Improving your Crossover Design - Z-Offset

[@Toid]

If you do not know how to use PCD: Check out this series here: kzbin.info/www/bejne/fJ6Zpaagbrmjp9U

[@BostonMike68]

The most informative and talented speaker builder on u tube

[@RB216220]

Nice video, I drove myself nuts trying to figure this out on my last speakers

[@jimthvac100]

Very good information I seen that feature in Xsim but was not sure how important it was. Thank you.

[@HowDareYouu]

a video from A to Z of how you should FRD Measure a bookshelf speaker (for crossover design) if you don't have access to million $$$ anechoic chamber or vast empty fields outside using OMNIMIC & Dats would be amazing

[@Toid]

I plan to have a video out soon on how to take measurements in house. I am hoping this will help. You can thank my Patreons, they asked for this video and that one.

[@alexvladescu9136]

@@Toid can't wait for that video too!

[@roelofvankoningsveld7885]

Hello Toid, Greetings from the Netherlands. What should I do with the PHASE graphs in xsim. Where do drivers need to be in phase? Around the crossover frequency? Across as big a frequency range as possible? I am capable of controlling the frequency response and the impedance in any Xsim design I want to build, but as long as I dont fully understand how to plan for 'phase' I don't feel I'm ready for the real world. How about a video on this?

[@kwacker132]

For all those seeking more information see Jeff Bagby's excellent article: How To Use Passive Crossover Designer To Find The Relative Acoustic Offset.

[@viincentfonkert2870]

Took way to long to find this, this was so damn helpfull! Thank you so much man!!!

[@mathieuhache1845]

Is there a risk to the tweeter to obtain a frequency response with the woofer/tweeter wired in series? When getting a tweeter FRD, I’ll start the sweep at high fre to protect it..

[@naibafabdulkobor4301]

They are not to be wired in series (which would have an effect, but a completely different one) but in parallel. Since you will usually do your measurements at very low wattage (0.5 to 1 W) there is usually no danger of blowing your tweeter.

[@DAVIDGREGORYKERR]

There is actually a DSP board that can do the crossover filtering that can drive two output amps that directly drive the loudspeaker and tweeter.

[@ИванПронин-о1ю]

hello. The link to the Passive Crossover Designer file for download is infected with a virus!!! It is impossible to download((( please make a link to a secure file! Thank you very much!

[@bogior896]

you are the best so far

[@PA5cAl1]

Great video, however, I think you can get away without measuring the z-offset by measuring the absolute phase of both drivers from the same position. You can do this via a two-channel measurement and I recommend this instead of a single channel measurement as it takes guess work out off the equation.

[@attainableaudio4533]

@Douglas Blake in my process which is pretty much the same as this, only I do it in one program so no need to import export and starting and stopping programs, once you have the relative measurements and enter the delays... The crossover simulation is pretty much spot on, with phase and frequency and Yes I've done this with a 3 way. But... Whatever gets you the results the easiest for people is all. Good👍

[@LCRLive687]

@@attainableaudio4533 Hi could you please kindly explain your process in this? Because this video is not making much sense to me. Toids video is measuring the speakers hooked in phase but it doesn't mean that the audio arrives at the same time based on the mic position in this video nor arriving at the samw time based on how in physics a tweeter triggers faster and transduces the signal more immediately. And how do you know that you are compensating for your delays right even in single unit measurements?

[@dam4274]

That’s all fine and good but I would need a hands on tutorial. You clicking here and there teaches me nothing. Does the free program come with a manual?

[@Toid]

It does. I'm not sure what you had a problem with. If you're more specific, maybe I can help. I am making the assumption that you know how to get your frd and zma files, I've done videos on this before. I've also done an entire tutorial on how to use the free program passive crossover designer. So maybe you need to watch those? It's so let me know and I can post a link for you.

