🚀 go-pugleaf

In article <1992Dec9.130311.18309@ms.uky.edu>, billq@ms.uky.edu (Billy Quinn) writes...
>I have an 8051 based design and I am not sure whether I should use an
>8051/EPROM/latch or just the single 8751 for mass production.  My design uses 
>no external memory and the code is less than 1K.  I'm looking for a decision 
>based on the cost of 1000 PCB's made (sort of a fictional situation).  I 
>will make the following statements assuming that I am correct, and let any
>one out there interested in commenting tell me why I am right or wrong.
> 
>1) Using an 8051 with an EPROM and latch will require more board space
>   and holes to drill.  This will also make PCB layout a little tougher
>   and a more expensive process.
> 
>2) The 87c51's cost way more than 8051/EPROM/latch but it will still be
>   cheaper to do it this way because the added cost of the PCB (using
>   8051/EPROM/latch) will be greater.

	You've overlooked a few terms in the equation.  
	1.  The cost of programming the EPROM vs the 87C51
	2.  The additional costs associated with the EPROM,
	    such as the address latch, sockets, assembly, procurement
	    costs, additional bypass caps, and possibly greater
	    power supply costs.  There may be other costs I've
	    overlooked.
	3.  Field upgradabilty.  Cheap with EPROM, expensive with
	    87C51.  Also, you can build the device with the possibilty
	    of using various EPROM sizes, and purchase based on
	    best cost more easily.  Same goes for the CPU if it's
	    decoupled from the PROM.
	4.  The quality of the device will *generally* be better if
	    it has fewer parts.
	5.  The 1000 part number is about where mask programmed parts
	    start to become viable.  These will be a lot cheaper than
	    EPROM CPUs.  Also, one-time-programmable (OTP) EPROM CPUs
	    are available, which can significantly reduce the cost.
	6.  An alternative to the ordinary EPROM is the Intel 87C64
	    and 87C257 EPROM parts, which have built-in address latch.

	Hope this gives you something to chew on...

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  Rod Nussbaumer, Programmer/Technologist       Bitnet: BOMR@TRIUMFER
  TRIUMF --- University of British Columbia,  Internet: bomr@erich.triumf.ca
  Vancouver, BC, Canada.                         Phone: (604)222-1047 ext 510
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~