we are speaking about simple home stereo systems only for "normal" listening, then proper components are NOT that expensive at all. If you want to build a 2 way speaker with 2nd order filtering, which may not require more, than 2 inductors and 2 capacitors, then a crossover for one box can be cheaper, than only one of your drivers. However, a proper quality 3 way crossover with 2nd order, where you need to go really low with the woofer, and cut the midrange driver from both sides, is a different story for sure... @@josephmojo123

[@IvanToman]

Is this a kind of joke? 20 db/octave for first order? And all other similar blunders, huh. No..... :(

[@ashish4123]

Awesome.... Make home appliances pcb theory as well as practical... Like ...AC, fridge, Washing machine, TV etc...🙏

[@valdovaldomero641]

Videos are GREAT but I honestly hate the ACCENT! I really don't understand what he's saying. Better hire a voiceover freelancer that has American accent like Filipinos for your videos man. Just suggesting, though.

[@34Gw]

Woof.

[@sjbechet1111]

Just built a pair of 3 way floor standing speakers - open source design (SB Acoustics Rinjani) I just bought the drivers and built the cabs myself - didn't bother with the passive crossovers - for just a $100 more bought an active DSP that let's you experiment with this stuff in real time - way better. Now with Parts Express selling really nice amps with on board DSP I can't understand why any home builder would bother with passive cross over any more.

[@marlodilburg257]

Very nice work. Please note, however, at ca. 8 min., 5 sec., your cutoff-frequency formula is inverted; it should be: f sub C is equal to R over 2 pi times L -- NOT 2 pi times L over R.

[@lsudan2670]

2:58 i am alone who saw that #tesla capacitor ?

[@fransdebruijn6763]

passive crossovers are legacy garbage. much better using active crossovers and seperate amplifiers for each transducer. the parameters shift substantially with temperature and throw everything out. also Silicon is cheaper then copper these days.

[@uwezimmermann5427]

another reason not to use simple RC-filters for loudspeakers is that the impedance of the loudspeaker (typically 4-8 ohm, almost real and not complex) has to be accounted for, since it is effectively in parallel with the output component of the filter (parallel with C for the lowpass and parallel with R for the highpass). Such a low resistance in parallel would mean that you would have to use veeeery large capacitors to get their reactance into the same order of magnitude for audio frequencies.

[@peterofenback9263]

The schematic of your crossover is very strange. Although inferior components were used, the developer obviously made a great effort to equalize the frequency response of the speakers used. In particular, the suction circuit parallel to the woofer, which contains a choke with the high value of 4.15 mH, is very unusual.

[@dedskin1]

The more coils and caps and stuff you add to your crossover , the more transients it will eat up . And power , but power today is not a problem , but transients are clarity and you dont want to lose that .

[@nickwallette6201]

So this is better, but it's still not the whole story. :-) Does a DIY audio crossover make sense? No. Not unless you have the following things: - A way to measure the reactive electrical response of the drivers you're going to use. Preferably in-box. (There are purpose-built LCR meters for this.) - A way to measure the mechanical properties of the drivers -- or, preferably and more realistically, the Thiele-Small (TS) parameters from the manufacturer. - A way to measure the distance from the common baffle surface to the middle of each driver's radiating surface. (I.e., how deep is the center of the cone from the front panel of the speaker? This helps adjust the time alignment of the drivers relative to each other.) - A tone generator capable of doing frequency sweeps. - An amplifier that you *KNOW* is safe to use on raw drivers (no DC offset, no power-on pops or thumps.) - A computer with a high-quality audio interface -- not built-in sound, not a Sound Blaster, but a professional-level audio interface with a balanced microphone input, and probably phantom power. - A calibrated measurement microphone that is either close-enough to perfectly flat, or tested and shipped with a corresponding per-unit calibration file that can be used to adjust the input to be close-enough to perfectly flat. (Alternatively to the last two, a calibrated measurement mic with a USB interface.) - Software that can analyze the mic'd sweeps and graph the IN-BOX response. * I can't stress this enough. IN-BOX response, not the response chart of the driver on an infinite-baffle in an anechoic chamber, which is what you'll get from the manufacturer. (Useful, but only for comparison to other potential drivers, or for designing your own enclosure to your specifications.) The box has an enormous effect on response. You can't ignore it and expect to get a usable result. - A room that is either: Large enough that you can keep the unit you're testing a few ft / at least 1M away from any wall or floor or other reflective surface; or is the exact room and location you're tuning the speaker's design for. - CAD software that can take not only the mathematically-ideal values of all these objects, but the _actual measured values,_ and allow you to insert reactive components (parts of your DIY crossover) until the predicted acoustic output *AND PHASE* is aligned at the crossover point(s) -- bearing in mind the effects that the box (which is also reactive), the drivers, and the crossover all have on each other. And then, you build the XO, insert it, and re-test the measured response -- which should be close to, but maybe not exactly, what you calculated. From there, tweak, rinse, repeat. Keep in mind that since no two drivers are perfect acoustic (nor electric) mirror images of each other, it's entirely likely that the order of the filter for one driver is going to be different than the order of the filter for the other driver that you're trying to match together. 2nd-order LP and 3rd-order HP, for e.g., is quite common. You may have to intentionally wire one or the other *with opposite polarity* to get the effective phase aligned properly. If that sounds like too much effort, you shouldn't DIY a crossover. And you definitely shouldn't just slap any old off-the-shelf one in there either. They're married to the drivers, which are married to the box, which ideally is married to the room it'll be used in. (

[@bazrazin1]

recently i built a 2.1 using that 2 dual tpa 3116d2 classd total 4 channels 0f which only 1 is not used i/p grounded, used 1 channel for the sub woofer, did not even make any extra LPF for it for it would require an extra dual voltage supply for the ne5532, just combined the left & right channel with two 56k resistors & input to the 3116d2, i have noticed even in my sony 3cd changer speakers no crossover for the woofer only non polar caps for the tweeter, also in my logitech Z5500 no inductor for the massive 180W rms sub.

