Overpowering very narrow BW, like maybe only an octave or less baffle step correction at > +6 dB.
I understand what you mean, but I think that an assessment with a cell phone recording is simply not possible. Here is the measurement at the listening position:
Understood and personally pay little attention to them beyond whether or not it compares well with a known reference, i.e. the two I posted, but for whatever reason yours literally sounds like a two separate track recording with the pronounced bass 'lit up' in 3D, so to speak, which based on your measurement strongly implies it has something to do with the phone going into a breakup/overload mode or in my signal chain for that matter except that it's still there regardless of my volume, though posting a video using the first 2-3 min. of the Legendary Profile should in theory 'tell the tale'.
Hi...
...well...
...back in May 2023 @Stanislav sent me the idea and a schematics of a series xo. I built it and liked what I heard, but at that time I was also experimenting with my parallel xo, which I finally finished developing and building. But I haven't completely forgotten the serial xo. I've been looking for literature over the last few months and luckily found a lot... a lot to read, but it was worth it. And a week ago the idea crystallized: next round - serial xo. But 3-way... I found older calculation formulas in one of the documents and used them to calculate the approximate values of the components, and then transferred the whole thing to the sim. I had to change some values, of course, but it was done quickly. The woofer is the Eminence Alpha 15A, which I already know well, the mids are covered by the TangBand 2145, and a PT2522-4 open back tweeter plays in the highs. For the first try I choose the crossing frequencies at 150Hz/3.5kHz.
I built the xo, and it wasn't bad, but far from perfect. It has great clarity and stage, but there was some "loudness-effect" - with missing bass... I think, the bass produced in my finished parallel system is somewhat boomy and I think @GM means this in his remarks.
So I changed the low crossing frequency to about 280 Hz and the big inductor from 10 mH to only 4.7 mH. Now I have more bass, but still less than in my parallel version. Clarity, definition, stage are great, the bass is well defined, but the amount and depth of it is less thanI used to..
It's funny, or weird, that changing only the xo has such an effect to the bass and now I have to think and read about how to improve it further.
The problem is (and @Stanislav warned me), you can't change a series xo very easily, like a parallel one, because everything changes... in a parallel xo - when using it in an open baffle - one have to choose a deep low crossing frequency in order to get a good bass ond one has to play with electrical vs. acoustical crossing and the use of a big inductor is a must... but the same inductor doesn't work for the same crossing frequency in the series xo...
Well, it's a long way to go... wish me luck :😀
...well...
...back in May 2023 @Stanislav sent me the idea and a schematics of a series xo. I built it and liked what I heard, but at that time I was also experimenting with my parallel xo, which I finally finished developing and building. But I haven't completely forgotten the serial xo. I've been looking for literature over the last few months and luckily found a lot... a lot to read, but it was worth it. And a week ago the idea crystallized: next round - serial xo. But 3-way... I found older calculation formulas in one of the documents and used them to calculate the approximate values of the components, and then transferred the whole thing to the sim. I had to change some values, of course, but it was done quickly. The woofer is the Eminence Alpha 15A, which I already know well, the mids are covered by the TangBand 2145, and a PT2522-4 open back tweeter plays in the highs. For the first try I choose the crossing frequencies at 150Hz/3.5kHz.
I built the xo, and it wasn't bad, but far from perfect. It has great clarity and stage, but there was some "loudness-effect" - with missing bass... I think, the bass produced in my finished parallel system is somewhat boomy and I think @GM means this in his remarks.
So I changed the low crossing frequency to about 280 Hz and the big inductor from 10 mH to only 4.7 mH. Now I have more bass, but still less than in my parallel version. Clarity, definition, stage are great, the bass is well defined, but the amount and depth of it is less thanI used to..
It's funny, or weird, that changing only the xo has such an effect to the bass and now I have to think and read about how to improve it further.
