Re: cine sound board

At 09:17 AM 6/12/97 -0500, you wrote:
>> Yup! That's the part! Thanks David.
>>
>> So do you have the full specs on the part? This is the one I played with=
>> in the
>> past, the fact that it uses a "pseudo-random" generator is indicative of =
>> a LFSR,
>> since a noisy semiconductor is a truly random source.
>>
>> -Zonn
>>
>
> The S2688/MM5837 ARE LFSR-based pseudo-random noise generators.
>
> I went to the UTexas library last night, and wasn't able to find a
datasheet, but I DID find a functional block diagram in a reference called
D.A.T.A. (forgot what it stood for)
>
> Anyways, the LFSR is 17 registers long, and the 14th and 17th bits are fed
back to the "Exclusive Or Circuitry" block. All the bits of the LFSR are
NORed together (zero detect.) The output of this NOR gate is also fed into
the "Exclusive Or Circuitry" block.
>
> This is an MOS part, and takes 3 voltages (2 supply -- Vdd and Vss, and
one control voltage, Vgg)
>
> The part has an internal ocsillator and clock driver to generate the clock
for the LFSR. The 17th bit of the LFSR is send through a push-pull output
buffer, and that is the noise output, so it looks like there is no
integrator on the output.
>
> Vgg controls two "Test Gates" which are probably nothing more than MOS
switches. There are two test pins (TESTA and TESTB) on the chip. It looks
like when Vgg is high, TESTA is input to the "Exclusive Or Circuitry" block
and TESTB is input to the "Internal Oscillator" block (Probably if you want
to supply your own clock.)
>
> Anways, it looks like I have what I need to emulate this thing. If anyone
can provide more detailed specs, especially about how that internal
oscillator works (i.e. is it just one frequency, or does it vary in
frequency) I'd appreciate it.

I'll bet if you put a scope on the unfiltered output you would easily see
residue of the clock. I'll also bet it's fixed, but fluxuates a bit with
temperature.

-Zonn
Received on Thu Jun 12 08:50:35 1997