Maker Pro
Maker Pro

scavenging opto-isolators

I

ian field

Jan 1, 1970
0
Pete Wilcox said:
Hence my previous comment about using a tiny piece of duct tape over the
gap to block out ambient light sources. Do try to keep up!

Cheers,
Pete.

Why go to all that trouble when suitable parts are easy to find in scrap
consumer electronics and are not expensive to buy new?!
 
D

Don Bruder

Jan 1, 1970
0
Pete Wilcox said:
Same difference. Alternative usage, perhaps, but essentially the same
type of device.

Cheers,
Pete.

Side note since the OP is scavenging, and rodents have been offered as
one possible source:

Most (if not all) modern rodent opto-coupler/interrupter units are
equipped with *TWO* separate detectors per opto unit ("A" and "B"), with
two separate units (one each for X and Y axis) per rodent - That's how
the rodent figures out which way the wheel is spinning. With two
detectors, determining direction is simple: A sequence of "pulse on
output X-A followed by pulse on output X-B" equals "1 pulse +X", while
"pulse on X-B followed by pulse on X-A" means "1 pulse -X". Ditto for
the Y-axis pair. Rodent keeps a running tally of the X/Y values and
periodically (either via "Hey, the mouse has moved!" interrupts, or when
polled by the computer) reports them to the computer, which then deals
with translating the counts into a cursor position.

If you're using one of these as a simple yes/no indicator in a system
with a fairly large motion, it should be no big deal, as the two
detectors are generally mounted so close together that it doesn't
matter, and you can use either the A or B output interchangeably.
However, if you're looking at really small amounts of motion, you may
need to pay attention to the separate outputs - or perhaps OR them
together so that if either pulses, you get a pulse regardless of which
one it was.

If you're using them for data transfer (turning the light source on and
off to transmit data across the air-gap perhaps) you also need to be
aware that *SOME* of them used not only two detectors, but two emitters,
and in some cases, those twin emitters were wired to allow them to be
turned on and off individually. Plan accordingly...
 
A

Allan Adler

Jan 1, 1970
0
Michael said:
If you don't need pin-compatible and can bear a bit larger footprint, look at
floppy drives as other suggest.

In an earlier posting, in another thread, I mentioned scavenging a 3.5"
floppy drive and other parts from a PC I found on the street. So, maybe I
can take it apart to study it and use parts from it.
By the way, the motor that moves the heads in that drive (made by Panasonic
for IBM) is a big, beautiful, Sankyo stepper. 1.8-degrees/step. I've never
done a hobby project with a stepper but this nice motor has inspired me to
do something with it.

Thanks, I'll keep that in mind.
 
A

Allan Adler

Jan 1, 1970
0
Don Bruder said:
Side note since the OP is scavenging, and rodents have been offered as
one possible source:

Most (if not all) modern rodent opto-coupler/interrupter units are
equipped with *TWO* separate detectors per opto unit ("A" and "B"), with
two separate units (one each for X and Y axis) per rodent - That's how
the rodent figures out which way the wheel is spinning. With two
detectors, determining direction is simple: A sequence of "pulse on
output X-A followed by pulse on output X-B" equals "1 pulse +X", while
"pulse on X-B followed by pulse on X-A" means "1 pulse -X". Ditto for
the Y-axis pair. Rodent keeps a running tally of the X/Y values and
periodically (either via "Hey, the mouse has moved!" interrupts, or when
polled by the computer) reports them to the computer, which then deals
with translating the counts into a cursor position.

If you're using one of these as a simple yes/no indicator in a system
with a fairly large motion, it should be no big deal, as the two
detectors are generally mounted so close together that it doesn't
matter, and you can use either the A or B output interchangeably.
However, if you're looking at really small amounts of motion, you may
need to pay attention to the separate outputs - or perhaps OR them
together so that if either pulses, you get a pulse regardless of which
one it was.

The more I learn about opto-isolators, the more ways I expect I'll be
interested in using them.
If you're using them for data transfer (turning the light source on and
off to transmit data across the air-gap perhaps) you also need to be
aware that *SOME* of them used not only two detectors, but two emitters,
and in some cases, those twin emitters were wired to allow them to be
turned on and off individually. Plan accordingly...

The original application is for a breakout box for the game port of a
sound card, just in case I need it.