[@Toid]

Here's the link to the basics of PCD: kzbin.info/www/bejne/fJ6Zpaagbrmjp9U

[@dam4274]

123Toid Thank you sir, I’m in the learning and research stages. I know I want to build a single full range pair first and then a nice 2-way set. I was familiar with the set back, the z-line as it’s called. I had an old pair of Zenith speakers from my brother-in-law who passed. I kept the speakers but trashed the particle board boxes that were nasty. I would have no specs on those but I’d like to make a nice set with them. This PCD would help to get them sounding right.

[@Toid]

@@dam4274 That is awesome that you still have those. PCD is a great program, and it is really awesome that Jeff Bagby gave it away for free. I hope the tutorial helps you restore those zenith speakers. It would be nice to get them back up and running.

[@ROCKNROLLMODS]

Thanks for the video. I apologize if I missed something, but I assume you will be using dsp to time align the drivers, or is there actually a way to handle this task in a passive manner? (Delaying hf signal, etc. via components in the passive xover) careful, what you tell me, may just blow my mind 🎉

[@adamdjd7882]

I was thinking the same thing, how do I actually implement this time offset on the speaker?

[@n.o.b.s.8458]

@@adamdjd7882 I'm still struggling with this concept a bit, but I think I've got an understanding now. I use VituixCAD instead of PCD, and in the options for the drivers, there are fields where you can fill out Z and H offsets. I've played around with these features, and they change the frequency response when inputted. So, my understanding is that before you design your crossover, you input these values and then design the crossover with these variables in mind. Essentially this is figuring out (with more precision) what the crossover will actually do when it's installed in the enclosure.

[@droidzhunterz6861]

@@n.o.b.s.8458 You cannot add delay with a passive crossover. All this method does is give you the option to see the effect an offset driver has on the frequency curve and then you can adjust the crossover component values to get the best looking frequency curve

[@n.o.b.s.8458]

@@droidzhunterz6861 absolutely! I’ve learned a lot more since. It essentially allows you to get a more accurate idea of what the listening window will sound like, and passive crossovers cannot change delay.

[@ИванПронин-о1ю]

The link to download the Passive Crossover Designer file is infected with a virus!!! Please make a secure link to download Passive Crossover Designer!!! Thanks!

[@jimthvac100]

isn't there a way to do this with just Using REW and XSIM without using PCD?

[@Toid]

Yes. I went over in this video. kzbin.info/www/bejne/rqCtm5iBadd-hK8

[@user-oz4kd8tn4m]

Thank you for some great videos! You do a nice job of communicating concepts. This video clearly explains what Z-offset is. But not so much why it matters. So what if the mic is closer to the tweeter than the woofer when creating the .frd files? Why is is this a problem? How does making the adjustment for z-offset improve the function of a passive crossover? Whatever that benefit might be, is it only relevant for measurements (or listening) on axis with the tweeter?

[@Toid]

It allows you to time align the drivers and make them in phase. Without this measurement when you go off axis of your speaker your speakers will not maintain linearity. Since this is a measurement of when sound starts, it would make sense that you have to have the mic in the same position. If you measure the tweeter and then moved the mic 6 inches closer and measure the woofer, you have no point of reference of when the sound left.

[@Nightjar726]

Hey Nick. How do we contact you now since I see the forum isn’t working properly? I would like some consultation work from you if possible Thanks

[@Nightjar726]

So if I want to do this with my MTM tower I position the mic at the center of the tweeter and measure from about a meter away for both the tweeter AND woofer? And I do it without the crossover? ( I can do this since I have an external crossover. Thanks!

[@Toid]

Assuming a 2-way MTM, then you position your microphone at tweeter level. Take a measurement of the tweeter, take a measurement of the woofers, then take a final measurement of the woofers and tweeter wired together in parallel.

[@Nightjar726]

@@Toid ok cool. Do I do this without the crossover attached? I can do this since I have made an external crossover. And then when I do both in parallel I just jumped the top and bottom binding posts correct? Thanks

[@Toid]

@@Nightjar726 you want to do this with the raw drivers and no crossover attached. For the woofers you need to wire them together, either in parallel or series, whatever your final wiring on those woofers is going to be.

