On Thursday, May 7, 1998 at 3:00:00 AM UTC-4, Paul Joannis wrote:
A few months ago, someone posted an article concerning a
failure on a D-10 or D-110 from Roland. I posted back saying I
had experienced the same failure intermittently myself, and
had to reset the unit to get it to work. Well, now the failure
is constant, and I was able to fiddle a little bit with the
keyboard to get it to work properly.
First, the failure: Upon turning the unit on, you always see
the upper row on the display light up, and after a second it
comes up 'Roland D-10 Linear Synthesizer'. When the failure
occurs, the upper row stays lit, and the message never
appears. This is sometimes accompanied by a deafening white
noise out of the analog outputs, as if all synth voices were
putting out garbage at the same time.
As I had previously posted, I suspected the problem to be
related to the central processing circuit. When you turn the
keyboard on, the program that's burned in the eproms and runs
the whole show makes (among other things) that little message
appear. If that doesn't happen, then the processing doesn't
happen. This is not an audio failure, even if the noise comes
out (it is probably just the way the signal generators react
when the unit is not booting up properly).
So, when mine began doing so permanently, I opened it up.
First, I checked the battery, 3.22 volts, fine. Then, I
cleaned out the dust and the components with standard head and
contact cleaner (Radio Shack, 8 bucks a can). Then I looked at
the main processor. It is a square chip (about one inch
square) on a black socket, Intel 80 series microcontroller. I
recognized this kind of socket, since my company used to use
them in some of our assemblies, and here's the problem with
them: these are 'Surface Mount' components, and that
particular socket makes a poor contact when the chip is
pressed all the way in. Looking at the chip in profile, the
contact is sort of like this :
chip pin -> _( <- contact on socket
When the chip is inserted all the way down, the pin makes a
poor contact on the bottom of the socket. By placing a small
piece of paper under the chip (either an ordinary piece of
paper folded 4-6 times, or a piece of cereal box cardboard or
two), you get something like this:
chip pin -> -( <- contact on socket
It's hard to explain on the internet, but all I can say is
I've seen this before on other circuits, and it works.
Carefully remove the chip form the socket by gently lifting it
out corner by corner with a small jeweller's screwdriver. Note
orientation (put it in backwards and you've just blown up your
unit!!! - one of the corners of the chip is cut off a bit, and
it is the same on the socket. Match them up!!!). Clean off
pins and socket with contact cleaner and toothbrush. If you
can connect your body to a building ground while you do this,
you will avoid electro-static damage. Place small cardboard or
folded paper inside socket, and insert chip in (with correct
orientation!!!) and snap into place. Make sure carboard is not
to thick, as chip will tend to pop out and you will have the
same problem.
It is possible that cleaning alone can alleviate this problem,
no need for the cardboard. However, just for good measure, I
also removed the two software eproms. These are DIP types,
side by side, with pins pointing down and a Roland Label on
them, a little over an inch long and 1/2 inch wide - just
remove carefully (note orientation!) by lifting on either side
with a small screwdriver, clean off socket and pins as above,
and reinsert carefully, noting orientation and not bending any
pins underneath - that would cause a similar problem again. If
any pins on the microcontroller that runs the program or the
eproms that contain the program aren't connecting, then the
synth will never boot up and diplay the message.
A faulty power supply might also cause this (not enough power
to the CMOS circuit will cause them not to work - as little as
a quarter of a volt DC can do this) but this is unlikely -
Roland stuff usually have good power supplies.
I can't garantee this is the perfect solution, nor do I
pretend to know everything. All I can say is, before I did
this, mine never worked, and since I've done this, it always
works fine. Maybe it just needed a good cleaning, but whatever
does the trick...
--
I've been looking for a solution to this for weeks! Thank you so much!
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)