There are two emitter-detector pairs on the board. There are no identifying
marks on them. Each pair has a part with a black head and a separate part
with a transparent head. I have no idea how to identify them and thereby
obtain spec sheets for using them. The Art of Electronics has a page showing
the most common types of opto-isolators but no pictures of devices. My best
idea at the moment with regard to the ones in the mouse is to figure out how
the ciruit works and work backwards to figure out what kinds of voltages and
currents are being applied in connection with the opto-isolators.
 
A

Allan Adler

Jan 1, 1970
0
Michael said:
If the PCB is phenolic you could easily break it in such a way that you could
get at the device you want to scavenge with a cutting tool, e.g. end-cutting
needlenose.(I have a pair of "end-cutters" that I got in the Air Force in the
1960's and that I reserve for this very use.) Snip away at the PCB material
until your device is free,its pins standing proud with nothing on any of them
except a solder ball with pad and a bit of board.

So far, these suggestions are helpful and reasonable. It looks to me like
phenolic but I'm not really sure.
Grasp device firmly with one hand, heat solder with iron in the other hand,
and when solder melts shake the device hard to throw off the PCB+pad. This
works best only when pins are straight, not bent against the PCB. Done this
operation zillions of times (with many different parts) over the past 40
years and it's now second nature!

As you point out, this isn't entirely safe, although it can be made safer
with goggles. Also, it makes it harder to clean up afterwards. But there are
things like absorbent threaded things to absorb solder. Also, there might
be purely mechanical ways to remove the solder and we might discuss that
in a separate thread.

However, I've been thinking that maybe I don't really have to remove the
optoisolators from the PCB. When Hewlett-Packard made the thing, they solved
a lot of design and construction problems that it would be hard for me to
solve in my own situation. So, it might be better for me to see whether I
can simply get the mouse to do something other than be a mouse. Here are
some ways it might be possible to make more of the useful features of the
mouse available to me:
(1) Use a soldering iron to remove components *other* than the opto-isolators.
That gives me the opto-isolators by themselves on a printed circuit board
that was designed to use them. They'll be properly placed, etc.
(2) Put the mouse back together. That will shut out the light more
conveniently than duct tape.
(3) Study an enlarged photocopy of the printed circuit to see how the layout
might be exploited by placing other components on it or attaching wires
to it.

I think I should do (3) first since it might tell me that I don't really
have to remove *all* the components other than the opto-isolators.
 
A

Allan Adler

Jan 1, 1970
0
Allan Adler said:
some ways it might be possible to make more of the useful features of the
mouse available to me: [snip]
(3) Study an enlarged photocopy of the printed circuit to see how the layout
might be exploited by placing other components on it or attaching wires
to it.
I think I should do (3) first since it might tell me that I don't really
have to remove *all* the components other than the opto-isolators.

I photocopied the printed circuit with a copy ratio of 4:1, and found it
wasn't very legible. I'll try to borrow a digital camera and take a picture
of the circuit. That might give me better enlargement and more control over
the quality of the picture. Meanwhile, I'm going to sit down with the PCB
and try to take a careful inventory of the components on it. I don't have
access to a multimeter at the moment so I can't conveniently determine which
components are connected to which, but maybe I can eyeball it with a
magnifying glass.

Does anyone have any idea what the IC on the chip is and where I might be able
to get a data sheet for it? The markings are:
EICI1274000
EM01 D 9948
I've googled it but can't find any online data sheets for it.
 
B

Baron

Jan 1, 1970
0
Old printers and computer PSU's are an excellent source.
 
A

Allan Adler

Jan 1, 1970
0
robb said:
hi Allen,

I've had a recent education by experimenting with finding
datasheets for obscure/obsolete part numbers for reuse parts and
pulls.

I find that if you find a datasheet archive / search engine site
and you use just the number portion or maybe the first letter
before the first number and include 3 mor 4 numbers you can get a
pretty good hit rate on at least an equivalent MFG part maybe not
exact maker but probably a compatible drop in replacement The
user (you) will decide if the **near find** is hit/miss
sometimes it is enough info to give some guidelines on usage and
circuit concerns etc

Example...

for --- EICI1274000

i went to "alldatasheet.com" and used "I1274" search parameter an
d got 12 hits
where 3 were Opto devices and the others were not related
here is one of the opto hits
http://www.alldatasheet.com/datasheet-pdf/pdf/140456/OPTEK/OPI1270-032.html

Thanks for these helpful suggestions. I went to alldatasheet.com and
searched for EICI1274000 and it automatically truncated my search to
EICI127 and said it couldn't find any. I realized also that I had been
led to this precise page by going to one of the places that said it
sells them and clicking on data sheets for it.

