Simple Inexpensive Passive 2-way with Dayton and SB drivers.

[Ktacos]

The vertical dispersion could be of some interest.

I would wager it's not great, but I'm pretty tired on measuring these things. Lot of work to drag my rig outside. Subjectively one really needs to sit with their ears between the tweeter and woofer to get good results.

I expect very low impedance at some higher frequency beyond the range you tested, but the (digital) amp might still ‚see‘ it. A small resistor might help without sacrifice.

I tested this, it didn't make any difference.

That would solve probs with directivity, and may rendern the nasty cap parallel to the tweet obsolete.

This tweeter seems to just have a rising response when you place it on a bookshelf sized baffle. Every DIY design I've seen shows a similar response up top, and even when using active filtering it needs a shelf to flatten it out. I genuinely cannot stand any amount of elevated treble. It sounds better with the parallel cap.

Lowering the xover point doesn't fix the DI, even if trying ~1khz which is just about on tweeter FS. I'm still showing a DI mismatch at this xover point.

I would need to remeasure the speaker and I might be able to fix it's problems then, but do I want to? Not sure I do. I've built like 6-7 speakers over the past 2 years and I'm kind of tired. I didn't have high expectations for these as frankly speakers without waveguided tweeters tend to not sound good to me. Another issue is the woofer has an elevated response around 700hz that is tricky to get around.

The bass tuning is a bit high, maybe an extended bass shelf is more desirable.

It would still need EQ. If one looks at the cnote data taken by Amir, that one is using a lower point tuning but it's still a good 5db down from the rest of the woofers response. I think this is just about all one realistically get out of this small woofer.

I frequently use some Kali LP6 which have no problem keeping their low end pretty flat down to the port tuning.

 

[a4eaudio]

...I would need to remeasure the speaker and I might be able to fix it's problems then, but do I want to?

I completely get your frustrations and whether or not to take more measurements...BUT if you choose to...

You need to know what is going on below 500Hz. There is a very good chance you just aren't accounting for baffle step losses accurately. (See screenshot below.)

You can get good quasi-anechoic measurements inside with 8-foot ceilings from 10Hz to 20kHz.
Farfield: [You can probably just skip this and use your original on-axis measurement from outside if you have confidence in it.] Set woofer at 48" from floor and measure at 3x the baffle width. Don't use anything to damp the reflections, you WANT to see them in your impulse response so you can set the gate/window accurately. Set your window according to the first reflection in the IR.
Nearfield: Reduce output 20dB and measure 1/4" from woofer dust cap and the same for the port.

You only need to perform the above for the on-axis measurement. Then use VituixCAD merger tool to merge the nearfield response to ALL of your on- and off-axis measurements with the click of the Export button. (You will need to simulate the diffraction response at 10m and 2m in the VituixCAD Diffraction tool.)

...It's evident that my speaker is quite bass deficient as it needs about +4-6db low freq shelf to bring it up to same level as the Kali.

It is often thought that baffle step compensation is just about the bass. But with a 7-1/2 inch wide baffle, baffle step loss begins at about 500Hz. The fundamental of most human voices and more than 50% of music instruments is below 500Hz, yet getting good measurements/simulations from 10Hz to 350Hz takes a bit of work to get right. Below is the Sd of the DSA135-8 on a baffle 7-1/2" wide by 11" tall. Everything 150Hz and below is down 4dB from 500Hz.

(You will also be able to easily see the port resonances in the port nearfield. Sometimes drilling four 1/8" holes at the midpoint of the port helps a lot, but in other cases it doesn't, in case you want to experiment with that. If the resonances are from standing waves rather than the port resonance, then damping is the solution.)

 

[Ktacos]

You need to know what is going on below 500Hz. There is a very good chance you just aren't accounting for baffle step losses accurately. (See screenshot below.)

Unlikely. Here is my measured response compared with diffraction sim.

