Optocoupler Teaser

Hi All,
I've been in contact with a guy who has a CNC machine who's controller uses
optocouplers. Here;s a picture:
http://home.fuse.net/consultec/misc/images/optocouplers_.JPG

The one on the right works and there WAS an identical one on the left that
failed and the manufacturer sent a new one that looks different and doesn't
work. You can see the new one on the left does have a different design. Upon
finding the new one doesn't work, the manufacturer sent another, identical
to the first one they sent and it still doesn't work. The guy is convinced
they're sending the wrong style because he's swapped optocouplers and
wherever he puts the original (the one on the right) it works.
Is there something obvious about what is going on? Is one for RS232 and the
other for TTL or 0-20milliamp? Is the output side of the chip the problem?
Is there enough info here to even make a determination?
This is beyond me, hoping someone recognizes the two different circuits and
can give some input.
Thanks in advance,
Bart

 

Well, it would help if you could supply the part numbers of the two opto
couplers. Then it would help also if you could tell us what they get at
their inputs. I see a 1k resistor at the left. If we assume about 10mA
minimum for the opto coupler LED for it to work then you would need at least
about 11.5V at the input (10mA*1k + Vled). But there is another LED, looks
like it is in series with the opto coupler LED, so that would add another
1.5V, say. Do you have that much (13V) at the input? If not then you should
probably replace the 1k series resistor by a lower value one.

--DF

 

Here is my simple take on it so far,
POINT: The working one (on the right) uses a voltage divider on the input
side so maybe some characteristic of the input side it was designed for
makes
the new/replacement one (on the left) fail immediately up energizing.

COUNTERPOINT: They both use what looks like 1/4 watt resistors and
TO-92 devices which probably means we're just dealing with smaller simple
signals like data transfer. Even the optocouplers I looked up on DigiKey
were rated in the milliamps so I don't think the new/replacement ones are
failing, they're just designed for a different application. Physically seen
differences are apparent on both the input AND output side.

MURPHY'S-POINT: The boards have just one simple task, when one
side is energized (yes, low power) then the other side is allowed to have
IT'S
energy flow (also low power). One board works and the other one doesn't.
Non-working board has transistor mounted backwards from working board.
Non-working board has transistor mounted backward from silkscreen image
on board. Non-working board has never worked, will never work, and
slipped through Brazil's (yes, that's where its made) stringent quality
checks.
My guy has received total of three of these replacement boards, all
identical,
none work.

Did they run out of NPN transistors and thought they'd use PNP and just
mount
it backwards?

Glad I'M not paying for the replacement boards,
Bart

 

hmm interesting devices? i noticed A1 A2 ? these for brush motors
by any chance?
in anycase, i did notice the transistor in the left unit looks
like its not in there correctly if you were to follow the board lay out?

 

That doesn't work. You have to pull them inside out to reverse them

 

Oh, I see, you call the whole board an opto coupler, I thought you meant the
6-legged components on the boards. So he swapped boards, not just
components. Ah, language...

--DF

 

My bad...you're right, I should have called it an optocoupler interface
board
throughout the post. I'm still working on the exactness of vocabulary this
discipline demands. Funny thing, when I WAS married my wife would
blunder with words, confusing me to where I'd ask a simple question
for clarification....that's when she'd exclaim with head cocked and hands
on hips "You KNOW what I meant!"
I never want to be like that,
Bart