Reading AllenB's guide to making a crossover without measurements, it says to flatten the tweeter impedance by adding a resistor. Just playing with some numbers I calculated that a 10ohm resistor on the 16 ohm 3FE22 would cause the impedance to be around 6.1ohm.
Using this number in a crossover calculator, I get around 3.8mH for the inductor on the tweeter. Is what I'm doing correct? For the woofer I just put 8ohms for now and this gets 5mH. Would the following two inductors be suitable?
5mH - https://www.soundimports.eu/en/jantzen-audio-000-5423.html
3.5mH - https://www.soundimports.eu/en/dayton-audio-ic183-5.html
These aren't the exact inductors I will get as I appreciate the values will probably be different if I'm measuring the drivers in the box. I am just a bit confused as it seems prices vary a lot.
I did read the DCR should be about 1/20 of the drivers nominal impedance and for non air-core indcutors the power rating should be something like double or more of the power rating you need to be safe.
Using this number in a crossover calculator, I get around 3.8mH for the inductor on the tweeter. Is what I'm doing correct? For the woofer I just put 8ohms for now and this gets 5mH. Would the following two inductors be suitable?
5mH - https://www.soundimports.eu/en/jantzen-audio-000-5423.html
3.5mH - https://www.soundimports.eu/en/dayton-audio-ic183-5.html
These aren't the exact inductors I will get as I appreciate the values will probably be different if I'm measuring the drivers in the box. I am just a bit confused as it seems prices vary a lot.
I did read the DCR should be about 1/20 of the drivers nominal impedance and for non air-core indcutors the power rating should be something like double or more of the power rating you need to be safe.
The forte of the Faitals is that they can handle a lot of power for the size, but they need some passive or EQ help as they can be quite bright.
I will mention again SBA for woofers. They have very good drivers at very good prices. Some very good ones half the price of the 12pw.
I will mention again SBA for woofers. They have very good drivers at very good prices. Some very good ones half the price of the 12pw.
I'm not too concerned about the treble at the moment as I think it'll help the dispersion for off-axis listeners. But when it comes to it I'll see how it plays out. I suppose I could try switching out the fullrange drivers as a further project if I'm not too convinced by the sound. I can always use the 3FE22 as little bedroom speakers or something.
I've had a look at the SBA woofers within my budget, the only one I can see is very cheap at around £30 but I can't find any info on it from users or any measurements. Is there any reason you think the HiVi F8 wouldn't work well? I have read quite a few positive reports about this driver and it's price is good for me
If you mean match-up to the 3FE22, it's even worse than the other woofer from a polar response/1st order XO POV.
Me personally? No, because the polar mismatch is too great in a typical home app, often requiring too much room damping, though if your app is such that there's no floor/ceiling/side walls reflections between you and the speakers, then it's a moot point.
To add, I'm not questioning you as I don't believe you, I am just asking as I would like to understand a bit further. I would really like to get the 8" woofer to work, now that I'm married to the idea in my head. I thought the dispersion would be similar up until around 2k so thought it'd be okay
You have to consider the baffle's polar response unless planning to hang the drivers in free air or at least make the baffle(s) just driver width, which impacts its higher frequency BW unless well damped. A way around this to some extent is to choose XO points that match up acoustically driver to driver (regardless of size) per my earlier posts, which in turn ideally starts by finding driver specs that do Vs the usual way, then make elaborate XOs to make them fit as best you can.
Please read these for now to hopefully get a better 'enough' understanding for the more technical docs I can post if need be WRT time/phase/polar response matching (hmm, not sure why I got just posts, I meant the whole threads):
Please read these for now to hopefully get a better 'enough' understanding for the more technical docs I can post if need be WRT time/phase/polar response matching (hmm, not sure why I got just posts, I meant the whole threads):
Regarding your method of choosing XO points, is that when you were talking about VC diameter?
I thought this was specifically referring to if I wanted a 1st order XO at 500hz, or do you follow this rule for any crossover?
I had a read of the thread you shared (there was only one rather than two) but I'm not sure if I am missing something.
Happy to have a look at any technical docs you can share when you get a chance, it might help a bit more as I'm not sure if I've misunderstood some things.
I appreciate all the helpYes. If I follow, yes. Oops! I'm always in a hurry nowadays..........
I'm now following your thinking and the maths you are using.
What sort of math do you apply to the woofer section. So I understand the crossover for the mid is based on VC diameter, but you did mention that the polar response being bad for the HiVi F8 too. I initially assumed you meant because it's 8" and I was initially looking at a first order, so by 3/4khz or so the dispersion would becompletely dropping off. If I went second order I would be well down on dB around 2khz so I thought this would be okay
You're welcome!
The same for woofer to mid and don't forget you can do iterations, so if too low, then take the mean of the 1st point, VC dia. to find the next point, etc..
The same for woofer to mid and don't forget you can do iterations, so if too low, then take the mean of the 1st point, VC dia. to find the next point, etc..
Last edited:
For an 8” woofer that’s inexpensive, fairly efficient and can cross over 1st order at 4-500hz, look no further than this
https://www.madisoundspeakerstore.com/approx-8-woofers/silver-flute-w20rc38-08-8-wool-cone/
The Faital you chose for the fullrange is a great choice considering your budget. Don’t worry about the rising on axis response as this is for HT and likely none of these will be on the listening axis……your assumption about the off axis benefiting is correct. Some minor room treatment on the side walls for the left and right might help at the first reflection points……everything is a tradeoff.
On a side note, the 8” woofer I listed could be used in a 2 way with a tweeter as well……you couldn’t use 1st order filters here but paired with a robust dome tweeter like the Dayton RS28, 2nd order at 1.8khz is easily achievable, ………this speaker would sound and perform MUCH BETTER than the full range/mid tweeter option but requires a little more crossover work.
I’d go for this option personally.
Another option as your first foray into speaker building is just on the modification/carpentry end and that’s to grab some KEF Q150 coax speakers…..they’ve been replaced with the new Meta line and are likely to go on sale this holiday season for $300 per pair. You’ll not find a better point source option or center channel performance than a coax and the KEF is a fantastic small box option. It’s weak point in the cheap enclosure…….it needs additional bracing and some stuffing……fun and easy stuff to experiement with to hear the changes as you go. When close to a subwoofer and crossed at 80hz, they can also play pretty loud and if you added a second sub (highly advise able for HT) you could cross them at 120hz where they’d chase you out of the room.
https://www.madisoundspeakerstore.com/approx-8-woofers/silver-flute-w20rc38-08-8-wool-cone/
The Faital you chose for the fullrange is a great choice considering your budget. Don’t worry about the rising on axis response as this is for HT and likely none of these will be on the listening axis……your assumption about the off axis benefiting is correct. Some minor room treatment on the side walls for the left and right might help at the first reflection points……everything is a tradeoff.
On a side note, the 8” woofer I listed could be used in a 2 way with a tweeter as well……you couldn’t use 1st order filters here but paired with a robust dome tweeter like the Dayton RS28, 2nd order at 1.8khz is easily achievable, ………this speaker would sound and perform MUCH BETTER than the full range/mid tweeter option but requires a little more crossover work.
I’d go for this option personally.
Another option as your first foray into speaker building is just on the modification/carpentry end and that’s to grab some KEF Q150 coax speakers…..they’ve been replaced with the new Meta line and are likely to go on sale this holiday season for $300 per pair. You’ll not find a better point source option or center channel performance than a coax and the KEF is a fantastic small box option. It’s weak point in the cheap enclosure…….it needs additional bracing and some stuffing……fun and easy stuff to experiement with to hear the changes as you go. When close to a subwoofer and crossed at 80hz, they can also play pretty loud and if you added a second sub (highly advise able for HT) you could cross them at 120hz where they’d chase you out of the room.
I would love to see a comparison between the KEF Q150, the TB w6-2313 or the w8-2314 and the SBA SATORI MT19CP-8.
That would be a fun comparison.
That would be a fun comparison.
So for example with the HiVi F8 and 3FE22
HiVi - 35mm voice - (34400/pi/3.5) / 2^4 = 196
3FE22 - 19mm voice coil - (34400/pi/1.9) / 2^4 = 360
Mean xover point = (360 + 196) / 2 = 278
Using the voice coil method you've described, it looks like a lot of two-way designs are out the window (unless I'm missing something here)
Thanks for the suggestion. I've seen this one mentioned a few times, but I think it's a bit of a nightmare getting it to the UK.
I have given up on first order, but I do really want to stick with WAW to help make things a bit easier in terms of matching up dispersion characteristics etc
Regarding the KEF Q150 driver, I already own some R3s so I feel like I'll be directly downgrading. At least if my project sucks then I am downgrading but sideways downgrading 😁
Hmm, not sure I follow.......
HiVi: 36 Hz Fs
35mm VC dia. = (34400/pi/3.5) = 3128.5 Hz
mean = (36*3128.5)^0.5 = 335.6 Hz
3FE22: 110 Hz Fs
3FE22 - 19mm = (34400/pi/1.9) = 5763 Hz
mean = (110*5763)^0.5 = 796.2 Hz a very poor match, so ...........
HiVi: (335.6*3128.5)^0.5 = 1024.7 Hz
3FE: (796.2*5763)^0.5 = 2142.1 Hz
primarily just to show it's near enough the same progression, so the mean between the 1st two is .......
(335.6*796.2)^0.5 = 516.9 Hz which meets your 500 Hz XO point plenty close enough, so no need to go further up the scale, especially since you're more likely to clearly hear its time 'smear'?/'stretch'?, i.e. the goal is PRaT (pace, rhythm & timing)
While I've never done any rigorous testing with point source drivers, these are just too far apart to acoustically sum coherently enough in time/phase for 1st order IME, though inverting the 3FE22 @ ~517 Hz/2nd order maybe OK, but misses your performance goals (and mine) by a 'country mile'.
Re: 2^x, this is for finding the ideal specs to replace either driver or both:
1st/90 = 2^4
2nd/180 = 2^2 (in phase with HF inverted polarity)
3rd/270 = 2^2.828
4th/360 = 2^1 (in phase with a 1.0 Hz offset)
HiVi: 36 Hz Fs
35mm VC dia. = (34400/pi/3.5) = 3128.5 Hz
mean = (36*3128.5)^0.5 = 335.6 Hz
3FE22: 110 Hz Fs
3FE22 - 19mm = (34400/pi/1.9) = 5763 Hz
mean = (110*5763)^0.5 = 796.2 Hz a very poor match, so ...........
HiVi: (335.6*3128.5)^0.5 = 1024.7 Hz
3FE: (796.2*5763)^0.5 = 2142.1 Hz
primarily just to show it's near enough the same progression, so the mean between the 1st two is .......
(335.6*796.2)^0.5 = 516.9 Hz which meets your 500 Hz XO point plenty close enough, so no need to go further up the scale, especially since you're more likely to clearly hear its time 'smear'?/'stretch'?, i.e. the goal is PRaT (pace, rhythm & timing)
While I've never done any rigorous testing with point source drivers, these are just too far apart to acoustically sum coherently enough in time/phase for 1st order IME, though inverting the 3FE22 @ ~517 Hz/2nd order maybe OK, but misses your performance goals (and mine) by a 'country mile'.
Re: 2^x, this is for finding the ideal specs to replace either driver or both:
1st/90 = 2^4
2nd/180 = 2^2 (in phase with HF inverted polarity)
3rd/270 = 2^2.828
4th/360 = 2^1 (in phase with a 1.0 Hz offset)
Okay I see how you're calculating the means now. I'm afraid this has added more confusion in the driver matching process.
Can you sort of break this process down of matching drivers in simple steps? Which calculation should be my go to?
Earlier we were just using VC diameter and using the formula (34400/pi/VC) / 2^4 and you said this is what I should use, even if I'm using a second order XO.
So using the above formula for the HiVi I got 196hz and for the 3FE22 I got 360. I thought this meant that each driver needs to be crossed around this point to maintain true pistonic motion. Therefore, lowering the speaker XO to 200hz would make the drivers match up (I assumed that 360 was the highest possible XO for the 3FE22 so going lower would be okay). Although I'm not saying I'd try to use a 200hz XO, just that I think this would solve the matching issue.
Now, with the mean formula we seem to be getting totally different numbers. I'm not sure what the actual mean value means itself, just how to get to that number so maybe this is causing confusion. But from what you've said it sounds like the means need to be a similar value.
But basically my question is which formula should I use first to check if the drivers match.
Really appreciate some answers to the above, I'd love to get some woofers ordered now but I feel stuck at this point as I'm totally confused!
Can you sort of break this process down of matching drivers in simple steps? Which calculation should be my go to?
Earlier we were just using VC diameter and using the formula (34400/pi/VC) / 2^4 and you said this is what I should use, even if I'm using a second order XO.
So using the above formula for the HiVi I got 196hz and for the 3FE22 I got 360. I thought this meant that each driver needs to be crossed around this point to maintain true pistonic motion. Therefore, lowering the speaker XO to 200hz would make the drivers match up (I assumed that 360 was the highest possible XO for the 3FE22 so going lower would be okay). Although I'm not saying I'd try to use a 200hz XO, just that I think this would solve the matching issue.
Now, with the mean formula we seem to be getting totally different numbers. I'm not sure what the actual mean value means itself, just how to get to that number so maybe this is causing confusion. But from what you've said it sounds like the means need to be a similar value.
But basically my question is which formula should I use first to check if the drivers match.
Really appreciate some answers to the above, I'd love to get some woofers ordered now but I feel stuck at this point as I'm totally confused!