This might be something that Hewlett-Packard has made specially for
itself and has no circulating data sheets. Do you think that is likely?

But the trick you mentioned of just searching for the number plus maybe
one letter is something to keep in mind.
 
A

Allan Adler

Jan 1, 1970
0
Allan Adler said:
I photocopied the printed circuit with a copy ratio of 4:1, and found it
wasn't very legible. I'll try to borrow a digital camera and take a picture
of the circuit. That might give me better enlargement and more control over
the quality of the picture. Meanwhile, I'm going to sit down with the PCB
and try to take a careful inventory of the components on it. I don't have
access to a multimeter at the moment so I can't conveniently determine which
components are connected to which, but maybe I can eyeball it with a
magnifying glass.

I've made a little progress, which also raises a few questions. One side
of the board has the printed circuit layout and the other has the big
components. The layout side has the following markings on it for components:
C2,C3,C4,C6,C8,C9,C10,C11,C12
R2,R3,R4
DZ1,DZ2
JP1, JP2, JP3, JP6, JP7

The component side has the following markings on it:
C1
R1
S1, S2, S3
U1, U2, U3, U4
JP6, JP7
CN1

Taken together, that gives:
C[1-12] - {C5,C7}
R[1-4]
DZ[1-2]
JP[1-7] - {JP4, JP5} with JP6,JP7 each occurring twice and maybe one of each
of the redundant JP6,JP7 supposed to be JP4,JP5 resp.
S[1-3]
U[1-4]
CN[1]

The letters seem to mnemonics or abbreviations. C for capacitor, R for
resistor, DZ for zener diode (?), JP for jumper, S for switch, CN for
connector, U for unusual component.

There are components on the layout side but they are very flat. One of the
DZ's looks like it might really be a diode. The resistors, capacitors and
jumpers on the layout side all look like black boxes of some kind, but
the jumpers have a O written on them, the resistors have numbers such
as 513 and 330 written on them (actually, they are block lettering and
upside down might spell some acronym like DEE or EIS) and the capacitors
have nothing written on them. Could these weird components be surface
mount technology? They are visibly and abundantly soldered.

The switches for the left and right buttons look a little like car batteries
and have written on the side

1A 125V AC
C NO NC

and each of C, NO, NC is above one of 3 leads going into the board. It appears
that the C and the NC are connected by the layout. The third pushbutton
switch looks more like a cubical building and is unmarked. The track wheel
has one end of its axis resting on that third switch but the other end of
the axis goes into something with a hole in the middle and which is apparently
turned by the rotation of the track wheel. Maybe it is a tiny generator? It
has three leads going into the board.

The labels are much bigger than the connections in the layout and it is
hard to be sure what refers to what. Also, I'm not sure what is connected
to what yet. The fewer components remain unidentified, the easier it will
be to guess the role of the others.

Since mice are abundant, one solution is to remove the components and
test them. I don't have any test equipment of any kind at the moment, but
it would be a reason to acquire some.
 
D

Don Bruder

Jan 1, 1970
0
Allan Adler said:
Allan Adler said:
I photocopied the printed circuit with a copy ratio of 4:1, and found it
wasn't very legible. I'll try to borrow a digital camera and take a picture
of the circuit. That might give me better enlargement and more control over
the quality of the picture. Meanwhile, I'm going to sit down with the PCB
and try to take a careful inventory of the components on it. I don't have
access to a multimeter at the moment so I can't conveniently determine which
components are connected to which, but maybe I can eyeball it with a
magnifying glass.

I've made a little progress, which also raises a few questions. One side
of the board has the printed circuit layout and the other has the big
components. The layout side has the following markings on it for components:
C2,C3,C4,C6,C8,C9,C10,C11,C12
R2,R3,R4
DZ1,DZ2
JP1, JP2, JP3, JP6, JP7

The component side has the following markings on it:
C1
R1
S1, S2, S3
U1, U2, U3, U4
JP6, JP7
CN1

Taken together, that gives:
C[1-12] - {C5,C7}
R[1-4]
DZ[1-2]
JP[1-7] - {JP4, JP5} with JP6,JP7 each occurring twice and maybe one of each
of the redundant JP6,JP7 supposed to be JP4,JP5 resp.
S[1-3]
U[1-4]
CN[1]

The letters seem to mnemonics or abbreviations. C for capacitor, R for
resistor, DZ for zener diode (?), JP for jumper, S for switch, CN for
connector, U for unusual component.

