Crumar Musical Instruments User Group.

And thank you for your great advice. I already had mine offered for sale with some defects. 
But I just couldn't leave the thing alone. I had a nasty hiss coming out of it and 
occasionally a voice 3 and 4 far too loud. I checked the datasheet of the 3328's and found 
some strange voltages on the 4th 3328 which I could trace down to a resistor near the 
LM13700N that needed resoldering. Voice 3 and 4 back were okay now. In a less 
professional way (touching everything possible point) I was able to find another point to 
be resoldered and now It's fully functional. INo more hisses and crackles. Not even one 
component replaced!

--- In crumar@yahoogroups.com, "neurokrash" <spamkiller@...> wrote:
>
> Yeehaw!
> 
> The CEM3328 Bit is now fully repaired.
> 
> This one had some bad CEM3328 filters, bad TL082 op-amps, and then was
> mostly functional aside from being out of tune and having one voice
> still out.  I decided that the problem was probably not from the lower
> filter board, since switching things around down there didn't change
> the bad voice, so I looked at the upper board.  There I found another
> row of TL082 op-amps, so I socketed these and began moving the chips
> around.  The bad voice followed a chip, so I replaced it and presto!
> 
> It is worth mentioning that some of the TL082s were running pretty hot
> - about 110-120 degrees f.  One of them was over 220 degrees!  I
> swapped these out until I got a full set that were 100 degrees or
> below.  It's 82 here, so that seemed acceptable ;)
> 
> I tuned the wacky voice by adjusting one of the potentiometers on the
> left side of the upper board, with the detune slider in the middle. 
> Seems okay now, but I'm not sure why it went so far out of tune(?)
> 
> Here are some of the things that I learned while repairing the Bits:
> 
> 1) Your finger is not a good diagnostic instrument.  I found this out
> by trying to check the temperature of an IC that was running at around
> 220 degrees.  Bad idea, get a temperature probe for your meter
> instead.  I got a meter with the temperature probe for $30 on ebay.  I
> would guess that you can get a probe that plugs into any meter for
> less.  If you really have money, Fluke makes a nifty device that uses
> a laser and infrared, so you point a laser pointer at the chip, and it
> tells you how hot it is!  I didn't spring for this, but may in the
> future.  In any case, finding chips that are running inexplicably hot
> is very helpful.
> 
> 2) Print out the datasheets for ICs that you plan to test.  It
> definitely helps to know where to look for supply voltage, ground,
> output voltage, etc... Everyone's help from here in the forum was
> invaluable in getting me started.
> 
> 3) Check the ribbon cables.  Some of the cables do not have connectors
> on both ends, but instead are soldered to the board.  Many of these
> failed while disassembling and assembling the keyboard.  They are more
> than 20 years old, so I can't complain - but if you take the keyboard
> apart, then put it back together and things get worse, be sure to
> check these since the little exposed wire likes to break.
> 
> 4) If you don't know what to test (like me) socket the board ;)  I
> didn't know how to read the schematics well enough to clearly trace
> the circuit path for each voice, so I desoldered all of the ICs on the
> board, put in sockets, then started to switch the ICs around and/or
> leave some out (with the power turned off!).  I could then test each
> voice individually and figure out which chips controlled which voices.
>  That made it possible to narrow down the problems, and to find the
> bad parts.
> 
> Thanks to everyone for all the help!
>