I have yet to do the real hardware task of making new sound chips for my 707,
but most of the groundwork is done.

You know how RAM and ROM chips use various pins for functions Data
lines, address lines, write enable, chip select, etc These are typically the
same for a given chip no matter what its application is. Roland did some
unusual things in their drum machines of this vintage (707, 909, probably
626), by switching the numbers of the address and several other pins. This
means, for instance, if one were to remove the sound chip and read its
contents into a computer (as Colin Fraser had done)- the data would be
scrambled. Of course, if one were to make chips of new 707 sounds, those
would need to be scrambled in just the same way.

There are a few ways I can proceed here. Probably the next thing I will do is
remove a few of the SRAM chips which contain the sounds, then put those in
28-pin adaptors so I can read the data off of them without getting too
confused. I guess if I am used to looking at raw audio data I might be able to
piece it together in a text editor, but I'd need to go through variations of this
process every time I wanted to read an IC, or write new samples. I have
decided to just build adaptors and study the data stored on the SRAMs.

I have not yet decided whether I want to use more SRAMs (62256), or make
adaptors for EPROMs (27256). I have heard that SRAMs can be quite a
headache, erasing themselves before making it to their circuit destination.
Also, the intel-type EPROMs are ubiquitous. With 62256 I could just drop them
in to try sounds (if I burned them scrambled). For 27256 I'd need a special
adaptor. I am leaning towards using a perfboard on the bottom of the case
with 28-pin sockets on it. The sockets could be connected to the SRAM spots
on the board with the wires arranged for the scrambled layout. I could then
just burn ordinary EPROMs and put them in with no further problems, once I
set it up. Another advantage is that the perfboard can face down towards the
bottom of the case- so once the bottom plastic cover is off, the chips can be
easily replaced without needing to get to the main PCB.

The annoying dilemma now is that I have no EPROMs or UV eraser. But I do
have 62256 SRAM chips. But >why< would I want to install sockets onto the
board to try them if I am only going to remove them soon later I am impatient
to try making new sound chips, but I feel I should wait until I get the reast of
the EPROM tools which I need for this and so many other projects.

I am trying to figure out how to make a chart to show how the chips are
scrambled, and how to make adaptors for them. I am busy, so it might take me
a week or so.
CJ

--- In tr-707@yahoogroups.com , "acousmatique" <acousmatique@y...> wrote:
> > Once I know how the drum sounds are layed out in memory, I can put
some
> > new ones! CJ
>
> Now y'all have done it... I broke down and ordered a basic EPROM
> programmer to jam this with. In theory, 62256 SRAM should work.
>
> Yes, the sound locations are in the service notes, on page 7 of the scans I
> have on hand.
>
> We should be psyched that Colin Fraser has saved us from reinventing a
> tremendous amount of wheel. On his site he describes in detail how he
made
> clones of the 909 cymbals, which just happen to be done on the same chips
> in a nearly identical circuit. Some mods are neccesary in order to use
garden-
> variety 27256 EPROMS, and those are described here.
>
> < http://www.colinfraser.com/tr909/909cyms.htm >
>
> Colin had more leeway, as he didn't need to fit this on a real 909 board. It
will
> require some work to apply this towards dropping an EPROM into a 707
and
> having it work.
> Laterz