The problem is (and @Stanislav warned me), you can't change a series xo very easily, like a parallel one, because everything changes... in a parallel xo - when using it in an open baffle - one have to choose a deep low crossing frequency in order to get a good bass ond one has to play with electrical vs. acoustical crossing and the use of a big inductor is a must... but the same inductor doesn't work for the same crossing frequency in the series xo...
Well, it's a long way to go... wish me luck :😀
Good luck, Istvan! It's a fun endeavor.
My OB 3-way was trial and error with parallel crossover, then DSP crossover, then series crossover. They all had their merits. It's far from perfect currently and I'm willing to bet I broke many 'rules' in crossover design... But it's MINE. There is pride in that.
My OB 3-way was trial and error with parallel crossover, then DSP crossover, then series crossover. They all had their merits. It's far from perfect currently and I'm willing to bet I broke many 'rules' in crossover design... But it's MINE. There is pride in that.
Last edited:
When I had a proper OB a long time ago, I had given up on the passive crossover idea. I couldn't get enought (Equalized) bass, and yet, the fullrange had to be tamed a lot (Eminence Beta 15 with Visaton B200).
Yes I know, passive system has its appeal/elegance, but I couldn't get it to do what I wanted, no matter what I did.
Active crossover (Hint: SLOB, NP) solved all the issues. The only cost, beside the active crossover itself, was an additional amp for the Betas, and additional wires (adding to the already cluttered rat's nest)...
Was it worth it? Oh yes...
Yes I know, passive system has its appeal/elegance, but I couldn't get it to do what I wanted, no matter what I did.
Active crossover (Hint: SLOB, NP) solved all the issues. The only cost, beside the active crossover itself, was an additional amp for the Betas, and additional wires (adding to the already cluttered rat's nest)...
Was it worth it? Oh yes...
Many thanks! Yes, it's fun - and the rules are meant to be broken 🙂 as long as the thing sounds good. I'm not a DSP guy, I have something against digitizing a high-quality signal with a small DSP, kneading it and then converting it back again. That's why I'm experimenting with the passive xo. But I may end up with an active xo, at least for the bass. We'll see.Good luck, Istvan! It's a fun endeavor.
My OB 3-way was trial and error with parallel crossover, then DSP crossover, then series crossover. They all had their merits. It's far from perfect currently and I'm willing to bet I broke many 'rules' in crossover design... But it's MINE. There is pride in that.
Just trying to understand you here... are you primarily using a turntable or other device that only outputs an analog signal? Otherwise it's taking a digital signal, "kneading" it in the digital domain, and then doing the D to A conversion, which doesn't seem particularly controversial.
As the cost for good amplification has gone down, and prices on passive components have gone up, my interest in active XOs and full DSPs has shot way up! Especially for OB environments. But we all have our own comfort zones. 🙂
I did not want a DSP, though it would be useful for experimenting with crossover points and slopes, EQ until you get what you like.
Then, build Nelson Pass's active crossover (jfets) with approximatively these slopes/points/EQ.
Another added benefit is that you can use a nice SE amp for the fullrange, while the bass can be driven even via one of these D - class amps 🙂
I use a turntable and an SACD-player, both high-4-digit units with beautiful (analog) sound. So I don't want to convert the analog signal of the turntable. And I don't want to use the "dumb" digital signal of the SACD, becsuse I only get CD-quality thanks to licence restrictions, but even this has better quality than a "cheep" DSP. It simply doesn't have the sound quality. Been there, done that, for me it's a no-go. That's it 🙂
Hi,
so... it's very interesting... after a few attempts I have now built a 2-way series xo... yes, 2-way, with 3-way versions there are major complications... but it will come... sometime.
