I have installed the drivers in one of the cabinets for testing. I will do near field, horizontal polar far field, impedance, burst decay, and distortion testing. This data will be used in simulations to design the filtering.
The paint came out pretty nice, but it has a little more gloss than I was expecting. I used semi-gloss. If I did it again I would go ahead and use a sanding sealer, and use a lower gloss paint such as satin or egg-shell.
My wife said "it's kinda blue... maybe you should call it Kind of Blue (wink wink)"... and I thought that was humorous, but then I realized how pretentious it would be to name my humble speaker after one of the greatest albums of the 20th century...
j.
The paint came out pretty nice, but it has a little more gloss than I was expecting. I used semi-gloss. If I did it again I would go ahead and use a sanding sealer, and use a lower gloss paint such as satin or egg-shell.
My wife said "it's kinda blue... maybe you should call it Kind of Blue (wink wink)"... and I thought that was humorous, but then I realized how pretentious it would be to name my humble speaker after one of the greatest albums of the 20th century...
j.
Nice!
And then there's Blue by Joni Mitchell, Blue Train by John Coltrane, Blue Lines by Massive Attack and The Blue Album by Weezer.
I, for one, welcome our AI overlords:
But my vote is for Forever Blue, by Chris Isaak
And then there's Blue by Joni Mitchell, Blue Train by John Coltrane, Blue Lines by Massive Attack and The Blue Album by Weezer.
I, for one, welcome our AI overlords:
- “Mr. Blue Sky” by Electric Light Orchestra 1: A cheerful and upbeat song that was written by Jeff Lynne while he was on a retreat in the Swiss Alps. The song is about the joy of seeing the sun after a long period of overcast weather.
- “Blue Bayou” by Roy Orbison 1: A soulful song about a person who longs to return home to their beloved Blue Bayou.
- “Blue Suede Shoes” by Elvis Presley 1: One of the first rockabilly songs ever recorded, this song warns people not to step on the singer’s blue suede shoes.
- “Forever In Blue Jeans” by Neil Diamond 1: A love song about how all the money in the world doesn’t matter if you can be in the comfort of your lover’s arms.
- “Blue Moon” by The Marcels 1: A classic song that has been covered by many different artists. The song is about a solitary figure wishing upon the moon for someone to love.
- “Behind Blue Eyes” by The Who 1: A song about a person who is misunderstood and feels like nobody knows what they are feeling.
- “Tangled Up In Blue” by Bob Dylan 1: A song about a person who is trying to make sense of a failed relationship.
- “Blue Monday” by New Order 1: A song about the feeling of dread that comes with the start of a new week.
- “Blue Eyes” by Elton John 1: A song about a person who is in love with someone with blue eyes.
- “Blue” by Beyoncé 1: A song about a person who is trying to move on from a past relationship.
- “Blue Ain’t Your Color” by Keith Urban 2: A song about a man who sees a lonesome lady at a bar and wishes to comfort her.
- “Blue Velvet”, Blue Jeans, Blue Banisters by Lana Del Rey 2: A song about a woman who is in love with a man who is bad for her.
- “Blue Eyes Crying in the Rain” by Willie Nelson 2: A song about a person who is heartbroken and crying over a lost love.
- “True Blue” by Madonna 3: A song about a person who is in love with someone who is always there for them.
But my vote is for Forever Blue, by Chris Isaak
Last edited:
Rhapsody in blue, Gershwin
Symphony in blue, Kate Bush
Blue Rondo a la Turk, Al Jarreau
For you blues, the Beatles
Blue Jean, Bowie
Cool Blue, Eurythmics
Famous Blue Raincoat, Leonard Cohen
Deacon Blues, Steely Dan
Blue Hotel, Chris Isaak
just some blue songs from my library…
but cheers to this thread, did I already say so?
Symphony in blue, Kate Bush
Blue Rondo a la Turk, Al Jarreau
For you blues, the Beatles
Blue Jean, Bowie
Cool Blue, Eurythmics
Famous Blue Raincoat, Leonard Cohen
Deacon Blues, Steely Dan
Blue Hotel, Chris Isaak
just some blue songs from my library…
but cheers to this thread, did I already say so?
and of course "Blue" by Leann Rimes... and "The Blue Danube" by Johann Strauss
Any of these are more poetic than LCCAM-10.3 .... which is the unofficial name I have assigned to the directory structure for all the folders, simulation files, and measurement data this project has generated...