You only need to perform the above for the on-axis measurement. Then use VituixCAD merger tool to merge the nearfield response to ALL of your on- and off-axis measurements with the click of the Export button. (You will need to simulate the diffraction response at 10m and 2m in the VituixCAD Diffraction tool.)

Already did all this.

 

[a4eaudio]

Already did all this.

You may have taken the measurements and followed the process, but you have not shown accurate quasi-anechoic measurements in any of your posts. Of course, you don't have to show them if you don't want to, but then no one can really offer useful advice either.

Unlikely. Here is my measured response compared with diffraction sim.

I'm not sure what is "unlikely". Your measured response shows, just like the diffraction sim, that output is falling 4 to 6dB below 500Hz. Since you don't show accurate quasi-anechoic measurements in your VituixCAD modelling, we can't tell much about BSC. But again, your created your thread to show your results and didn't ask for help or opinions, so if you are sure you have BSC correct than that is okay. (It may just be a coincidence that you found them to be 4-6dB too low in the low-end and a simulation as well as your own woofer measurements would suggest that if BSC wasn't implemented fully then the low end would be down 4-6dB. But we can't tell with the data provided.)

 

[Heinrich]

Unlikely. Here is my measured response compared with diffraction sim.

View attachment 460299




Already did all this.

O/k, you won‘t pick up the suggestions, fair enough. I personally wouldn‘t accept to dump that much, and decidedly successful work on the paint job ;-)

Regarding bass, you could easily adjust to the in-room response, instead of taking free field measurements. As an experienced DIY guy you already know that bass is never right out of the box.

 

[Ktacos]

I'm not sure what is "unlikely". Your measured response shows, just like the diffraction sim, that output is falling 4 to 6dB below 500Hz.

No it doesn't, where are you getting this from? Response looks pretty linear to me with respect to some room reflections, the speaker wasn't really near any boundaries.

Here is a close mic of the woofer with diffraction sim applied and filtering applied. I'm sorry but you are simply incorrect in your assumptions. The low bass response is likely just a limitation of passive filtering with this cabinet volume+this woofer.

 

[a4eaudio]

...Your measured response shows, just like the diffraction sim, that output is falling 4 to 6dB below 500Hz....

No it doesn't, where are you getting this from?

Your picture (dark red line) in post #23.

See the yellow shaded rectangle below:
Your SPL at 100Hz is 4dB less than it is at 500Hz.
Your SPL at 60 Hz is 9dB less than it is at 500Hz.
For full baffle step compensation, you would have a final speaker with quasi-anechoic response targeting close to flat at 75-77dB, like the bottom thicker pink line (based on the image, I know it is not necessarily SPL-accurate).

Here is a quasi-anechoic measurement of farfield, merged with nearfield of woofer and port with full BSC. Port tuned to 42Hz, F3 of 47Hz. Your final speaker should not look THAT different from this, other than different bumps and squiggles due to different drivers and a slightly higher F3.

...I'm sorry but you are simply incorrect in your assumptions...

I haven't made any assumptions. I merely stated that you haven't provided the measurements necessary to analyze the problem and also admitted that you didn't ask for me or anyone else to analyze it.

...The low bass response is likely just a limitation of passive filtering with this cabinet volume + this woofer.

However, I can assure you that THIS is not correct. The passive filtering is a low-pass around 2.7kHz so the only thing that could matter in the passive filtering is the BSC. The woofer (the same) and cabinet volume (the same or very close) to the C-Note and we know they can produce acceptable bass. You have tuned yours higher than the C-Notes, but not by that much, so the port tuning is probably not the problem.

 

[Ktacos]

Your SPL at 100Hz is 4dB less than it is at 500Hz.

Yeah, duh, those are raw responses of the driver before any filtering...

 

[Ktacos]

I haven't made any assumptions. I merely stated that you haven't provided the measurements necessary to analyze the problem

I did, in post 26. My own close mic with diffraction put against your sim (no port in mine).

 

[a4eaudio]

I did, in post 26...

Exactly. This is AFTER I posted that you hadn't provided the measurements.

 

[Ktacos]

O/k, you won‘t pick up the suggestions, fair enough.

I'm open to suggestions, just not too interested in ones that I feel confident aren't the problem. I think there was a bit of miscommunication which is why so much time was spent suggesting inadequate baffle loss compensation.

I would agree that my xover point could be much lower which would probably fix some of my complaints about the speaker. I may try that in the future.

 

[Salt]

It's the standard: xover at 2.X KHz, BR enclosure ...

 

[Ktacos]

It's the standard: xover at 2.X KHz, BR enclosure ...

I know, I'm really pushing the boundaries aren't I?

 

[Salt]

Bass is first : look at bass level tuning and what it provides for upper frequencies, adjust LP according and addressing bafflestep.Then HP and attenuation of tweeter should be adjusted to the 'base'.
That is basic.
Don't see pushing boundaries but well done basic work.

 

[Ktacos]

Don't see pushing boundaries but well done basic work.

It was a sarcastic comment.

 

[motomech]

It's a nice finish but it's not very durable.

I hooked up my Kali LP-6 today for a comparison with my speakers. Honest opinion from me about my speakers, not good lol. I would not suggest building them. It's evident that my speaker is quite bass deficient as it needs about +4-6db low freq shelf to bring it up to same level as the Kali. I considered trying some passive radiators instead of the port but it doesn't make sense to me to spend an extra $70+ on those. The port resonance for this speaker is around 1000hz and it's pretty noticable when playing piano. The imaging on mine seems worse, it overall just feels like a far less coherent speaker, I've certainly never been a fan of speakers with tweeter not time aligned with the woofer and I feel that annoys me here. There's definitely just something off with the speaker. Despite the on axis appearing quite flat on my speaker, the mid range sounds scooped. I assume this has to do with the speakers DI. If it is to believed, it's too wide in all the wrong places. I can listen to Kali all day with no complaints, they actually make me want to listen to music, my speakers, not so much. My speakers seem to just lose entire layers of the music that the Kali represents very well.

I'm a little disappointed but not surprised. It's not easy making a passive speaker and the benefits of active were certainly made even more apparent than they already were. Thought I could make something pretty good but at least to my standards I didn't get it right. I think the lack of a waveguide really hurts the speaker for instance.

Here's the DI from my sim (tweeter's upper range not accurate).

View attachment 460006

So yeah, little bummed to say the least. Thankfully I didn't spend too much money on these but I did lose a good bit of time. Definitely has put a damper in my DIY efforts but I feel it's important to be as honest with everyone here as well as myself so I wanted to share my feelings about the speaker. Granted, I'm a mix engineer and musician who is typically using stuff like genelec and nuemann monitors so perhaps I'm being a bit hard on myself. I think the average person would find the speakers to be fine, but I want to do better than fine.

There may be a way to salvage this project for little money.
Quite a few years ago, I built a pr. of C-notes when they first came out, but didn't care for them and they sat around. After awhile, on a lark and more as an experiment than anything else I modified them by lopping the ND25 tweeter off the wave guide and siliconed a ND28F-6 in it's place and then I installed the cross-over that Matt Grant designed for his NEXUS series of speakers.

Project NEXUS... A timbre matched family of Hi-Fi/HT speakers

Timbre matched family of Hi-Fi/HT speakers all based on the same midwoofer (Dayton DS135-8) and tweeter (ND28F-8). These designs came about from my...

www.hificircuit.com

When I posted what I did I took a lot of heat, but I can understand a designer could be a little aggravated when some idiot starts messing around with his design, so I left the whole thing drop(as far as posting about it).
But as I sit here listening to them a couple years later, they still blow me away, better than all my other sm. bookshelf's, including Aviatrxs designed by Curt Campbell, my S2000's designed by Paul Carmody and my Hivi Swans 2.1's.
The waveguide boosted the freq. response from the 2.1kHz. X-over up to around 4K to 5k Hz. so I had to add an L-pad to tone things down a bit, but since your tweeter holes is cut for the SB, you could just install the Dayton ND28 as is and leave Matt's X-over stock. The only prob. I see is that the ND28 is about 7 mm smaller than the SB, b ut I still think your existing cut-out would work and the gap could be filled.
The Nexus design calls for the DS woofer instead of the DSA, but they spec almost identical and I can't hear anything bad from the DSA. Actually, it's a little beast.
The Dayton ND28's are very inexpensive and the X-over parts aren't bad either.

 

[motomech]

I would wager it's not great, but I'm pretty tired on measuring these things. Lot of work to drag my rig outside. Subjectively one really needs to sit with their ears between the tweeter and woofer to get good results.


I tested this, it didn't make any difference.


This tweeter seems to just have a rising response when you place it on a bookshelf sized baffle. Every DIY design I've seen shows a similar response up top, and even when using active filtering it needs a shelf to flatten it out. I genuinely cannot stand any amount of elevated treble. It sounds better with the parallel cap.

Lowering the xover point doesn't fix the DI, even if trying ~1khz which is just about on tweeter FS. I'm still showing a DI mismatch at this xover point.

I would need to remeasure the speaker and I might be able to fix it's problems then, but do I want to? Not sure I do. I've built like 6-7 speakers over the past 2 years and I'm kind of tired. I didn't have high expectations for these as frankly speakers without waveguided tweeters tend to not sound good to me. Another issue is the woofer has an elevated response around 700hz that is tricky to get around.



It would still need EQ. If one looks at the cnote data taken by Amir, that one is using a lower point tuning but it's still a good 5db down from the rest of the woofers response. I think this is just about all one realistically get out of this small woofer.

View attachment 460267

I frequently use some Kali LP6 which have no problem keeping their low end pretty flat down to the port tuning.

View attachment 460269

Working on perfecting all the little factors like baffle step, acoustic centers, etc. is all well and good, but I have found using the X-over design of a well known designer is the best way to DIY
And Matt Grant is a Master x-over designer.
The prob. with the C-note is the X-over is too high and the woofer is allowed to ring.
Matt chose to X over the drivers at 2.1K Hz. and there is no 700 Hz. resonance.
And no EQ is required with these speakers, they are that gd.

 

[Ktacos]

C-notes when they first came out, but didn't care for them and they sat around. After awhile, on a lark and more as an experiment than anything else I modified them by lopping the ND25 tweeter off the wave guide and siliconed a ND28F-6 in it's place and then I installed the cross-over that Matt Grant designed for his NEXUS series of speakers.

Wow what a shockingly terrible idea.

The prob. with the C-note is the X-over is too high and the woofer is allowed to ring.

Doesn't seem like a problem, I have c notes and added the cap mod for breakup and noticed zero change. Real problem with cnotes is the shelved lower region of the woofer which is easily alleviated by a single filter. C notes sound pretty close to 8030c IME so uh, not sure I want to entertain what you're saying much. Perhaps you just don't like neutral speakers.

 

[motomech]

Wow what a shockingly terrible idea.



Doesn't seem like a problem, I have c notes and added the cap mod for breakup and noticed zero change. Real problem with cnotes is the shelved lower region of the woofer which is easily alleviated by a single filter. C notes sound pretty close to 8030c IME so uh, not sure I want to entertain what you're saying much. Perhaps you just don't like neutral speakers.

You are the one that has spent time, money and effort to produce speakers which you admit "don't sound good".
I only suggested you replace the tweeters and use a X-over for a known design that is highly acclaimed to recoup some of your investment
Suit yourself.

 

[Ktacos]

Suit yourself.

I will thank you.

Major reason I don't like diy anymore and tend not share anything with people, endless terrible suggestions from others. I genuinely enjoy building speakers but man I do not like interacting with the community at all. Probably the last time I share anything TBH.