And now the real reason why I really wanted to change the xo... I wanted to demonstrate my finished speakers to the best of all wives about 3 weeks ago and after just a few seconds she said: that sounds dull, not as good as the things before ... of course I was sunk into the ground. But - somehow I thought the same since I finished the final xo with high-quality components (the components were selected based on hearing tests, etc.). So I took a wrong turn somewhere. The idea of the series xo came at the right time 😀
As I said, after about 3 (failed) attempts I have now built a 1st order series xo - with only 3 parts. Okay... the big 111 uF cap consists of 5 smaller caps, because it sounds better this way, I've had this experience during my countless previous tests... okay, back to the xo, or the whole thing - it sounds great. And the basses are also back, but not the bloated ones, but the properly controlled, musical and very well differentiated basses, for me a big improvement in this area. The mids: beautifully clean, transparent. And the highs as well, even though I've now done without a tweeter. The whole music has a great consistency, maybe because there are no phase issues, being 1st order.
I'm thrilled 😀
Of course I don't have to fine-tune the whole thing, measure, etc., but the direction is right. There will be a demo soon 🙂
so... it's very interesting... after a few attempts I have now built a 2-way series xo... yes, 2-way, with 3-way versions there are major complications... but it will come... sometime.
And now the real reason why I really wanted to change the xo... I wanted to demonstrate my finished speakers to the best of all wives about 3 weeks ago and after just a few seconds she said: that sounds dull, not as good as the things before ... of course I was sunk into the ground. But - somehow I thought the same since I finished the final xo with high-quality components (the components were selected based on hearing tests, etc.). So I took a wrong turn somewhere. The idea of the series xo came at the right time 😀
As I said, after about 3 (failed) attempts I have now built a 1st order series xo - with only 3 parts. Okay... the big 111 uF cap consists of 5 smaller caps, because it sounds better this way, I've had this experience during my countless previous tests... okay, back to the xo, or the whole thing - it sounds great. And the basses are also back, but not the bloated ones, but the properly controlled, musical and very well differentiated basses, for me a big improvement in this area. The mids: beautifully clean, transparent. And the highs as well, even though I've now done without a tweeter. The whole music has a great consistency, maybe because there are no phase issues, being 1st order.
I'm thrilled 😀
Of course I don't have to fine-tune the whole thing, measure, etc., but the direction is right. There will be a demo soon 🙂
Nothing like an honest spouse to pop our little bubbles! 😉
It's also interesting how easy it is to get used to the sound of something, including heading in the wrong direction...the audio equivalent of the frog not realizing the temperature in the pot is slowly increasing.
Okay, well... the whole thing is even worse... Two weeks ago, I bought one (several...) SACDs at the Vienna High End and immediately listened to the first song from one of the CDs. It was okay, but... Today, just now, I listened to the song again... wow, it is accompanied with a jazz broom and it is beautifully defined with the new xo and, above all, recognizable - with the old xo it was just a noise-mush... I'm just shocked 😱 The difference is huge, the old system is - bad... Tomorrow the old xo will be ceremoniously dismantled 🙂 Here is a short demo recording of the song - as I said, fine-tuning is still to come, so please bear with me 😀
I have come to this interesting thread lately. These comments reflect on what I have read, and some more generalized thoughts about OB's.
I want to reinforce a point that was mentioned here early-on by the OP: that he likes to have some significant distance between his full range driver, and his woofer. It looks like his center to center distance might be around 24” or so. Like the OP, I also find that separating woofer/full range can help to preserve the single driver type of presentation. In my experiments, I ended up liking 28” OC.
One thing that seems to be mentioned rarely, is that a multi-way, parallel, passive crossover can be made in two (or more?) sections that can be fed independently. Since bass typically needs more power---especially OB bass---separate stereo amps, with level controls, can be separately wired to each passive section, reducing the need for passive BSC. This can allow the use of a low-power tube/valve amp for the mid/HF, and powerful solid state for the bass,without the need for DSP or electronic crossovers. This is not a new concept, but it does seem particularly useful for DIY OB projects.
About rear tweeters: If the speaker is dipole, the HF should be also. If a mono-pole tweeter is used on the front, then another similar one should go right behind it on the back. A rear tweeter L-pad could be used to allow tailoring of the sound to the room position, reflectiveness, etc. For a full-range,maybe compare the output, front and rear, then compensate the lack of treble on the rear, by putting a small tweeter near the back of the full-range, with an appropriate filter.
