Hi Jesse,

Things are startin' to look a lot better if we have you onboard :)

Here's the situation, compared to a standard 27c256, the Roland mask
ROMS differ in this regard (i compared schematics).

-(12) of the (15) address lines do not follow the jadec standard
arrangment. However, every address pin except *(1) correspond to
address pins on a 27c256.

-*Roland ROM utilises Pin 26 as a CS (chip select) & Pin 20 as an
address-line. These Pins would have to be reversed if replacing stock
ROM's with 27c256's and also when reading the original mask ROM's.

-Roland OE (aka output enable & chip enable) is active "HIGH", while
standard 27c256's are active "LOW".

-One of the (2) ROM's used for the bulk of the sounds needs an active
"HIGH" for CS (chip select), while the other needs active "LOW".
27c256's are active "LOW" for CS.

Everything else is the same, Data pins D0->D7 line up, VSS & VDD same
and (3) of the (15) address pin's line up to the the jadec standard. I
believe we could design a converter board with not much space and
minimal external parts, some resitors and PNP transistors to invert
the CS/CE states. I already started on listing what pins of the Roland
ROM's correspond to what addresses.

Seeya

Ryan

P.S I need manage once to get the ROM to read, but the data wasn't
useable and I thought it was an error, this was before i thought their
was a compatability issue.

--- In tr-707@yahoogroups.com , Jesse Smith <stoi2m1@...> wrote:
>
> Im not giving up, I have at least three other machines
> that implement the same type of method to protect
> their samples as well. And I may possibly have more.
>
> I do have experience using CAD software for designing
> PCB's I use a program called eagle soft which can
> output the file typ some manufactuers need but I think
> another one is PCB123 (or something like that).
>
> I like your ideas they come from a new angle then most
> approches i have taken in the past. Dont be
> discouraged Ive been looking and messing with this for
> at least a year.
>
> Ill work on some schematics for ways to rearrange the
> address pins and do some talking to see what may be
> the easiest way to switch them.
>
> All of this will not help you with your situation, if
> the address and data pins are all rearranged but the
> address pins are sill in what is suppossed to be a
> Jadec posistion and the data pins are also in a Jadec
> posistion then you should be able to read from it,
> even if the data is jumbled. So the issued lies with
> is the chip powered and is one of the enable pins
> active high instead of low, and is your programmer
> reading properly.
>
> We now your programmer is ok, for you have read other
> chips in with it. So it seems like a chip enable
> problem again, but the pins could be jumblem much more
> then we think, so hunting and poking will probably be
> the best method and build a chart of all possible pin
> variations.
>
> Jesse
>
> --- plutoniq9 <Plutonique9@...> wrote:
>
> > Ok,
> >
> > I desoldered & pulled my TR-727 sound ROMS and
> > thought I'd get my
> > hands dirty with ripping the sound data from 'em to
> > see how they are
> > all organized. After mnay hours of trying this and
> > that to get 'em to
> > read, I went back and looked at the schematics.
> >
> > 9.9 times out of 10, all eproms of same # of pins
> > have an indentical
> > pin layout, this even holds true for parrallel SRAM.
> > It's called the
> > Jadec standard and was created for compatibility &
> > ease of mind for
> > developers...much like many standards.
> >
> > But if your a paranoid company like Roland, and
> > think people are gonna
> > poke their noses in where they don't belong
> > (hardware wise), you'd
> > step away from this standard, re-arrange all your
> > pins (or most of
> > em)....and thus make them incompatible with eprom
> > readers and make new
> > custom chips more than just blowing a new ROM and
> > replacing the other
> > chips. That's what route they took with the TR-707,
> > TR-727, TR-909 &
> > probably the 505 & 626.
> >
> > They've basically implemented a pretty good copy
> > protection system.
> > Basically, in order to rip & burn new ROM's for the
> > 707/727, one would
> > have to build a custom PCB that re-routed (12)
> > address-lines & a CS
> > (chip select) line. In addition, both CE
> > (chip-enable) pin, and a CS
> > of one of the two ROMS, would have to be converted
> > from rolands
> > "active = high" to standard Jadec "active =
> > low".....
> >
> > Not impossible, but we'd need to have some PCB
> > boards made up to even
> > take the first step. But if we were to go this far,
> > you may as well
> > design the board to accomidate a 32-pin eprom for an
> > additional
> > 16-banks of sounds. I see (2) small pcb boards, with
> > a 32-pin DIP
> > socket and 28-pins for soldering into the 707/727
> > PCB + a couple PNP
> > transistors & resistors for inverting the CE & CS
> > states.
> >
> > We'd probably need 10 people willing to pitch in to
> > have the boards
> > made up (because most PCB board producers have
> > minimums), and someone
> > who has experience using PCB design software, most
> > PCB producers
> > actually have their own software they require people
> > to design on.
> >
> > Wow! Is it worth it Are people completley bummed
> > out and are now
> > losing interest, if not who could help on this
> > project. It would suck
> > to talk about it this much and then to give up.
> >
> > Ryan
> >
> >
> >
> >
> >
> >
>