These are all standard layout and parts-list abbreviations. As you note,
R = resistor, C = capacitor, etc. however, U = "Integrated circuit".
There are also others, such as "Q" for transistors, or L for coil, but I
don't expect you'll see that on a mouse board. Possible, but I don't
expect it.

There are components on the layout side but they are very flat. One of the
DZ's looks like it might really be a diode. The resistors, capacitors and
jumpers on the layout side all look like black boxes of some kind, but
the jumpers have a O written on them, the resistors have numbers such
as 513 and 330 written on them (actually, they are block lettering and
upside down might spell some acronym like DEE or EIS) and the capacitors
have nothing written on them. Could these weird components be surface
mount technology? They are visibly and abundantly soldered.

Yes, these are surface mount components. The "513" and "330" notations
you mention are two values of resistor. The "0" you noticed for jumper
is because a 0 ohm resistor IS a jumper. (or, thinking the other way, a
jumper is the same thing as a 0-ohm resistor) Black bodies with solder
on ends are almost always resistors. Tan/reddish bodies with solder on
the ends are almost always capacitors, though they might be diodes.
Black bodies with more than two connections (two on one side, one on the
other, for instance) are transistors (and will be labeled "Q#" on the
board) while integrated circuits will have anywhere from 4 to 50 or more
individual connections to the board.
The switches for the left and right buttons look a little like car batteries
and have written on the side

1A 125V AC
C NO NC

and each of C, NO, NC is above one of 3 leads going into the board.

When referring to pin function on a switch, C = common. NO = normally
open. NC = normally closed or no connection.
It appears
that the C and the NC are connected by the layout.

Not uncommon. Basically, pushing the button hits the switch, which
connects the "NO" pin to the "C" pin. The "NC" pin is irrelevant
The third pushbutton
switch looks more like a cubical building and is unmarked. The track wheel
has one end of its axis resting on that third switch but the other end of
the axis goes into something with a hole in the middle and which is apparently
turned by the rotation of the track wheel. Maybe it is a tiny generator? It
has three leads going into the board.

That will be the rotary encoder for the wheel. Probably another opto
device, but may be something else. It's NOT a generator, though...
 
A

Allan Adler

Jan 1, 1970
0
Thanks to you and Robb for explanations about the components.

There is a schematic of a Symbolics/Logitech mouse at
http://lispmeister.com/blog/hardware/Symbolics-Logitech-circuit.html
I couldn't find a comparable one for the Hewlett-Packard mouse.

I stared at the layout on the PCB and made a crude drawing of all the
places where there is solder. Many of them don't seem to have any components
connected to them. For example, switch 3 is clicked by pressing the wheel
and sits over 6 soldered spots on the layout side, but there are only two
visible leads sticking through the solder. The other 4 are completely flat
and I think that someone just decided to fill holes for some reason. Some
of the flat solder places are under the middle of the mysterious
EICI127400 chip.

Then I tried to draw boxes around varous groups of solder dots as I
figured out how to associate them with components. I succeeded with
all but maybe two.

Next, I'll try to see if I can tell which dots are connected directly to
which, without using any test equipment. That will at least help to establish
relevance of one component to another. For example, after that it might be
plausible to guess that the mysterious EICI127400 chip with 16 pins serves
to debounce the various switches and provide other logic.

In the case of the opto-isolator pairs, the transparent one seems to have
two of its 3 leads connected, while the black one seems to have all three
leads disconnected.

Once I can guess how everything is connected, wrong as those guesses might
be, it will be possible to stare at the resulting circuit diagram and pretend
that I'm designing a mouse with certain components and trying to figure
out what might be wrong with my design.
 
M

Michael

Jan 1, 1970
0
Michael said:
By the way, the motor that moves the heads in that drive (made by Panasonic for
IBM) is a big, beautiful, Sankyo stepper. 1.8-degrees/step. I've never done a
hobby project with a stepper but this nice motor has inspired me to do something
with it.


FWIW, I mis-spoke re: the maker of that floppy drive. That nice stepper motor
is from a floppy drive made for IBM by HITACHI, not Panasonic. The
IBM/Panasonic drive that I have has a tiny, cheaply-made stepper motor for
moving the heads.
 
