[sdiy] 909 hh sample selection

[Colin f]

> i'm knee deep in 909 mods, and i'm wondering how the hi hat 
> samples are 
> selected.

> circuit details (once again) are at 
> http://hardware.freepage.de/raf909/ 
> under "circuit analysis" and "hihat and cymbals."

> from the description, d196-198 sounds like the area i need to 
> alter, but i 
> don't understand how that configuration of diodes is working. 

The important thing to note is that the closed hi hat sample gets one
quarter of the ROM, and the open hi hat gets three quarters.

There are two logic signals shown on the schematic - RUN/~STOP and
CLOSED/~OPEN
When RUN is high, the clock runs and the sample plays back.
When CLOSED is high, the closed hi hat sample is selected, if it's low, the
open hat plays.

The 4040 and 4520 counters generate a 15 bit address, driving the sample
ROM.
But the top two bits of the address don't come directly from the counters.
The diodes D192-198 OR the top two address bits with the CLOSED line.

This means if the open sample is selected, the top two address bits could
00, 01, 10 and 11.
But if the CLOSED line is high, then the top two bits will always be 11.

The way the inputs for the NAND gate that generates the RUN signal is wired,
if CLOSED is low, RUN will go low when the top counter bits reach 11. So
when a trigger resets the counters, RUN goes high, the counters step through
addresses 000000000000000 to 110000000000000, at which point RUN goes low
and the sample stops.

If CLOSED is high, RUN will go low when the counter bits reach 01 - the top
bit input to the NAND gate is forced high by CLOSED.
But since the ROM has the top two address bits forced to 1 when CLOSED is
high, the closed sample plays addresses 110000000000000 to 111111111111111.

Convert the addresses to decimal, and you'll see the open hat is the bottom
3/4, and the closed is the top 1/4 of the ROM, with one byte of overlap.

Cheers,
Colin f