Hello fellow DIYers and audio enthusiasts,
For comparison the two graphs look like this:
While not really what I wanted to see, it seemed like the new driver would do the job I was hoping for. Eventually, I decided to build this MKII version of the same Swans speaker kit with the new mid-woofer from HiVi. As for the tweeter, I have opted for HiVi Q3. This tweeter has relatively low resonant frequency and its frequency response is nice. Besides, it matches the visual aesthetic of D6.8MKII and it fits perfectly well within the recessed space which is originally meant for Q1R:
THE CABINETS
A couple more pictures for perspective:
THE CROSSOVER
Since the value of R2 and R3 are a bit unusual, I’ve put 8.2 and 22 Ohms resistors instead - don’t think that there’s a big difference.
Below are the snapshots from XSim with normal and inverted tweeter polarity:
SYSTEM MEASUREMENTS
A few years ago I shared my attempt to improve the crossover of the commercial DIY speaker kit HiVi Swans DIY2.2-A. Back then I suggested a modification based on the crossover design of Swans D2.1SE speakers which employ the same drivers (HiVi D6.8B & HiVi Q1R) in almost the same enclosure. In case anyone is interested, here is the thread about that.
Anyway, as much as I like these speakers, I decided to sell them and move on to my next DIY project, which is the topic of this post. It was around the same time I was tinkering around with the Swans speaker kit, when I noticed that there’s a new version of the mid-woofer D6.8B, namely D6.8MKII. The new model has very similar dimensions to its predecessor, so I started entertaining the idea of using it with the same cabinets from the original Swans kit. Luckily, one can source only the cabinets without the need to buy the whole kit. I am not blessed with a workshop where I can build my own speaker boxes, so I have to rely on prefabricated cabinets, or flatpack kits ready to be assembled. I am well aware that this is far from ideal, but the alternative of using the services of a professional carpenter/joiner makes it much more difficult for me.
Next, I was curious to see how the new driver would perform in the same enclosure volume and vent dimensions, so I ran some WinISD simulations based on the provided data. What I was hoping for is less bass compared to its predecessor, mostly because of some unpleasant room modes I was experiencing with the Swans speakers before. Don’t get me wrong, it’s not that I don’t like bass, but what I wanted to accomplish with the new speakers is to overcome the typical problem with so many vented speakers nowadays – the excessive bass as a result of non-flat bass reflex alignment:
Many speakers are intentionally designed that way. I reckon, this is a great selling point for the speaker manufacturers – everyone loves bass and this is a way to squeeze lower F3 from a relatively small enclosure. Unfortunately, it comes at a price – degraded transients, lack of punch and somewhat loose bass. The way I understand it, this could be a consequence of using a driver with higher value of Qts (>0.4). With slightly lower Qts than its predecessor, the new D6.8MKII has prompted me take a closer look at it as a potential candidate for my new project and the WinISD simulation turned out to be promising:
Anyway, as much as I like these speakers, I decided to sell them and move on to my next DIY project, which is the topic of this post. It was around the same time I was tinkering around with the Swans speaker kit, when I noticed that there’s a new version of the mid-woofer D6.8B, namely D6.8MKII. The new model has very similar dimensions to its predecessor, so I started entertaining the idea of using it with the same cabinets from the original Swans kit. Luckily, one can source only the cabinets without the need to buy the whole kit. I am not blessed with a workshop where I can build my own speaker boxes, so I have to rely on prefabricated cabinets, or flatpack kits ready to be assembled. I am well aware that this is far from ideal, but the alternative of using the services of a professional carpenter/joiner makes it much more difficult for me.
Next, I was curious to see how the new driver would perform in the same enclosure volume and vent dimensions, so I ran some WinISD simulations based on the provided data. What I was hoping for is less bass compared to its predecessor, mostly because of some unpleasant room modes I was experiencing with the Swans speakers before. Don’t get me wrong, it’s not that I don’t like bass, but what I wanted to accomplish with the new speakers is to overcome the typical problem with so many vented speakers nowadays – the excessive bass as a result of non-flat bass reflex alignment:
Many speakers are intentionally designed that way. I reckon, this is a great selling point for the speaker manufacturers – everyone loves bass and this is a way to squeeze lower F3 from a relatively small enclosure. Unfortunately, it comes at a price – degraded transients, lack of punch and somewhat loose bass. The way I understand it, this could be a consequence of using a driver with higher value of Qts (>0.4). With slightly lower Qts than its predecessor, the new D6.8MKII has prompted me take a closer look at it as a potential candidate for my new project and the WinISD simulation turned out to be promising:
For comparison the two graphs look like this:
While not really what I wanted to see, it seemed like the new driver would do the job I was hoping for. Eventually, I decided to build this MKII version of the same Swans speaker kit with the new mid-woofer from HiVi. As for the tweeter, I have opted for HiVi Q3. This tweeter has relatively low resonant frequency and its frequency response is nice. Besides, it matches the visual aesthetic of D6.8MKII and it fits perfectly well within the recessed space which is originally meant for Q1R:
THE CABINETS
The baffle cutout diameter seemed to be a bit tight for the new mid-woofer, but nothing to really worry about. The bass reflex ports for this kit are provided with the component pack and are not included when you purchase only the cabinets, so finding a replacement might prove tricky.