A

Allan Adler

Jan 1, 1970
0
Allan Adler said:
Thanks to you and Robb for explanations about the components.

There is a schematic of a Symbolics/Logitech mouse at
http://lispmeister.com/blog/hardware/Symbolics-Logitech-circuit.html
I couldn't find a comparable one for the Hewlett-Packard mouse. [snip]
For example, after that it might be
plausible to guess that the mysterious EICI127400 chip with 16 pins serves
to debounce the various switches and provide other logic.

I just found the following website.
http://www.fpga4fun.com/QuadratureDecoder.html
They aren't working with the same mouse as mine, although it seems pretty
similar. The IC in it definitely has different markings. However, the page
is pretty clear about asserting that the IC is responsible for quadrature
decoding. Whatever it is, the author understood its function well enough
to be able to replace it by a FPGA. That makes it seem likely that it is
a pretty standard chip, even if this particular EICI127400 seems obscure.
 
D

Don Bruder

Jan 1, 1970
0
Allan Adler said:
Allan Adler said:
Thanks to you and Robb for explanations about the components.

There is a schematic of a Symbolics/Logitech mouse at
http://lispmeister.com/blog/hardware/Symbolics-Logitech-circuit.html
I couldn't find a comparable one for the Hewlett-Packard mouse. [snip]
For example, after that it might be
plausible to guess that the mysterious EICI127400 chip with 16 pins serves
to debounce the various switches and provide other logic.

I just found the following website.
http://www.fpga4fun.com/QuadratureDecoder.html
They aren't working with the same mouse as mine, although it seems pretty
similar. The IC in it definitely has different markings. However, the page
is pretty clear about asserting that the IC is responsible for quadrature
decoding. Whatever it is, the author understood its function well enough
to be able to replace it by a FPGA. That makes it seem likely that it is
a pretty standard chip, even if this particular EICI127400 seems obscure.

"Quadrature decoding" is a 5 dollar phrase for the 25 cent concept I
touched on in another message: "If we get a pulse from A, then we get a
pulse from B, we know the wheel is turning clockwise, but if we get a
pulse from pulse from B, then we get a pulse from A, we know the wheel
is turning counterclockwise"
 
A

Allan Adler

Jan 1, 1970
0
Don Bruder said:
"Quadrature decoding" is a 5 dollar phrase for the 25 cent concept I
touched on in another message: "If we get a pulse from A, then we get a
pulse from B, we know the wheel is turning clockwise, but if we get a
pulse from pulse from B, then we get a pulse from A, we know the wheel
is turning counterclockwise"

Yes, of course. But the author also had some idea of what to expect from
the specific pins on the IC. That's what I meant by function, although I
should probably have used a different word. This particular chip has
16-pins in two rows. After I know more about the connections on the
board, I might be able to make better guesses about the pinout of the IC.

I'm getting a little better at staring at the photocopies of the board and
can now make out some connections. But it is still pretty hard to read. I can
see most of them pretty clearly when I just hold the PCB. The trouble is
that I like to sit in a coffee shop while I do this. I don't think there
is anything inappropriate about reading photocopies in a cafe but I don't
know how comfortable I or the other people in the cafe (customers, employees)
would feel about my studying an actual printed circuit board.

Just for the experience, I may photocopy it on a color copier and see if that
helps.

I realize there are websites that discuss these various SMD devices but I
like to read books. Can you recommend any that go into detail about the
various SMD and SMT devices I might find, and books that might explain
the rotary encoder devices that show up in mice, and books that discuss
chips such as the quadrature encoder IC? I have the Analog Devices
Analog-Digital Conversion Handbook, 3d ed, from 1986. On pp.181 and
443-445 it says a little about encoding, including a wheel with black and
white regions.

Since it is hard to make an accurate drawing of the PCB layout and since
I don't know enough yet about the devices and their connections to
make a circuit diagram, the next thing I'll do is to allocate a piece
of paper to each of the roughly 31 devices on the board. Each piece will
show a representation of the device and its connections and each will
be annotated in as much detail I can provide about the nature of the
device and which pins of which other devices each pin is connected to.
That will at least serve to organize the dribs and drabs of information
I can glean from the board and maybe will result in some kind of circuit
diagram. Also, maybe there is some kind of CAD program that can take that
compendium of information and produce a circuit diagram. I don't have one
and probably can't install one on my machine, but maybe someone here can
do it if I post the information.
 