About dipole peak at about 2khz (I'm not sure if this has been mentioned): One OB build of mine uses a dedicated midrange, positioned in a narrowing baffle section. It has a peak in the upper mids. Putting absorbent material around the backof the mid-driver calms down the peak. I prefer this to adding components in the form of a notch filter.
I want to reinforce a point that was mentioned here early-on by the OP: that he likes to have some significant distance between his full range driver, and his woofer. It looks like his center to center distance might be around 24” or so. Like the OP, I also find that separating woofer/full range can help to preserve the single driver type of presentation. In my experiments, I ended up liking 28” OC.
One thing that seems to be mentioned rarely, is that a multi-way, parallel, passive crossover can be made in two (or more?) sections that can be fed independently. Since bass typically needs more power---especially OB bass---separate stereo amps, with level controls, can be separately wired to each passive section, reducing the need for passive BSC. This can allow the use of a low-power tube/valve amp for the mid/HF, and powerful solid state for the bass,without the need for DSP or electronic crossovers. This is not a new concept, but it does seem particularly useful for DIY OB projects.
About rear tweeters: If the speaker is dipole, the HF should be also. If a mono-pole tweeter is used on the front, then another similar one should go right behind it on the back. A rear tweeter L-pad could be used to allow tailoring of the sound to the room position, reflectiveness, etc. For a full-range,maybe compare the output, front and rear, then compensate the lack of treble on the rear, by putting a small tweeter near the back of the full-range, with an appropriate filter.
About dipole peak at about 2khz (I'm not sure if this has been mentioned): One OB build of mine uses a dedicated midrange, positioned in a narrowing baffle section. It has a peak in the upper mids. Putting absorbent material around the backof the mid-driver calms down the peak. I prefer this to adding components in the form of a notch filter.
Last edited:
I agree with your assertions, save the first one which I don’t have an opinion other than that most OB systems I have built and listened to, I prefer 3-way designs over the current broad band driver + woofer style.
I would add that your point about amplifiers also points to high damp factor amp for the low end (class d, class ab) and probably a low DF for the full range (valve or ss single ended class a).
Regarding the midrange hump, you might experiment with a partial side behind the midrange (think one side of a U-frame).
Tweeters… I think the best bet is a single dipole tweeter as opposed to dual front / rear firing.
A lot of this is personal preference.
I would add that your point about amplifiers also points to high damp factor amp for the low end (class d, class ab) and probably a low DF for the full range (valve or ss single ended class a).
Regarding the midrange hump, you might experiment with a partial side behind the midrange (think one side of a U-frame).
Tweeters… I think the best bet is a single dipole tweeter as opposed to dual front / rear firing.
A lot of this is personal preference.
I remember when the Ohm Walsh first came out, it was just the 360 radiant cone. At some later point, they added a monopole front firing tweeter, I assume crossed appropriately. I'll guess that addressed the personal preference of many potential customers, hearing trained on speakers with that aspect for so long.
Audio sure follows the PT Barnum addage. After all, it is entertainment. It's probably safe to say all styles and solutions align with personal preference. What's interesting is when you try out anothers and it becomes yours as well.
That is the direction I have gone also. So far, my two completed OB's have been 12” 3way+18”sub and 12”4.5way designs.
My full range+woofer experiment, was a shallow (4”) sealed/stuffed BI-pole, with OB dipole woofer support.The bi-pole had wide range Fostex drivers: 4.5” front, and 6.5”back. Reasons for choosing this design: Power response, frequency balance, excursion limiting, driver availability, and feeling that these drivers sounded “thin” in open baffle implementations. Though it was only a one channel rough prototype, it had a similar deep sound stage, like an open baffle. A good listen, even in mono.
I have noticed some single-side wings; they seem to make sense, to lessen the peak in any one band.