Archive of the former Yahoo!Groups mailing list: Discussion about the Korg PolySix synthesizer

previous by date index next by date
previous in topic topic list  

Subject: Re: [PolySix] checking connections

From: <backshall1@bellsouth.net>
Date: 2018-07-16

I absolutely agree with this. A lot of the bad traces on the board just look like they go to a “via” that goes through the board to the other side. Unless you test all the way from chip to chip, you haven’t really proven anything. Since most of the damage is on the D00-D07 data lines and there are eight of these, and each one goes to multiple chips around the board as well as a couple of connectors to other boards, there is a whole lot of testing to do. I think it was Andy Jury who first mentioned to me that he tests everything from connector CN10 back to all of the other chips. That means that in order to prove D00 is working, the following continuity checks must be done:
CN10-4 to IC31-6
CN10-4 to IC26-13
CN10-4 to IC28-4
CN10-4 to IC33-12
CN10-4 to IC34-4
CN10-4 to CN06-3
And that’s just for the first data line out of eight. I have been using this method ever since and it’s really the only way to make sure you don’t miss anything. Oh, and since the two 74LS08 chips look like they are in bad shape, there is likely damage on the DB0-DB7 lines as well, between the 8048 microcontroller and IC30/31
 
Don B.
 
Sent: Monday, July 16, 2018 11:34 AM
Subject: [PolySix] checking connections
 
 

When checking for continuity of questionable traces, IMO it's best
to figure out which pins of which IC's are being connected by the suspect
trace. Then use the DMM to measure directly between the involved pins.
Connecting the pins of the IC's together is the ultimate goal, so it's best
to confirm that it's being achieved. In other words, you want to check from one
end to another end, since that's what really matters. Also, trying to get a good
connection from your probe to a corroded trace is not easy.

Really, on something like the PolySix, what I would suggest is to print out the
schematic of just that one board, and "buzz out" with a meter the connections in
the area that is corroded. Highlight the traces that have been tested as you go.
If you just try to test the traces that you notice that look bad, it's easy to miss
something.

My $0.02

Bob