Post by thread:[PIC]: UNUSUAL D/A APPLICATION

There is a digital chip called a 'priority encoder' which will take 2**N digital inputs and return an N bit code identifying which is the 'highest priority' one that is active (or a special output if none is active). The 74LS348 (I assume an HC equivalent exists) takes 8 active low inputs and encodes them to 3 outputs. The chips support cascading so you can build a network of them that will directly decode your 50 input lines to a 6-bit binary number. The first level of the network contains 7 chips which provide 56 inputs. You only use 50 of them. The second level of the network contains 1 more chip, the inputs of which are connected to cascade outputs from the first level chips. The first level chips provide the low-order 3 bits of the resulting binary value. The second level chip providews the high-order 3 bits of the 6-bit value. You can then use that 6-bit binary number to drive a D/A chip. No PIC needed. Bob Ammerman RAm Systems (contract development of high performance, high function, low-level software) ----- Original Message ----- From: "Douglas Butler" <spam_OUTdbutlerTakeThisOuTIMETRIX.COM> To: <.....PICLISTKILLspam@spam@MITVMA.MIT.EDU> Sent: Thursday, August 30, 2001 9:07 AM Subject: Re: PIC UNUSUAL D/A APPLICATION > My first thought would not be to use a processor at all. I would run > the 50 inputs to 50 open collector buffers pulling down a resistor > network. Something like a R2R network but it might be simpler as you > don't have to distinguish when multiple lines are low together. Finally > use an op amp to buffer and filter the output. > > Sherpa Doug > > > {Original Message removed}

> I have 50 seperate logic lines that switch in sequence from > # 1 to # 50 then reverse from # 50 to # 1 > at any one time only one line is in a low state and all > others are high state. what I need is an analog output > voltage generated from the pic in response to which switched > line is enabled. 50 resistors and a current source? ******************************************************************** Olin Lathrop, embedded systems consultant in Littleton Massachusetts (978) 742-9014, olinKILLspamembedinc.com, http://www.embedinc.com -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

I previously wrote:

Since the lines are fired in a fixed sequence one at a time, it might be easier to just put open collector buffers on the odd numbered inputs and tie all outputs together. This will give a single input that tells whether the line number is even or odd and the processor can count up or down every time the even/odd line changes. Line 1 and 2 could be tied to additional inputs in order to allow the processor to lock onto the line number and determine whether to count up or down. The processor would not be able to determine the line number and direction until the line number reaches line 1 or 2, but this may be acceptable if this is a case where the processor can be programmed to not do anything until after a lock has occurred.

This can be further simplified by making your own buffer. Just put a diode from each odd line to a common point and tie that point to the base of a transistor through a current limiting resistor. Add a pull up resistor to the collector and run the output into the I/O port of your processor.

Bob Buege

-- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads