TTL SRAM interfacing

[Ingo Debus]

cyborg0 at GlobalEyes.net wrote:
> 
> I have perhaps what is a somewhat unusual question.
> I would like to make a analog step sequencer with memory.
> I just wonder if what I am about to try is possible, or if there is a
> better way. I would also like to do it without a micro.
[...]
> I would like to know if it is feasible to just feed the output of the
> binary counter into the address lines of SRAM, and, in turn, have the
> SRAM data out feed into a ADC.

:-) This is how my first sequencer back in 1979 (or so) worked. I used
two 256*4 RAMs. Four bits were for CV, one for gate and one for reset
(so the sequence could be shorter than 256 steps). Input was done via
the synthesizer keyboard, I added contacts at 16 keys and used 74148s to
generate the data for the RAM. I used a self-made 4 bit DAC for CV
generation, just a summing opamp and R, 2R, 4R and 8R resistors. There
was also a row of 16 LEDs above the keyboard, driven by a 74154 decoding
the SRAM data, so the key "played" by the sequencer was displayed.
It was very uncomfortable to use. To enter a sequence, I had to press a
keyboard key, a "write" button, eventually a "gate" button, and a "step"
button for every step. Memory was volatile (NMOS RAMs).
I had plans to extend it to 5 bits CV resolution, but then I decided to
build a polyphonic synth with a microprocessor, so the sequencer thing
could become much easier.

No, I don't think I still have schematics of that sequencer.

Ingo