I am interested in a circuit board to control a two digit 7 segment LED.  Information will be coming from my software.

Tim,

What you are trying to do is likely to prove quite a challenge.  What you are saying you need is addressable 2-digit, 7 segment displays.  Not only that, you also need to be able to send data, or some sort of interrupt signal, back to the controller to indicate the parts have been pulled and to clear the display.  There are a couple of options I can think of:

First, you can attempt to find some form of I2C 2-digit (or 2 one-digit) 7-segment LED display drivers.  The I2C interface has a 7 digit address bus (plus one read/write bit) for a possible of up to 128 of the same device (all differently addressed, of course) on the same bus.  Unfortunately, many devices do not allow the use of the full address.  For example, the MAX6955 happens to be an I2C LED display driver, but the first three bits are fixed at 011, meaning you can only give each chip one of up to 16 unique addresses, meaning you can only have 16 of these devices on one bus.  You can display up to 16 7-segment digits using one of these chips, which would give you 16x16 = 256 DIGITS or 128 2-digit sets per bus.  The addition of a multiplexer, or multiplexers would allow you to expand that, theoretically, to several buses, provided you follow the provisions in the I2C specification for driving the bus.  You could use a single micro-controller as the interface between your computer (via usb) and the I2C interface.  You would have to write code for the Arduino to handle the bus multiplexing and I2C stuff (there should be pre-written I2C libraries available).  The MAX6955 chips are pretty pricey (~$25 for one on DigiKey, but they get down to ~ $11 when buying in bulk (8 separate buses would yield 1024 2-digit displays, needing 128 of those driver chips, costing an estimated $1200).  The SAA1064 is a cheaper alternative driver chip, but only allows 4 digits per chip, and only 4 chips per bus.  In quantity, they cost about $2 (from e14, may be cheaper from DigiKey or Mouser), but would require 4x as many, as wall as more multiplexers, significantly increasing part count, and, thus, complexity of both the layout and the software.  In the end, the cost would be about the same as the MAX6955 route, probably with more headache.

The second option is a more simple 7-segment driver chip, such as the CAT4008.  This does not use an I2C interface, it is simply clock and data in, with a serial out.  This is basically just a shift register.  This needs 4 wires to each driver for communication(as opposed to only 2 for I2C based devices) which are clock, data, latch and blank (opposite of enable).  These chips are cheaper than the MAX or CAT chips, but you will need one per segment.  If you have 1000 displays, you need 2000 chips.  They are about $.075 each if you buy a reel of 2000, which would run about $1500 just for those chips.  Not to mention that it would take considerable time to shift out the 16000 bits to all of those cascaded serial devices.  You could use multiple pins from a micro-controller to do, say, four parallel streams of 250.  Again, you would have to write the micro software and integrate comms with that into your custom program.

Either way, you are looking at a considerable task, both in terms of development and cost.  And we haven't even touched on the idea of a "reset" button yet.  It would probably be easiest to have the part picker simply enter into the computer that they pulled x number of y part, and let the computer set the display to off or zero.  Additionally, the button could be wired into the display controller/driver "reset" or "clear" line, and simply clear the display's memory, turning it blank or to zero, without interfacing with the computer.  If you really, really need a button at each parts bin to interface with the computer, the simplest method MAY be a massive matrix (like your keyboard) arranged in a row/column configuration with an interface to the computer.  It may even be possible to have a micro-controller send an ascii character(s) to the computer via USB, acting like an input device (think keyboard).  Otherwise, you would have to integrate it into the I2C bus, which would be a MAJOR pain, though probably possible.

If you are not a engineer of some sort, this is a massive undertaking, and I'd recommend seeking a professional firm for the design, and possibly the integration, if you have the capital.  What you want to do is extremely easy on a small scale, but scaling it up that high makes it exponentially more complicated.

Best of luck,

Joel

Joel,

I really appreciate all your information. It sounds like you really understand this type of issue. I would like to talk to you in more detail. Please contact me ASAP and we may be able to work out a compensation situation that you would be interested in.  Please call or email me at

timcloninger@cs.cugg

Thanks,

Tim Cloninger

MAccuchart

Sent from my I-Phone

Hi, this interests me.  Sadly I have misplaced my phone.  I am searching

for it now, but have a lot of ground to cover.  This will resolve itself by

morning.  I am on Central Standard Time, plus DST.  What's your time zone?

Don Hersey,

I am central time also, Chicago area.  Where are you located?

Thanks,

Tim Cloninger

timcloninger@cs.com

Tim,

To be honest, that info came from just a few minutes of searching, and really just scratches the surface of the scale and complexity of this project.  I do not intend to discourage you, in fact it is very interesting, but it will be quite a challenge at least due to the scale of the project.

Tim,

To be honest, that info came from just a few minutes of searching, and really just scratches the surface of the scale and complexity of this project.  I do not intend to discourage you, in fact it is very interesting, but it will be quite a challenge at least due to the scale of the project.