AMZ FX ANALOG ALCHEMY GEOFX MEF ROG DIY WIKI

OT: 4000 series ICs - how do you determine if they're short?

Started by Gil, October 25, 2003, 07:13:08 PM

Previous topic - Next topic
Print
Go Down Pages1
User actions

Gil

October 25, 2003, 07:13:08 PM
I'm in a process of repairing a vintage analog synth, and I read an article about recommeded steps for repairing analog synths, which says to replace all the ICs of the 4000 series, as they may get shorted over about 15 years period. I got about 6 chips of this series and before I get to replace them, I'd like to know if there is a way to determine if such IC is short or misfunction before going to replace it.

The ones I got in the circuit are :

CD4001
CD4011
CD4049
CD4069

Thanks.

http://www.oldcrows.net/~oldcrow/synth/tips.txt

Paul Perry (Frostwave)

#1
October 25, 2003, 07:51:03 PM
WOAH! ***NEVER*** replace chips in an old synth if they aen't busted already!! You'll likely wreck somethng as you go along. It's rare for digital chips to crap out on their own, and no guarantee that a 'replacement' chip will be a genuine replacement (varying edge rise times etc).
Now some op amps could be replaced with quieter versions perhaps.. but again, I wouldn't touch it if it works!
Likelier source of problems is filter caps.

Peter Snowberg

#2
October 25, 2003, 08:04:32 PM
I've never heard of CMOS chips going dead after time. In hundreds of years we can expect all semiconductor devices to be kaput, but 15 years??? I don't believe it. :?

The easiest way to check those chips is with a logic probe. If you don't have a logic probe, you can easily make one with a CD4049 or any CMOS logic gate that can drive an LED. LEDs take more current than CMOS logic wants to give so make sure you use a 4049 or 4069.

It helps to use clear, high brightness LEDs for this so you can watch for quick pulses. Wire one led from power to a resistor then to output A. Wire a second LED and resistor from power to output B. Wire input B to output A. Connect a 4.7meg resistor from ground to input A, then connect a 220K resistor to input A and leave the other end flying. That flying resistor tip is your probe.

Look up the data sheet for each of the chips in question by searching for "cd40xx datasheet" at google and then carefully us the probe to check each of the pins. If you see inputs and no outputs, it's a dead chip. You'll see outputs stuck at + or ground quite often. In any case the data in should cause data out and you can check to make sure that at least something is happening and that it makes sense according to the function of the chip.

-Peter
Eschew paradigm obfuscation

Gil

#3
October 25, 2003, 08:06:31 PM
Are you referring to power supply filter caps ?

The problem is that the synth plays a scattered tone randomally,
that is changes when tweaking the EG element on the VCF modulation.
As I lower it's value the scattered tone rate is slower.... this happens only when hitting certain key combination, randomly. Any idea ?

Paul Perry (Frostwave)

#4
October 25, 2003, 08:10:26 PM
Might be tarnishing on the key contacts,
Which synth is it?

Tim Escobedo

#5
October 25, 2003, 10:55:10 PM
Actually, there seems to be a lot of conjecture about the "lifespan" of many semiconductors. Some of it seems alarmist, particularly of the types of devices folks around here use. I think most of the concern is over more modern designs and manufacturing process. Still, it's kind of interesting to keep an eye on.

Gil

#6
October 27, 2003, 12:56:07 PM
QuoteI wouldn't touch it if it works!

Well.... it doesn't actually...

As I said the problem started as the NICD battery leaked,
than the synth started to play random notes which I relate to the acid that had leaked over and which probably shorted things which make it to play automatically.... Now I scratech the area on the computer board (which holds the battery) and recover the whole area by point to point wiring and replacing all parts of course.

I probably defected the LS367 that is responsible for the operation of the alpha dial wheel, but fixed it as I replaced it.

The problems that still persist are -

for the 8th note that is assigned, the EG attack and decay (and maybe more) doesn't function (it ignored those parameters like they are not there at all).

Also when playing some key combination the tone will get scattered, and the rate of that "scatterness" will slow as you lower the value of the EG paramenter at the VCF "Modlulation" scetion.

I'll really appreciate any help as I don't want to start replacing ICs without knowing which ones are probably the defected ones...

I would attach the schematics and block diagram of the board if there was such an option on this board...

Some more info - there's a LS374 chip which is connected directly to two ULN2003AN chips that where probably shorted by the leaking acid,
and are all at the section described as "Gate out and EG pulse out".
Do you think that it's the one causing the problems ?
Or any IC's on the Key Scan in and out ?

Thanks in advance!

Peter Snowberg

#7
October 27, 2003, 04:45:42 PM
You can get a free web site without pop-ups for putting schematics on at:

http://www.freewebs.com/

You only have to include a small tagline at the bottom of each page and you can't link directly to pictures from here, but you can always link to a picture page on that server.

It's up most of the time....
Eschew paradigm obfuscation

Paul Perry (Frostwave)

#8
October 27, 2003, 08:05:28 PM
I would be very suprised if any of the chips were damaged by the acid. More likely, either traces are broken by acid, or effectively shorted by moist deposits of metal salts (which mah have seeped under chips & caps). So it is more a painstaking cleaning and connection checing task, probably.

Gil

#9
October 28, 2003, 02:28:53 AM
Well, I actually re-built the wole battery area, i.e took all parts off, *scratched* the upper layer so there's only  the plastic left. Than rebuilt with new components using point-to-point wiring (the same technique I use to build my DIY  effects) so I guess this is not moist... I cleand the PCB a dozen times so far ! One thing I noticed is that one EPROM leg (#24) doesn't read the correct voltage (1.8v instead of 5v ....)

Gil

#10
October 28, 2003, 02:40:03 AM
To get the schematics,

go here :

http://photos.walla.co.il/gilwein

Click on "schematics"

There are the block diagram and schematics of the computer board (the problematics one holding tha battery)

Thanks all !

Paul Perry (Frostwave)

#11
October 28, 2003, 06:43:42 AM
Gil, that link didn't work for me..

Mike Burgundy

#12
October 28, 2003, 06:56:57 AM
Just a quick idea:
A NiCd battery in a AC-fed synth? Sounds like memory chips in there - which will be wiped if the battery fails. Most likely you've just lost your presets, but it would be a good thing to rule this out (is an empty preset set to "0"?)

Paul Perry (Frostwave)

#15
October 31, 2003, 08:01:45 AM
Here's how to subscribe to synth-diy:
www.euronet.nl/~rja/Emusic/Synth-diy/
and here's how to sub to analog heaven (all about using & maintainig analog synths).
Maybe someone there can help!

Paul Perry (Frostwave)

#16
October 31, 2003, 08:05:52 AM
Here's how to subscribe to synth-diy:
www.euronet.nl/~rja/Emusic/Synth-diy/

and here's how to sub to analog heaven (all about using & maintainig analog synths).
machines.hyperreal.org/Analogue-Heaven/

Maybe someone there can help!
Print
Go Up Pages1
User actions