A

Allan Adler

Jan 1, 1970
0
Allan Adler said:
Since it is hard to make an accurate drawing of the PCB layout and since
I don't know enough yet about the devices and their connections to
make a circuit diagram, the next thing I'll do is to allocate a piece
of paper to each of the roughly 31 devices on the board. Each piece will
show a representation of the device and its connections and each will
be annotated in as much detail I can provide about the nature of the
device and which pins of which other devices each pin is connected to.
That will at least serve to organize the dribs and drabs of information
I can glean from the board and maybe will result in some kind of circuit
diagram. Also, maybe there is some kind of CAD program that can take that
compendium of information and produce a circuit diagram. I don't have one
and probably can't install one on my machine, but maybe someone here can
do it if I post the information.

I spent a few hours last night carrying this out. I allocated one piece
of paper to each component and its representation, including labeling the
pins. Then I used a 7x magnifying monocle (maybe jewelers use them) to
trace the paths on the PCB to write down, for each pin, at least one other
pin it might be connected to on another device. It may not be perfect but
it should be pretty accurate. Now that I have collected all this data, I'll
try to make sense of it this evening.

Just out of curiosity, is it practical to make an x-y vision device that will
look at a PCB as I have done and automatically collect all the data about
connections? It took me hours (not counting the days to get oriented to the
kinds of devices) of staring at the board to collect this information for a
simple mouse. It would be too much work to analyze a more complicated device
in this way.
 
D

Don Bruder

Jan 1, 1970
0
Allan Adler said:
Just out of curiosity, is it practical to make an x-y vision device that will
look at a PCB as I have done and automatically collect all the data about
connections? It took me hours (not counting the days to get oriented to the
kinds of devices) of staring at the board to collect this information for a
simple mouse. It would be too much work to analyze a more complicated device
in this way.

That would likely be due to the fact that most of these devices are
never intended to be analyzed the way you're doing. The intent is "use
the device in the way we say it's supposed to be used, and when it
breaks, throw it away and buy a new one." The only detailed circuit
diagrams available are the ones used by the folks who designed the
beast, and you've got a snowball's chance in hell of prying those out of
the maker's hands...

In fact, depending on the exact interpretation of the law, attempting to
reverse engineer such a device the way you're doing MAY be a
patent/copyright violation that could put you on the hook for a whole
bunch of money.

Meanwhile, I don't recall hearing what the point of this whole mess was?
You've gone off on a tangent that makes little or no sense if all you
want is the devices. Why are you trying to map a mouse circuit board,
rather than simply grabbing the devices you need off it and pitching the
stripped carcass in the junk?

What, exactly, are youre efforts tying to accomplish? Why did you want
these things in the first place? Or has that original project fallen by
the wayside as you explore a seemingly fun puzzle?
 
A

Allan Adler

Jan 1, 1970
0
Allan Adler said:
I spent a few hours last night carrying this out. I allocated one piece
of paper to each component and its representation, including labeling the
pins. Then I used a 7x magnifying monocle (maybe jewelers use them) to
trace the paths on the PCB to write down, for each pin, at least one other
pin it might be connected to on another device. It may not be perfect but
it should be pretty accurate. Now that I have collected all this data, I'll
try to make sense of it this evening.

I have made more sense out of it, but not enough yet. There are roughly
95 pins and the complete accounting of what is connected to what will fill
a 95x95 incidence matrix. I'm prepared to do that, if necessary, and to
write simple programs to draw some inferences from the information. However,
it will be much better to find out more about the components. Using the
monocle, I've gotten a better look at the rotary encoder and at the
opto-isolators. So, let me describe them and see whether anything sounds
familiar.

The rotary encoder consists of a kind of wheel, i.e. a cylinder, with a
hexagonal hole in the middle. That hexagonal hole is matched by the
hexagonal head of one end of the the track wheel axle, which drives it.
One face of the cylinder seems to be perfectly flat and fits against a kind
of face plate of the same material. Both are held together by a metal
frame that is anchored into the PCB and seems to be grounded. The other
face of the cylinder has alternating radial hills and valleys; this is
the face that fits against the metal frame. The top of the metal frame
has a kind of dip in it, like the spout of a pitcher, pointing towards the
cylinder. The spout fits into one of the valleys of the cylinder. As one
rotates the track wheel, the cylinder turns and the spout is moved slightly
as it passes over the hills on the cylinder. I haven't counted the number of
hills and valleys yet, but it would be easy to do so.