[@f3wbs]

I thought he was going to do an actual crossover with someone.

[@fidodidotech2789]

nice vid, everytime my notification pop-up fo your vids i always remember the scientist in back to the future saying "great scott!" hehe

[@ProtoMan137]

Could you *please* show simple schematics? You're saying "We're connecting an impedance". But where, in series, in parallel, plug it into the f** outlet? Same with "We're creating a voltage divider". Where? How? please don't get me wrong I love your content but it's just hard to follow.

[@gregorkluth2857]

It doesn't work like this. As the resistance of the loudspeaker changes depending on the frequency ( sometimes by 100 ohm's or greater) you will have huge dips in the actual frequency responce. Try messuring the speaker with a microfon. You'll see what I mean

[@yamahantx7005]

That original 'complicated' crossover had a Zobel network. It's designed to help present a constant load to the crossover, compensating for the need to account for the equivalent driver circuit in the crossover. Modeling a woofer as an 8 ohm resistor is a bad idea, considering its inductance will turn a 2nd order filter into a poorly tuned 3rd order filter. With a 2nd order network, you'd invert the terminals on one speaker to account for the 180 degree phase shift. But on a 3rd order network, there would be no point to reversing polarity on one speaker. Fun fact: I replaced an off the shelf crossover with my own design a week ago. The difference is that I designed the speaker enclosure, and chose the drivers, but got cheap with the crossover when I built them. I finally have money for a proper crossover. 16 AWG air core inductors and polypropylene caps for a 60W monitor. It helped that I measured the impedance curves for both drivers and knew what circuit I was really dealing with, along with having a 3D printer for printing the inductor cores for winding them.

[@Tallamango2024Pallamango]

Tweeter can sound good if tuned well, but woofer is going to lose more. Using short cooper thick cables, can help.That is why active cabinets sound better

[@on1ytheb3st]

I’m a little confused when using a passive RC filter for audio applications. If making a RC HPF, does the speaker driver become the resistor in the circuit and only a capacitor is needed in series with the power lead? This doesn’t totally make sense because the driver is an inductor and even with a constant ohm rating, it’s actual impedance varies with frequency unlike a resistor. From what I could tell in the video though that’s exactly what he insinuated so I am confused.

[@michaelc.ateoate979]

Crossovers??... We don't need no stinking crossovers..... Full range folded horn...nuff said...

[@CommanderCrash]

@GreatScott! ok maybe you can help me. I have a RPI3 B that I use headless I have a Audio amp like this one in the link shorturl.at/twLR9 I have the audio amp connected to 5v pin on the RPI GPIO and to ground. I have a speaker that is 3 watts. Sound comes out and it plays music but I have this issue with noise from the cpu, network activity and other workings that go on in the rpi3 B Its like a hissing sound almost like a 56k modem sound when its conneted. How do I get rid of this noise. I have limmited sapce in the case and would not want to power the audio amp on a sperate power source I also do not want a ground loop filter I have tryed one and no effect in noise.

[@EasyOne]

good

[@klangfreund5244]

An risistor is not a speaker, that makes it mutch more komplex because the speaker has a resonant frequency. That's a short komment but teorie and practice are Not the same, if you wanna to make a crossover you have to adjust manually with a callebratet microphone. To get a good result you have to combine dis basic filters with a lot of different types of filters. For example there are high shelf, low shelf filters

[@yjbmwsc]

The fact that amplifier boards are dirt cheap nowadays means there is no need to use beefy crossover as the old-school ways... The filtering can be done at the input stage using simple R-C circuits then the input signals are fed to separate drivers. That is, the woofer and the tweeter have their own amplifiers... Much easier to achieve similar results but of course, one has to appreciate how people in the past made the best out of their current technology, and carefully designed circuits often outperform their more recent counterparts...

[@215alessio]

i tought rc filters had more resistance than Rl as low pass filter thats why for low pass they use rl passive instead of rc passives

[@janoschlolmann8822]

Nice video. Is it possible to calculate or measure the frequency response of a speaker ??

[@hennochoi]

Most likely you can't just calculate the frequency response the best you can do is guess for what range they are, if instead you want to know the value and don't have any datasheet of the speaker the only way is to drive the speaker with different frequency and measure it with a microphone ( probably better use a condenser mic) and use some software to see the audio captured by the mic

[@marcosantunes5242]

This is one more good video!

[@MeisterQ]

Even i didnt like the video, ill give you a thumbs up. Because you are doing almost every week nice videos. This wasnt so interessting.. more boring.

[@milesdufourny4813]

The larger inductor should be mounted perpendicular on the board. Never mount inductors the exact same way on the board.

[@Lfunk1983]

I bought a couple Amazon crossovers and soldered them in and test drove which one to keep. This is for old vintage speakers that I restored

[@valentin.s1352]

nice video! i used to learn about bandpass and band stop filters in school. the equations get really crazy with all the complex parts