The speaker dimensions are 205x350x305 mm (WxHxD). The cabinets are made of 18 mm MDF and there’s bracing inside. The vent is 51 mm diameter and 140 mm long. This tunes the boxes to around 49 Hz and provides a cut-off frequency (f3) of 45 Hz.
The speaker dimensions are 205x350x305 mm (WxHxD). The cabinets are made of 18 mm MDF and there’s bracing inside. The vent is 51 mm diameter and 140 mm long. This tunes the boxes to around 49 Hz and provides a cut-off frequency (f3) of 45 Hz.
A couple more pictures for perspective:
THE CROSSOVER
With these speakers I wanted to challenge myself and see if I could do a better job in terms of proper phase tracking and baffle step compensation. The result is a combination of 2nd and 3rd order electrical filters (for the woofer and the tweeter respectively). I have aimed to cross at around 2 KHz:
Since the value of R2 and R3 are a bit unusual, I’ve put 8.2 and 22 Ohms resistors instead - don’t think that there’s a big difference.
Below are the snapshots from XSim with normal and inverted tweeter polarity:
SYSTEM MEASUREMENTS
Attachments
Yes, that’s right – they bear a striking resemblance to some Dynaudio models. For someone who used to have Dynaudio speakers before, I have to say that there’s more than just the physical similarities between them. That being said, it was the disappointment of the Dynaudio brand in general that drove me to the world of DIY. I don’t say that they can’t make decent speakers, but for that one has to climb up the price ladder significantly. In my experience, their entry-level models are rubbish, and they are not cheap at all.
Thanks, you’re welcome. As for the dual purpose look of the stands, I don’t have a cat if this is what you’re implying. It’s just for decoration purposes as my partner can’t stand the original look of the stands. It’s simply the spouse approval factor in action 
Thanks, unfortunately I haven’t taken any off-axis measurements. Maybe I’ll do that at some point in foreseeable future and post the results here.
Well done. I really like the result.
The similarity between HiVi (Swans) D series and Dynaudio or Totem is not accidental. But that is history and another story.
The large diameter Tweeter Plate is a long-term issue at HiVi (Swans). K1, SD1, Q series, X1-A had a diameter of 116 mm.
The only solution I can think of (I haven't tried it in practice) for using a tweeter with a smaller Plate diameter is an intermediate ring. Probably best as 3D printing. Such a solution presupposes adjustment of the depth of the embedment and the mounting hole. For tweeters with a small diameter Plate then also a mounting ring for a small tweeter.
It should be clear from the picture what I mean. Due to the topic, I used HiVi tweeters, but it can certainly be applied to other brands as well.
The color of the intermediate ring is a matter of taste. The gray color is for illustration only.
The similarity between HiVi (Swans) D series and Dynaudio or Totem is not accidental. But that is history and another story.
The large diameter Tweeter Plate is a long-term issue at HiVi (Swans). K1, SD1, Q series, X1-A had a diameter of 116 mm.
The only solution I can think of (I haven't tried it in practice) for using a tweeter with a smaller Plate diameter is an intermediate ring. Probably best as 3D printing. Such a solution presupposes adjustment of the depth of the embedment and the mounting hole. For tweeters with a small diameter Plate then also a mounting ring for a small tweeter.
It should be clear from the picture what I mean. Due to the topic, I used HiVi tweeters, but it can certainly be applied to other brands as well.
The color of the intermediate ring is a matter of taste. The gray color is for illustration only.
Yes, a diameter of 116 mm is rather odd and it’s difficult to find a replacement. I have seen some Wavecor tweeters with diameter of 115 mm and that’s the closest I could find. The proposed ring is a potential solution, indeed.
Obviously, the best way is to design the speakers from scratch and use custom made cabinets. Using alternatives like I did with this project is a compromise. That being said, the overall result is quite satisfying and I am generally happy about how it turned out, especially the crossover part. But in all fairness I think the volume and the tuning of the cabinets is not optimal. My complaint is that the mid-bass is somewhat weak and not particularly articulate. Don’t get me wrong, they seem to dig deep enough, but I would appreciate some more punch and clear definition in the mid-bass region. I am considering building smaller cabinets with different vent dimension and exactly the same size of the front baffle, so I can use the same crossover again. It would come at the price of a higher F3 frequency, but I hope for better mid-bass performance. It would be an interesting experiment.
Obviously, the best way is to design the speakers from scratch and use custom made cabinets. Using alternatives like I did with this project is a compromise. That being said, the overall result is quite satisfying and I am generally happy about how it turned out, especially the crossover part. But in all fairness I think the volume and the tuning of the cabinets is not optimal. My complaint is that the mid-bass is somewhat weak and not particularly articulate. Don’t get me wrong, they seem to dig deep enough, but I would appreciate some more punch and clear definition in the mid-bass region. I am considering building smaller cabinets with different vent dimension and exactly the same size of the front baffle, so I can use the same crossover again. It would come at the price of a higher F3 frequency, but I hope for better mid-bass performance. It would be an interesting experiment.