I am setting up my rotating turntable as I write this... j.
Any of these are more poetic than LCCAM-10.3 .... which is the unofficial name I have assigned to the directory structure for all the folders, simulation files, and measurement data this project has generated...
I am setting up my rotating turntable as I write this... j.
The chamfers and the colour (color!) do make it look discretely compact, very well done! I look forward to seeing the measurements, and that they meet your bar for bass production! 
I must check at what level mine max out at low frequency - not something I have previously attempted to measure - but won't be 99db at 40Hz for sure.
I imagine that once you've worked out the filters these will offer a great deal of listening pleasure.
I must check at what level mine max out at low frequency - not something I have previously attempted to measure - but won't be 99db at 40Hz for sure.
I imagine that once you've worked out the filters these will offer a great deal of listening pleasure.
Well, I have some doubts about achieving the design F3 and F6. The spec for the RS270-4 is Fs = 21 Hz. With 6 dB of gain at 45 Hz, I get a response F3 of 38 Hz, and F6 of 29 Hz.
I measured an Fs of 27 Hz, which is 30% higher than spec. This affects the native (non-eq'd) response of this sealed-box system. In theory, it should not matter: With an Sd of 346 cm^2 and an Xmax of 6mm, the driver is capable of 99 dB at 40 Hz, and it should be able to be Eq'd to meet the same F3 and F6 target. However it will require more aggressive Eq. We shall see how it works out.
j.
It is the 2-pi response, as calculated by standard TS methods. I use the "enclosure" menu in VituixCad.
It is hard to say how much boundary reinforcement we get in an average room at 40 Hz with a wavelength of 8.5 m. Certainly we get more than 4-pi given the floor + ceiling + wall. In my room I get more than 2-pi because one woofer cabinet is in a corner.
j.
It is hard to say how much boundary reinforcement we get in an average room at 40 Hz with a wavelength of 8.5 m. Certainly we get more than 4-pi given the floor + ceiling + wall. In my room I get more than 2-pi because one woofer cabinet is in a corner.
j.
@hifijim, I am really interested in your process of measuring and integrating the driver responses. I'm currently stumped on my own Hybrid XO project which I've posted on another online community. I'll be following your progress closely! My speakers are not that similar to your Blue Mondays (wink), but I'm still hoping that the work you're doing will help me find a way to get the best response from my existing speakers. Thank you for being so generous with your knowledge and expertise!
Sixto.
Sixto.
I was planning to dedicate some time to explaining how I manage both the active and passive filtering. Hopefully my process and explanation will make sense to you and help you move forward.
Making accurate and repeatable measurements is one of the most challenging aspects of speaker design. So don't be discouraged if it seems hard.
Making accurate and repeatable measurements is one of the most challenging aspects of speaker design. So don't be discouraged if it seems hard.
Initial simulation results
I made near field scans of each driver, and horizontal polar far field scans, and merged them together in the standard way. I will go into more details in future posts, but I wanted to start with the good news… It appears so far that my performance goals will be achieved.
The initial active/passive filter is just that, initial… I plan to refine it over the next several weeks in simulation before I order crossover parts. But this is what I have come up with so far. The first filter is flat on-axis, and the second has a smoother PIR response. The second filter is also set up to have no gain above 0 dB, as @b_force advised.
Second filter optimized for PIR
I made near field scans of each driver, and horizontal polar far field scans, and merged them together in the standard way. I will go into more details in future posts, but I wanted to start with the good news… It appears so far that my performance goals will be achieved.
The initial active/passive filter is just that, initial… I plan to refine it over the next several weeks in simulation before I order crossover parts. But this is what I have come up with so far. The first filter is flat on-axis, and the second has a smoother PIR response. The second filter is also set up to have no gain above 0 dB, as @b_force advised.
Second filter optimized for PIR
Without listening, I’d lean forwards the 2nd, and agree with @fluid’s assessment.
But without betting, could you please please show a CTA2034 with a 30dB scale (equivalent to difference between 0.1W to 100W), or a non-normalised polar map with a scale of +3dB to -18dB?
The differences will be most obvious, I would think.
