AW: VP-330 (analog speech synthesis)

[Mark Smart]

> The BBD chips have a HF loss all by themselves ?
> I thought it was all because of the anti-aliasing filter,
> but I have never thought any further ... So, do you really
> loose HF in the buckets ?
> 
> JH.

Yep.

My test was very simple. After I got an Omni II it became clear how much
better the chorus sounded on the Omni II than on the Omni I. I poked around
to figure out why. The sampling frequency is actually slightly HIGHER on
the Omni I, so that wasn't it. Listening to different places on the chorus
board of the Omni I, I found that the signal sounded fine when it went into
the analog delay chip, and muddy when it came out (before the antialiasing
filter). It seemed pretty clear that the MN3002 chip was the entire reason
why the Omni I sounded muddy. Since I don't have data on either of these
chips, I can't really speculate as to why this is true. But the 16-pin
SAD512's on the Omni II do not degrade nearly as much between input and
output. I read somewhere on the net that 16-pin SAD512's are actually
SAD1024's which only have one working half. The chips were
mis-manufactured, then the working half was wired to the pins the same way
every time so they could still use them.


************************************************
*     Mark Smart                               *
*     Electronics Engineer                     *
*     NovaNET Learning, Inc.                   *
*     smart at medusa.nn.com                      *
************************************************