Sony SS-CS5 crossover upgrade discussion

[Velocipede_Acoustics]

Hey all,

I uploaded this video where I discussed some of my findings regarding a crossover upgrade for the Sony SS-CS5 bookshelf speaker.

Sorry if it's a lot rambley and a bit long winded. I thought about making a write up, but that would take hours and I thought it easier to just brain dump into a microphone. I hope you don't mind!

 

[Velocipede_Acoustics]

Just put together the 3 way crossover with a few value tweaks. This looks, and sounds, good so far. I'm impressed at the $88/pair speaker. Basically a 90 degree listening window maybe? I'll flirting with a 2.5 way design too.

directivity 0,15,30,45 deg.

That 15 deg measurement in black looks great.

Separated for better visibility.

 

[alex-z]

How does it do vertically?

Also, if you do a full CEA-2034 spinorama you can import that data into VituixCAD and it will calculate the Harman preference score for you. You will need to splice in the near-field low frequency response for that though.

 

[Velocipede_Acoustics]

I've not decided to test vertically yet and I'm not sure how I'm going to do it in an accurate and meaningful way. Doing a full spinorama looks like a ton of work too. We'll see. I may need a dedicated measuring space, which I don't have right now.

 

[Pete Basel]

Nice work! It would be interesting to see the reverse nulls, if they are in-phase XOs.
2.5 way? Are you aware that Sony has a tower version with 2 woofers, was that
your planned starting point? We saw the tower at Best Buy yesterday.

 

[Velocipede_Acoustics]

Nice work! It would be interesting to see the reverse nulls, if they are in-phase XOs.
2.5 way? Are you aware that Sony has a tower version with 2 woofers, was that
your planned starting point? We saw the tower at Best Buy yesterday.

Thanks!

The phase alignment isn't perfect and may never be, especially trying to keep this budget friendly. They're pretty close. The drivers are summing. Not perfectly, but they are. I'm not terribly interested in nitpicking this build much more. $88 a pair with as much in crossover parts to get a pretty flat speaker that plays fairly wide. Even if phase isn't perfect, this result out to 45 degrees is pretty good. I'm going to build this out, stuff the box better, may add some bracing and we'll see if it gets any better.

2.5 way going up. Sort of a 2.5 way where we run the 1" driver with just a high pass and the 0.5" on on the same network with another high pass pass but at a higher frequency to fill in past 8khz... ish... I've since played with this a little and determined it wasn't feasible. It was a potential to save a few parts. I'm trying to keep this in budget the best I can, and doing a 3 way cheaply is difficult.

As for the towers, I would not mind getting a pair to tinker with. Those may also be a budget hero if I can accomplish the same thing. These do have distortion issues 250hz going down so another woofer would go a long way to help cut that down I think.

 

[Pete Basel]

Perhaps you tried this, but, did you consider putting the tweeter attenuator resistor
after the high pass filter, then connecting the super tweeter to the tweeter's high
pass through a single cap. Try it with whatever polarity works better on the super
tweeter.
It saves an inductor and capacitor obviously, not much but simple is better.

 

[Velocipede_Acoustics]

Perhaps you tried this, but, did you consider putting the tweeter attenuator resistor
after the high pass filter, then connecting the super tweeter to the tweeter's high
pass through a single cap. Try it with whatever polarity works better on the super
tweeter.
It saves an inductor and capacitor obviously, not much but simple is better.

I did. A resistor after a filter attenuates both below and above the crossover point, or what ever the parts sum too, so you end up with a peaky response. The T's impedance drops far to low too. The ST's response is only drops 5db from 20khz to 3.6khz, so there are significant problems.

This is what the raw responses look like on axis. It's a bit interesting.
ST = Blue
T = Yellow
W = pink

See what I'm working with here? The ST has way too much low frequency energy to have a shallow slope, and while I would love to cross it low ish, it's distortion below 5k skyrockets.

This is the current prediction for the three way I tested above

here is the schematic

Predicted impedance

 

[alex-z]

I've not decided to test vertically yet and I'm not sure how I'm going to do it in an accurate and meaningful way. Doing a full spinorama looks like a ton of work too. We'll see. I may need a dedicated measuring space, which I don't have right now.

If you have a decent measurement stand, just flip the bookshelf speaker sideways for the vertical responses. With a horizontally mirrored speaker like the SSCS5 you save a lot of measurement time, because you only need 18 horizontal measurements and 18 vertical. You can even cut down the vertical measurements to 8, because past 90 degrees the horizontal and vertical are usually similar, and the initial 0 degree measurement should be identical, if the mic alignment is correct.

 

[Velocipede_Acoustics]

If you have a decent measurement stand, just flip the bookshelf speaker sideways for the vertical responses. With a horizontally mirrored speaker like the SSCS5 you save a lot of measurement time, because you only need 18 horizontal measurements and 18 vertical. You can even cut down the vertical measurements to 8, because past 90 degrees the horizontal and vertical are usually similar, and the initial 0 degree measurement should be identical, if the mic alignment is correct.

Thanks. That makes the process seem easier.

 

[Pete Basel]

A small resistor in series with C4 can be used to control peaking and the shape (damping) of the rolloff.

 

[Velocipede_Acoustics]

A small resistor in series with C4 can be used to control peaking and the shape (damping) of the rolloff.

Indeed it could. The rolloff would be less steep. It's already not as steep as I want it and the peak is fairly tame.

 

[Ericglo]

Did you look at BenB's mod on these speakers?

 

[Velocipede_Acoustics]

Did you look at BenB's mod on these speakers?

Would you happen to have a link?

 

[Ericglo]

https://www.avsforum.com/threads/improving-the-sony-ss-cs5.3068362/

@BenB

 

[Velocipede_Acoustics]

Very Cool. Maybe we can get a new crossover designed for his mods too.

I'm going to work on pulling the working original crossover from my other speaker today and tomorrow to map it out. Then I'm going to study up on getting a real spinorama done, if I can.

 

[Velocipede_Acoustics]

I can't edit my above posts, but I noticed the crossover schematic I posted isn't correct. The tweeter should be in phase. Otherwise, I'm still on that design that I want to get a spinorama of.

 

[Pete Basel]

I've not seen a schematic for the stock crossover but it would obviously cut costs if
components could be reused in the new design. Even if you had to wind down some
of the inductors.

 

[Velocipede_Acoustics]

Yeah I think so. I admit I was a bit overzealous and tore the original one apart and broke the PCB trying to get the parts off the glue. I'll be more thoughtful when I take apart the second speaker.

 

[Dennis Murphy]

I've not seen a schematic for the stock crossover but it would obviously cut costs if
components could be reused in the new design. Even if you had to wind down some
of the inductors.

Here are the schematics. Note that the main tweeter doesn't have a low pass filter. It runs full range at the top, with the "super" tweeter just tacked on with its own high pass filter. I've done a simple crossover mod that remedies the lack of adequate baffle step compensation, but the speaker is just too difficult to disassemble to make crossover mods attractive.