[@LCRLive687]

Thanks but for the frequency Frd measurement isn't it dangerous to send a full range sweep to the tweeter? it can blow it

[@Toid]

Dome tweeters and compression are fine. Planar or ribbon tweeter on the other hand, you will want to use a capacitor on them and limit the response.

[@LCRLive687]

@@Toid ok thanks. Or i guess maybe to only set the sweep range from above a certain frequency if using a software which has that option of what hz to start the sweep.? but in addition, I always thought measurements should be from further away. At least with REW. And i think based on 1 meter. And lastly, I wouldn't know how you would be compensating for the frequencies that the room would be imparting onto the measurement.

[@Toid]

@@LCRLive687 if you have an anechoic chamber and care about your actual sensitivity then 3’ is great. But completely unnecessary and can introduce more issues (room interaction) when done in a non anechoic environment. By getting closer you cut down on room interaction, typically closer to 500hz is a clean measurement. However, one of the reasons I like the Omnimic software over REW is the ease of use of getting good accurate full measurements all the way down in frequency due to it’s combined response and ease of use to set your impulse.

[@LCRLive687]

@@Toid for sure. Im just trying to go back and forth between this video and your xsim tutorial. Its just this video that still getting to me. I wasn't sure why you are measuring both tweeter and woofer in parallel as opposed to just using the seperate tweeter and woofer measurements to then compensate the offsets with the individual files overlayed. Why the blend of the two in parallel?

[@Toid]

@@LCRLive687 the xsim tutorial and thus both show why. If you just import both measurements import them into xsim your combined response will not be accurate. That is why we import the combined response in xsim and set the zoffset.

[@migueldajer6726]

how to measure the mechanical offset when you have a pair of woofers and and an array of tweeter just above the two woofers? from center tweeter to top woofer? in between woofers?

[@__-fm5qv]

is there a way to do phase alignment electronically? Currently have little to no space to move drivers in and out (current project is a compact soundbar, so theres not room for physically moving drivers around).

[@Quotenbrtchen]

There is with a Digital Signal Processor (DSP). That allows you to EQ and phase-align different frequency ranges.

[@ericschulze5641]

Just mount your crossover on the back so you can tweak it, without repeatedly removing drivers,

[@Bucefal76]

I think you get it wrong. It is not z offset problem when you do all measurements at tweeter axis in the same point. The problem is with time 0 for all measurements you took. But the way how to adjust the time 0 in this video is correct. I think much better than I usually do by measuring crossover prototype with Twitter set to be 180 deg out of phase and then adjusting time offset to match measurement (null reverse). By the way. It is very hard to fix the time delay wity passive crossover. I guess that you try to say is to adjust the phase. Non of your crossovers on this channel adjust the time. They are just RL phase adjusted. Please familiarize with step measurements. You will find that in your design, when phase is aligned first cycle will come from the tweeter then it transit (in a smooth way I hope) to woofer. But sound will never came from both at the same moment.

[@viincentfonkert2870]

But how do you adjust the time delay? And how can you see when they are both 'in time'?

[@Bucefal76]

@@viincentfonkert2870special box shape or passive delay circuit for tweeter or active crossover.

[@viincentfonkert2870]

@@Bucefal76 thanks! Didn't find a video on passive crossover delay circuit yet but I'll find it! Thansk!

[@qrdaudio]

Hello, when we measure with mic on axis with tweeter, usually the tweeter come earlier than the woofer? Why you ad positive delay on woofer with xsim, rather than negative delay on tweeter? And is it after we correctly set the delay, we keep it then design passive circuit?

[@Toid]

We are stating the sound is coming x inches before the tweeter, so it stays positive. I have never used a negative number on the tweeter, but I wouldn't see why it would make a difference.

[@qrdaudio]

@@Toid ah I see, so + value at woofer, its pretend that we move acoustic center forward, is it correct?

[@Toid]

@@qrdaudio exactly

[@qrdaudio]

@@Toid what happen when we design passive crossover without correcting that delay. Would it be wrong result or else?

