Sound Sculpture Switchblade Question

[tommy.genes]

Hey all, popping back in to the forum after a few years away. Looking into a few new projects and hoping to ask a few questions.

Does anyone know what switching method the Sound Sculpture Switchblade uses? JFETs? Relays? Audio switch matrix ICs? I'm not looking for complete schematics, not looking to clone the device etc. But since these are considered to be the "best in class" for this type of device, I'm curious what technique they use.

Thanks,
-- T. G. --

[WhittakerE]

 I suggest reaching out to Sound Sculpture directly or searching for technical reviews and articles that may provide more insight into the device's switching method. Best of luck with your projects!

[tarafoster]

Thanks for the information...   

[tommy.genes]

Yes, thank you for the suggestion. I will share here if I learn anything.

-- T. G. --

[idy]

Here's what we know: It seems Sound Sculpture never talks about what is inside. Unauthorized opening voids warranty...
...so possibly no one who owns one even knows what is inside.

They do say certain things that we can take to mean: all analog signal path with digital control. You can cross fade. So it isn't just relays. It is midi compatible, so it is a computer that controls analog signal paths. There are lots of chips that allow this. Discrete FETs kind of unlikely.

[FiveseveN]

Unauthorized opening voids warranty...

Looks like someone bit the bullet:



Sadly it's too small to make out the chip labels and there doesn't seem to be a larger version.

[idy]

...and there are some relays for things like reverb and amp channel switching, mimicking footswitches...

[tommy.genes]

My first thought was that the round logo on the large chips looked like the Maxim Integrated logo (now owned by Analog Devices)


Based on those being DIP-28 packages, that implies that they are the MAX457X series (1 through 4). This is further corroborated by the fact that these parts specify -90 dB of Off Isolation and Crosstalk, which matches the published specs of the Switchblade.

Why so many? Assuming this is the balanced 8 X 8 version, to do a true crosspoint switch would require 128 separate switches (8 x 8 x 2). With sixteen of those chips, that would mean they are using 8 switches per chip. This adds up since the MAX457X can have up to 11 switches per chip.

Lastly, the MAX457X chips are unbuffered, which implies that a lot of the other devices are op-amps and/or transistor arrays to provide buffering for all the channels.

It actually is an impressive feat of brute-force engineering if you think about it. Thank you to everyone for contributing to the discussion.

-- T. G. --