Due to the completely new installation environment of the drivers, you also need to develop a completely new crossover. Also, the cabinet set-up (Cavity in the door) will be completely different, a setup like this probably wouldn't get by without a subwoofer.
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Small 2-way speakers with linear on-axis and power response characteristics (Scan Speak and SB Acoustics drivers). H&V off-axis measurements included
- Thread starter XMechanik
- Start date
- Thread Starter
- #663
Thanks for sharing your results. Wonderful finish and great shot!
My speakers are officially done. Dampening material is 1" Denim Insulation from PE. Seems hard to track down a good thick material that you can just throw in so I didn't mind the high cost. Very nice stuff, honestly if you're going to put the effort into building a speaker might as well go all out. Not sure how much is ideal for a PR box but I just lined the walls and put a chunk in the middle.
As for the passive radiator tuning, I saw one user settle on 16 grams and vcad shows that be being fine. I said why not go higher and added 20g to my PR's. Should make the tuning around 41-43hz. I measured in room with and without the added weight and it looked pretty much identical so who knows if it helped the low end. Sounds like it helped a little.
Subjectively it's hard to comment on the speakers, not because I don't think they're great but because I currently use them on stands on a desk and I really hate desk reflections, just ruins speakers for me. I have other stands I will try but that requires moving a lot around in the room. I'm facing the long wall as well so not a great setup.
With that said I find the driver integration to be superb. The tiny vertical lobes are basically inaudible, sounds about as coherent as I think one could get a passive two way. Drivers were all really well matched, final imp and FR are nearly identical. I feel the woofer is a bit lacking in oomph but maybe my expectations are too high for the driver size. I usually add a 3db low shelf at 300hz with my speakers and I did the same with the Mechano23 so maybe I just like some more bass and lower mids. For my listening distance and volume the woofer doesn't seem to care about the additional gain. I feel with that added low shelf the speaker is tonally correct and I found the imaging to be great and engaging, pretty addicting speaker to listen to. The closeely matched drivers helps a lot to establish a strong center. Oh and the Midrange is great, smooth DI through a region that is normally not great makes the mid range crystal clear.
First time using a PR, honestly hard to say if I prefer to a port. I'd like to think the mid range around 700-1000hz sounds cleaner and I'd feel confident attributing that to a lack of port resonances. Definitely noticable if you're playing piano through them. I did setup them one time in front of my keyboard and they sounded impressively convincing for that use. As for the the low end I don't know, it's there but it feels a bit subdued. EQ kind of brings it back but it still feels lacking. At times it feels like the impact of the bass is fine but the sustain of bass notes is poor.
I may build some bass module stands and use them in a 3 way system which kind of defeats the point of the PR but oh well. If I were to make them again I'd build them sealed and save some money on the PR's and just add some subs. I think they accomplish their goal for me, which was a passive near field speaker that was neutral with good DI and klippel verified performance. I like building speakers but not a big fan of the analysis and design phase. If I would change anything I'd probably try and use some super high xmax woofer like a dayton epique so people like myself with big amps could get more punch and sub bass and not care about the sensitvity loss. I'm just so used to speakers that tend to start rolling off at around 30hz so my only con with the mechnao23 is the low end. But yeah it's a good neutral speaker with good DI, what is there to say. If you're in the market for a good bookshelf design just make these, probably better than most other stuff out there. I hope to move things around and get them placed better and return with my impressions.
As for the passive radiator tuning, I saw one user settle on 16 grams and vcad shows that be being fine. I said why not go higher and added 20g to my PR's. Should make the tuning around 41-43hz. I measured in room with and without the added weight and it looked pretty much identical so who knows if it helped the low end. Sounds like it helped a little.
Subjectively it's hard to comment on the speakers, not because I don't think they're great but because I currently use them on stands on a desk and I really hate desk reflections, just ruins speakers for me. I have other stands I will try but that requires moving a lot around in the room. I'm facing the long wall as well so not a great setup.
With that said I find the driver integration to be superb. The tiny vertical lobes are basically inaudible, sounds about as coherent as I think one could get a passive two way. Drivers were all really well matched, final imp and FR are nearly identical. I feel the woofer is a bit lacking in oomph but maybe my expectations are too high for the driver size. I usually add a 3db low shelf at 300hz with my speakers and I did the same with the Mechano23 so maybe I just like some more bass and lower mids. For my listening distance and volume the woofer doesn't seem to care about the additional gain. I feel with that added low shelf the speaker is tonally correct and I found the imaging to be great and engaging, pretty addicting speaker to listen to. The closeely matched drivers helps a lot to establish a strong center. Oh and the Midrange is great, smooth DI through a region that is normally not great makes the mid range crystal clear.