[@Toid]

​@@qrdaudio The crossover you design (response you think you will get) will not be accurate. Might be off by a lot. Play around with the program and check how it changes with and without delay.

[@howardstever6676]

Good info. Thank you.

[@jimmbonz]

for a 3way do you put all 3 drivers in parallel at the same time or do you work on pairs? (woof/mid then mid/tweeter)

[@Toid]

You can do it either way. I typically work in pairs.

[@bryanbassett2110]

Does it not damage the tweeter to play a full frequency sweep through it?

[@Toid]

The typical soft dome tweeter isn't typically going to get damaged by something like this. But there are some tweeters that you should protect with a cap.

[@bryanbassett2110]

@@Toid ok gotcha. Thanks!

[@evaldas7346]

What about tweeter? Does it not blow away the tweeter if you measure the same like woofer from 20hz to 20K?

[@noahtheviking1]

it is done at a very low power level

[@barniboy9336]

Will this work with coaxial drivers ? Will the PCD software be able to handle that ?

[@Toid]

Yes it would

[@Toid]

Also if you'd rather use x Sim, I showed how to do it in that program here. kzbin.info/www/bejne/rqCtm5iBadd-hK8

[@markobalatinac7237]

What is up with muddy audio

[@Toid]

I'm not sure. It's not muddy on my phone.

[@markobalatinac7237]

It lacks some 13khz+ detail

[@Toid]

@@markobalatinac7237 Looks like I didn't use the Low Frequency cut-off on the microphone like I usually do.

[@markobalatinac7237]

That must be it

[@attainableaudio4533]

It's funny because I always seem to find that the people with the worst audio (on KZbin) are always people who deal with audio. Lol😳

[@user-qg2uf8jp3h]

What did he say ?

[@crztrn1]

omg more basic . where do i go

[@Toid]

Wear are you talking about?

[@OnePotMeals]

Practice before the video

[@janinapalmer8368]

I learnt absolutely nothing from this .... just mount the tweeter further back so that the speaker's are all time aligned .. forget the XO skullduggery that was not shown here ...

[@Toid]

Even if you want to mount the Tweeter back further, you still have to figure out how far back you need to mount it. It's not like you can just measure it with a tape measure. Which is why you need this video. This video shows you how to put that offset in, so when you design your crossover you can properly align it. if you don't put that delay in, you'll think that you're going to get one response build to crossover and you'll get something completely different.

[@Toid]

This may help you understand the phase: kzbin.info/www/bejne/bXupimqsmMSto5Y I did an entire series on crossover design which I pinned to the top of these comments if you are looking for crossover help.

[@janinapalmer8368]

123Toid ~ I'm sorry to say but nowhere in this video is there any description of how to introduce this mysterious electronic time alignment circuit. What effect would this circuit have on the system group delay ? There IS no better way to time align drivers other than by physical displacement. You don't have to be super precise just align the centre of the driver's magnetic pole plate... that's quite sufficient.. and the group delay won't get compromised!

[@Toid]

@@janinapalmer8368 I'm actually done videos on time alignment and phase in the past which is why this one didn't cover it. I'm definitely going to disagree with you on physical time alignment. You really need that time alignment to be close to exact, you can't just measure with a tape measure. You have to take measurements to get that. But even when you physically move it back you're going to now have issues with diffraction and lobing. those issues will always be present with physical time alignment. It typically creates a lot more issues then you're trying to resolve. There's a reason why a lot of people don't use it. However you can if you want to. That's always a designers choice. Here's a good read on the basics of time alignment both physical and electrical. It'll hopefully help you get the basis of it. en.m.wikipedia.org/wiki/Loudspeaker_time_alignment

[@attainableaudio4533]

How does the saying go... Knowing just enough to be dangerous... J/K but seriously yes of your going to want something to be linear phase then you have to measure. I personally think just getting the relative measurement so you can actually see in the sim what your crossover will be in the real world is worth it alone. And if you can't get what you want from this then you can move things around to get things to "line up" in the sim ie: take some delay from the woofer and see what it looks like, then you can build it in the real world and know what you'll be getting.