As for the opto-isolators, each pair consists of a transparent piece,
which I think is the transmitter and which I will denote TR, and a black
opaque piece, which I think is the receiver, and which I'll denote BL.
Both pieces seem to have exactly the same shape, to wit:
(1) Both have three leads descending from the device. The leads are in
a row, but the middle one seems to be bent to point in the direction
that the light travels. So, they sit on the PCB as tripods.
(2) The device itself looks like a television set. It is roughly cubical.
The bottom has the leads coming out of it. One of the faces not in
the plane of the leads has a nearly inscribed disk which is slightly
recessed. In appearance it is like the TV screen. On the face opposite
that one is a kind of bulge, reminiscent of the bulge at the back of
many TV sets. On this device it looks like the top of an LED.

What seems strange is that, even though TR and BL face each other, they
are not placed so as to be mirror images of each other, geometrically
speaking. BL is placed so as to be a *translation* of TR along the line
joining them. Thus, if TR is to the left and BL is to the right, then the
"tv screens" of both TR and BL are on the right; the bulges of both TR and
BL are on the left; and the tripods formed by the leads both have the middle
lead bent to the right.

Apparently the bulge in TR transmits the light to the disk in BL facing
it, but the disk in TR and the bulge in BL have no function.

Now, suppose I am looking at the face of TR containing the bulge and I
see the three leads below it. Labeling the three leads 1,2,3 is not very
insightful into the function of TR. What would be a more insightful labeling?

Likewise, if I am looking at the disk of BL and I see the three leads
below it, what would be an insightful labeling of the leads?

If you know where there are data sheets for opto-isolator pairs that fit
the descriptions I have given, please let me know. Ditto for the rotary
encoder.
 
M

Michael

Jan 1, 1970
0
Allan Adler wrote:
(snip)
I don't think there
is anything inappropriate about reading photocopies in a cafe but I don't
know how comfortable I or the other people in the cafe (customers, employees)
would feel about my studying an actual printed circuit board.

Heh! I second that misgiving! Here in USA, one is hard pressed to think of
anything that couldn't get you face-down-and-cuffed because of some know-nothing
acting on his/her perceptions.

Michael
 
A

Allan Adler

Jan 1, 1970
0
Allan Adler said:
As for the opto-isolators, each pair consists of a transparent piece,
which I think is the transmitter and which I will denote TR, and a black
opaque piece, which I think is the receiver, and which I'll denote BL.
Both pieces seem to have exactly the same shape, to wit:
(1) Both have three leads descending from the device. The leads are in
a row, but the middle one seems to be bent to point in the direction
that the light travels. So, they sit on the PCB as tripods.
(2) The device itself looks like a television set. It is roughly cubical.
The bottom has the leads coming out of it. One of the faces not in
the plane of the leads has a nearly inscribed disk which is slightly
recessed. In appearance it is like the TV screen. On the face opposite
that one is a kind of bulge, reminiscent of the bulge at the back of
many TV sets. On this device it looks like the top of an LED. [snip]
Now, suppose I am looking at the face of TR containing the bulge and I
see the three leads below it. Labeling the three leads 1,2,3 is not very
insightful into the function of TR. What would be a more insightful labeling?
Likewise, if I am looking at the disk of BL and I see the three leads
below it, what would be an insightful labeling of the leads?

After several attempts, I tried searching for
3-pin infrared emitter led
and eventually found
http://www.robots-dreams.com/2005/01/controlling_you.html
It looks reasonably like my devices. The pins on it are, from left to right,
output, gnd, vcc. I'm not sure I can assume the pins have the same meaning
on both TR and BL.

The following occurred to me, but might be nonsense: Maybe the fact that TR
is transparent and BL is black is partly irrelevant. In other words, maybe
both are essentially the same device, with a bulge that emits and a recessed
disk that receives. Maybe TR is a lousy receiver and a good transmitter, while
for BL it is the other way around, for reasons having to do with one being
transparent and the other not. Taking a wild guess, the output pin is the
signal received by the recessed disk. On a device that receives badly, the
output should be very low and it might be safe to tie it to ground. That might
be consistent with the way some pins are tied together on the two opto-coupler
pairs. And it would also imply that the pins have the same meaning on both TR
and BL.

Anyway, that is just speculation. I'll keep looking for data sheets. I'm
also trying to find books on these optical devices.
 
Top