First time using a PR, honestly hard to say if I prefer to a port. I'd like to think the mid range around 700-1000hz sounds cleaner and I'd feel confident attributing that to a lack of port resonances. Definitely noticable if you're playing piano through them. I did setup them one time in front of my keyboard and they sounded impressively convincing for that use. As for the the low end I don't know, it's there but it feels a bit subdued. EQ kind of brings it back but it still feels lacking. At times it feels like the impact of the bass is fine but the sustain of bass notes is poor.
I may build some bass module stands and use them in a 3 way system which kind of defeats the point of the PR but oh well. If I were to make them again I'd build them sealed and save some money on the PR's and just add some subs. I think they accomplish their goal for me, which was a passive near field speaker that was neutral with good DI and klippel verified performance. I like building speakers but not a big fan of the analysis and design phase. If I would change anything I'd probably try and use some super high xmax woofer like a dayton epique so people like myself with big amps could get more punch and sub bass and not care about the sensitvity loss. I'm just so used to speakers that tend to start rolling off at around 30hz so my only con with the mechnao23 is the low end. But yeah it's a good neutral speaker with good DI, what is there to say. If you're in the market for a good bookshelf design just make these, probably better than most other stuff out there. I hope to move things around and get them placed better and return with my impressions.
agnostic1er
Member
- Joined
- Feb 22, 2025
- Messages
- 28
- Likes
- 54
Hello,
To XMechanik: I send you a message on your "xmachina.ai"at"gmail.com" adress but I don't know if you received it.
I have two questions about mechano23, please:
I also tried to download XMachina passive crossover designer here: https://xmachina-ai.blogspot.com/p/current-version-info.html
but page seems to no more exist.
Thanks for your time.
Hubert
To XMechanik: I send you a message on your "xmachina.ai"at"gmail.com" adress but I don't know if you received it.
I have two questions about mechano23, please:
- What a distance to mic did you use for the frequencies measurements of its drivers ?
- On your rotary table, do you use a system to maintain the same distance driver to mic for all your off-axis measurements ?
I also tried to download XMachina passive crossover designer here: https://xmachina-ai.blogspot.com/p/current-version-info.html
but page seems to no more exist.
Thanks for your time.
Hubert
The axis of rotation should be in the horizontal center of the baffle, which leads to a design like the one I once designed and built for the rotary table (or better: rotary tower
) :Sorry for the mess and the poor quality of the photos, but the principle should be clear, I think.
The measurement distance depends on the size of the speaker. For a speaker like the one in question here, I would consider one meter to be sufficient to include the influence of the front baffle.
Last edited:
agnostic1er
Member
- Joined
- Feb 22, 2025
- Messages
- 28
- Likes
- 54
Azrael, thanks for your reply.
In the mean time, I saw this: https://www.audiosciencereview.com/.../diy-spinorama-turntables.53480/#post-1934440
which matches what you describe.
And yes, 1M could certainly be the measurement distance XMechanik used for this speaker.
In the mean time, I saw this: https://www.audiosciencereview.com/.../diy-spinorama-turntables.53480/#post-1934440
which matches what you describe.
And yes, 1M could certainly be the measurement distance XMechanik used for this speaker.
- Thread Starter
- #668
I think it's safe to assume that the software is discontinued.
I tried to keep 1m distance indeed. And by the way I have to admit that my rotary table has a drawback: it's not 100% rigid. In case of non simmetrical tabletop loading (esp. for the V orbit measurement, when the speaker lies on side) some precession of the rotation may occur. I ususally try to compensate this with padding something under but I think it doesn't rule off the error completely. Azrael's solution seems to be much better in that regard but on the other hand it has 4 pipes instead of 1, creating more potential reflection sources.
agnostic1er
Member
- Joined
- Feb 22, 2025
- Messages
- 28
- Likes
- 54
agnostic1er
Member
- Joined
- Feb 22, 2025
- Messages
- 28
- Likes
- 54
I downloaded your engine on diyAudio and I'm currently playing with it, even including off-axis frd files. I'm used to create passive x-overs since 20 years and found the engine to create very uncommon circuits
agnostic1er
Member
- Joined
- Feb 22, 2025
- Messages
- 28
- Likes
- 54
Hello,
Here is an example of a 3-way floor speaker using U18 woofer, RS125 mid and a 27TTFNC tweeter, so common and relatively easy drivers to work with.
My current crossover developed with H and V off-axis 0-90° measurements use 15 components.
edit: the 5kHz dip is due to Vituix option 1M distance in spite of the real 2.6M listening distance because woofer is the listening high and mid, then tweeter lower. The dip is filled at 2.6M distance.
The XMachina crossover use 47 components, despite a very simple task/strategy: 12/12/12dB, spl target +/-3dB, Zmini 4ohm, sys SPL 50%, pwrResp 50%, 250 to 15000Hz for the pwrResp, 15minutes design time and above all 15% target compliance 85% circuit simplicity and reduce number of parts.
Any idea why such a complicated design?
Here is an example of a 3-way floor speaker using U18 woofer, RS125 mid and a 27TTFNC tweeter, so common and relatively easy drivers to work with.
My current crossover developed with H and V off-axis 0-90° measurements use 15 components.
edit: the 5kHz dip is due to Vituix option 1M distance in spite of the real 2.6M listening distance because woofer is the listening high and mid, then tweeter lower. The dip is filled at 2.6M distance.
The XMachina crossover use 47 components, despite a very simple task/strategy: 12/12/12dB, spl target +/-3dB, Zmini 4ohm, sys SPL 50%, pwrResp 50%, 250 to 15000Hz for the pwrResp, 15minutes design time and above all 15% target compliance 85% circuit simplicity and reduce number of parts.
Any idea why such a complicated design?
Attachments
Last edited:
agnostic1er
Member
- Joined
- Feb 22, 2025
- Messages
- 28
- Likes
- 54
Some minutes ago I made XMachina working on exactly the same task as in my previous post with the only difference being 5% target compliance 95% circuit simplicity (vs previous 15 and 85%). Astonishingly the result is even a little better with only 20 components !!!
Attachments
- Thread Starter
- #674
There must be something wrong with your settings. First, you need to have spl target below drivers spl, since the crossover is passive. It appears that this rule is often broken. There could be other issues as well.
agnostic1er
Member
- Joined
- Feb 22, 2025
- Messages
- 28
- Likes
- 54
Tried my speakers out in a few different rooms but every space I felt like the bass was just kind of, not there at all. Current setup looks average response wise other than the low end, not sure if it's the room or the speaker. I did note the speaker sounded like it had a high pass in every space I used it in. Kind of strange.
Red is non-EQ'd response and you can see it kind of dives down just under 200hz. Blue is if I add a 100hz low shelf filter of 10db. Kind of a lot. With the EQ the speaker sounds great, without it, it sounds lifeless and thin.
Red is non-EQ'd response and you can see it kind of dives down just under 200hz. Blue is if I add a 100hz low shelf filter of 10db. Kind of a lot. With the EQ the speaker sounds great, without it, it sounds lifeless and thin.
Last edited:
I finished my rear speakers a couple of weeks ago and I´m very pleased how they turned out. I built them with closed enclosure (little under 4 l) and mounted them to the ceiling. This time I used Kimmosto´s crossover because it takes less space and it turned out to be very close to the original in sound perspective. I tested the speakers in the front speakers position and bass was definitely weaker, but not as much i feared. The ceiling position on the other hand gives them noticeable bass boost and the sound is very close to the original. They are very good match in surround sound and now the sound field is uniform (the former rear speakers are Wharfedale Diamond 9.0sr). The response chart´s blue curve is the original speaker (with passive radiator). Measurement was taken with REW in the center of the room ( microphone distance 70cm).
Attachments
- Thread Starter
- #678
You may try to verify this by taking near-field measurements. They're relatively easy to perform and valid up to about 200Hz.
I did that, mine looks close to Amir's. Red line is Amir review close mic. Mine does look like it has a few humps here and there especially at 300hz. I wonder if the PR bleed's the resonance a bit.
Sine gen's are showing that I'm getting output down to about 40hz then it falls like a rock, seems fine for the woofer size. I think the main culprit is really this tiny room I have the speakers in and the desk they sit in front of. Just way too many reflections. It's a long but narrow room and the width is the same as the height. I plan on tossing in A LOT of treatment into this room.
Last edited:
- Thread Starter
- #680
I don't know if this will bring anything new to the explanation of the phenomenon, but it remains to determine the NF characteristic as a whole. Could you shift the PR response by about 3 dB and add it to the cone response? This shift comes from the difference in diaphragm piston area. For the midwoofer it's 87 cm2, and 178 cm2 for the PR . Therefore, shift = 10 * log(178/87) = 3.1 dB.