But without betting, could you please please show a CTA2034 with a 30dB scale (equivalent to difference between 0.1W to 100W), or a non-normalised polar map with a scale of +3dB to -18dB?
The differences will be most obvious, I would think.
BTW, what is your thought behind a smooth on-axis response?
Even with the baffle you have, there will always be diffraction errors as well as the predicted (small) mismatch in directivity.
Which can be clearly seen in the off-axis is response.
So you literally can't optimize for one OR the other, but it will always be a compromise.
That's why I said to have a listen, to figure out how you want to compromise.
Even with the baffle you have, there will always be diffraction errors as well as the predicted (small) mismatch in directivity.
Which can be clearly seen in the off-axis is response.
So you literally can't optimize for one OR the other, but it will always be a compromise.
That's why I said to have a listen, to figure out how you want to compromise.
30dB scale?
How far do you guys want to pixel peep?
I do all my filtering on a 90dB scale.
Why?
Because it's the overall directivity and response that counts!
(Go read mister Toole again!)
Once that looks good, you can maybe fix some small things here and there.
Although I doubt that most people would hear it, especially small high Q dips less than 2dB are close to impossible to detect.
Way to much psycho-acoustics and masking going on.
Polar plots are only usable when normalized.
Non-normalized graphs in general can give you the wrong impressions about things that are totally out of context.
Same goes for distortion.
Very nice design, very well driver choice and thoughtful baffle geometry - thanks for sharing!
I'm just not a fan of the (intentionally i know, I'm familiar to the rationale) wide tweeter-mid spacing...
Some observations from the "Initial simulation results" in #492:
Would be interesting to see just the passive filter results without DSP, or both separately. Did you just shape the filter characteristics in the transfer area between mid + tweeter with the passive components, and do all remaining equalizing with the DSP?
Best regards
Peter
I'm just not a fan of the (intentionally i know, I'm familiar to the rationale) wide tweeter-mid spacing...
Some observations from the "Initial simulation results" in #492:
- series notch for SB15NBAC cone breakup resonance is a worthwile measure, but it seems not to hit the target frequency exacly (a bit too low tuning or to shallow Q)?
- there is some potential for phase matching at the crossover frequency between tweeter and mid.
- think about a parallel notch for tweeter impedance linearisation at fs to avoid the hump below 1kHz, that may also help with the phase.
- sacrify even a bit more on-axis linearity than in the second approach between 2,5-5,5 kHz, until the power response has the same or even half a dB lower level than in the 600Hz-1,2kHz range, from my experience this sounds most neutral.
Would be interesting to see just the passive filter results without DSP, or both separately. Did you just shape the filter characteristics in the transfer area between mid + tweeter with the passive components, and do all remaining equalizing with the DSP?
Best regards
Peter
Speaking of phase, what I also noticed, is that you don't have the same shelving filter for the woofer and for the rest.
In this case it's obvious, because the baffle is not the same width, but it does create a bit phase difference sometimes.
Although there is so much stuff going on that it's a little hard to tell.
In general, I would personally optimize better, this can be done with less biquads.
In this case it's obvious, because the baffle is not the same width, but it does create a bit phase difference sometimes.
Although there is so much stuff going on that it's a little hard to tell.
In general, I would personally optimize better, this can be done with less biquads.
If you want to do your crossovers at 90dB scale that’s your prerogative.
I’m catching up on Jim’s posts whilst on vacation. Do you have a problem if I request to see it a bit more easily on my phone’s 6” screen?
Now I understand why other people become tired of your commentary sometimes.
PS. There's no difference in the directivity for the 2 plots- it's already baked in.
@hifijim
I see a dip of about 1-2dB for of an octave between 2.5K and 5Khz, such that the predicted in-room response and power response are flat in the 2nd option. I concur with fluid. I'd start with the 2nd, listen and tweak to your listening material's / sonic preference.
I’m catching up on Jim’s posts whilst on vacation. Do you have a problem if I request to see it a bit more easily on my phone’s 6” screen?
Now I understand why other people become tired of your commentary sometimes.
PS. There's no difference in the directivity for the 2 plots- it's already baked in.
@hifijim
I see a dip of about 1-2dB for of an octave between 2.5K and 5Khz, such that the predicted in-room response and power response are flat in the 2nd option. I concur with fluid. I'd start with the 2nd, listen and tweak to your listening material's / sonic preference.
Last edited: