Sica 5.5 Coaxial 3-way towers and center speakers

[Windigo]

I recently completed a couple of floor standers and center channel using the Sica 5,5 C 1,5 CP and SB17NRX2C35-4 (two in series).

They are a sealed design since they will be used in a HT with subs. Volume of the woofer enclosures is about 24 liters.

Cabinets are made from MDF with internal damping materials and bracing. The baffles are 20cm wide and the towers stand about 1m tall while the canter is 60cm wide. The side panels are angled in at around 7 degrees and the woofer part of the towers is tilted back at 5 degrees. The center channel has an internal separate enclosure for the coaxial driver.

The crossover is around 450Hz and 2.7kHz and is based on impedance and frequiency measurements in the correct boxes (for the towers, the center channel is using the same crossover).
They are hot glued to MDF boards and soldered directly together. Not pretty but it works. The midrange parts are placed in the coax cabinet up top and the rest in the bottom of the speaker. I know placing inductors close to eachother and oriented the same way is not optimal, but the effect is negligible. Iron core is used for low DCR for the woofers and in parallell with the midrange. Aircores are used in series with the midrange and for the tweeter. All caps except for the one on the woofers are film/foil.

I took some measurements of them and also compared to my Kef Q550 that I was using before.

All measurements shown are using Dayton IMM-6 from about 60cm away. Gating is -1ms and +4.5ms and no smoothing applied unless otherwise is mentioned. Volume/gain was not changed at any time during measurement.

First, the DIY speakers on axis and at 15, 30, 45 and 60 degrees. The down slope from 600Hz is not there in near field measurements or with wider gating. May be related to the distance to the woofers.

Next, the same measurements for the KEFs:

And both on axis (Teal: DIY, Red: KEF)

And with wide gating (>30ms) and psychoacoustic smoothing. The dip around 300Hz seems to be reflection related as it is not consistent when moving the speaker.

Lastly, since I forgot to do any kind of in room measurement at the listening position when I had the gear set up, here is both speakers as measured by my AVR, around 2.8 meters away. Seems that the AVR compensates for the downwards response that we should see, but it may give a rough idea of the behaviour anyways.

I don't have an SPL meter so I did not include any distortion measurements. I was doing the test at roughly 90dB and both speakers showed less than 1% above 200Hz so it should really not be an issue.

Sensitivity is very similar as you can see from the graphs, but the Kefs are lower impedance so the DIY speakers should be easier on the amp.

Overall, I am happy with how these turned out and I really like the sound subjectively. In the future, I may do another crossover version to try and flatten out the upper mids and lower treble a bit, but to they sound good to me as they are.
For what subjective impressions are worth, here are some of my observations. Note that I have not been able to do any real A/B testing, volume matching etc. The Kefs sound a bit sharper but can get sligtly "metallic" timbre. The DIY speakers sound compairativly fuller/smoother, probably due to the slight upper mid dip and less elevated treble. Both speakers have good off-axis sound dispersion so they image well and are easy to get to "dissappear" in the room. The DIY speaker, probably helped by larger drivers and a 3-way design rather than 2.5 like the KEFs, have more satisfying dynamic "punch".

 

[JohnnyNG]

Wow, what awesome work. That Sica is such a killer driver and at an amazing price to boot. Crossed at 450Hz and used as a mid, dynamics are almost limitless.

 

[Matt_Holland]

Really great work!
If you try some crossover changes, some mono A/B comparisons should be fairly easy.

 

[Windigo]

Wow, what awesome work. That Sica is such a killer driver and at an amazing price to boot. Crossed at 450Hz and used as a mid, dynamics are almost limitless.

Thank you! Yes it is a really great driver. Looks and feels nice as well. And along with the SBs which are crossed to subs (2x dual opposed 12”) at 60Hz there should be very little compression and distortion even at 100+dB.

 

[Windigo]

Really great work!
If you try some crossover changes, some mono A/B comparisons should be fairly easy.

Thank you!
Yes that would be very interesting to see how flat tehy can get. The tweeter does not have the most even response so it will require some good trail and error.
For now I am using them with and AVR so basic correction/EQ can be applied so I am not in a rush to improve them.

 

[ppataki]

First, the DIY speakers on axis and at 15, 30, 45 and 60 degrees. The down slope from 600Hz is not there in near field measurements or with wider gating. May be related to the distance to the woofers.

Next, the same measurements for the KEFs:

Congratulations, really cool project and beautiful woodwork!

Just one question: looking at the above off-axis graphs, the KEFs seem to have a wider dispersion above 12kHz - can you hear any difference in that regard (subjectively speaking)?

 

[voodooless]

Those are some nice looking speakers! Congrats! Excellent craftsmanship!

These Sica coaxials are really excellent!

DIY speaker, probably helped by larger drivers and a 3-way design rather than 2.5 like the KEFs, have more satisfying dynamic "punch".

The Q550 almost looks small next to your DIY speaker It obviously has only the Uni-Q and a single 5.15” driver to help down low, vs your two high quality 6.5” woofers.

I think the newer KEF Q meta series will be a much tougher challenge. It’s a true 3-way with a smaller coaxial and you can have more “real” drivers for the bass. So I was lazy and have the Q11 meta on order If I would not have gone the DIY route, the Sica would have been really high on the list for sure.

 

[Windigo]

Congratulations, really cool project and beautiful woodwork!

Just one question: looking at the above off-axis graphs, the KEFs seem to have a wider dispersion above 12kHz - can you hear any difference in that regard (subjectively speaking)?

Thank you!

The KEFs generally sound little bit brighter and sharper (sometimes just a tad too sharp). Not a massive difference and I suspect it has more to do with the balance of mids to lower treble. It is hard for me to pinpoint exactly, could certainly be that it plays a role, but I would say generally variations above 10kHz need to be fairly large to have a significant impact on the sound for most people.

 

[JohnnyNG]

Thank you! Yes it is a really great driver. Looks and feels nice as well. And along with the SBs which are crossed to subs (2x dual opposed 12”) at 60Hz there should be very little compression and distortion even at 100+dB.

Yeah, tough to beat a 4-way like that.

 

[Windigo]

Those are some nice looking speakers! Congrats! Excellent craftsmanship!

These Sica coaxials are really excellent!

The Q550 almost looks small next to your DIY speaker It obviously has only the Uni-Q and a single 5.15” driver to help down low, vs your two high quality 6.5” woofers.

I think the newer KEF Q meta series will be a much tougher challenge. It’s a true 3-way with a smaller coaxial and you can have more “real” drivers for the bass. So I was lazy and have the Q11 meta on order If I would not have gone the DIY route, the Sica would have been really high on the list for sure.

Thank you! I do my best with limited woodworking experience.

Yeah not really fair in the dynamic range department but fun to compare anyways

The new Q meta speakers look fantastic! Bet they will sound amazing

 

[Colonel7]

The KEFs generally sound little bit brighter and sharper (sometimes just a tad too sharp).

Nice job! What’s the intended listening access for your DIYs and what do you do for your KEFs? Many coaxial speaker designs aim for 10-15 degrees off-axis to deal with the diffraction issues about 8-11k

 

[Windigo]

Nice job! What’s the intended listening access for your DIYs and what do you do for your KEFs? Many coaxial speaker designs aim for 10-15 degrees off-axis to deal with the diffraction issues about 8-11k

Thank you!
They are used with wide seating since they are in the living room, but I designed the crossover for on axis, even if the 15 degree response looks a bit better up top. I keep them angled just a bit out from the main listening position but there is not much difference moving directly on axis or slightly more to the side.