TB303 RAM problems

Trevor Page trevor at resonance.fsnet.co.uk
Fri Mar 31 19:52:13 CEST 2000

Hi all,

I recently picked up a rather mangled TB303. So far I have had to repair the
main PCB which had cracked in half. Someone had attempted to do this with
bits of wire before, but made a messy job of it. Anyhow, the Pitch Mode and
Normal Mode LEDs were always on together - which should not happen. Thinking
the shoddy wiring was to blame, I replaced it all.

Having repaired the PCB, I found that both of these LEDs still remained on -
although curiously the machine did seem to work pretty normally otherwise
from what I could tell. Cutting a long story short, I discovered that one of
the uPD444C RAM chips seems to be at fault. The RAM IC's share the same I/O
port lines on the microcontroller as the mode indicator LEDs. On one of the
RAMs, two of the data lines were permenantly asserted to 5V. These were
causing both LEDs to stay on all the time. Removing this RAM IC restored
normal operation, but with the machine only being able to store patterns in
the upper two banks. Whether or not it can work reliably like this for the
time being I don't know.

My first question is, how hard would it be to get hold of such a device? I
admit I have not tried Roland yet(I made an order from them this morning,
before I discovered the faulty RAM! dammit!). If anyone might have any of
these devices sitting in their junk box, I'd like to buy them off you.

Secondly, are RAM ICs like this particularly prone to failure? I am worried
about the main controller IC; if this ever goes then I'm in trouble. But is
there any reason why RAM ICs might be the first to fail?



More information about the Synth-diy mailing list