[sci.electronics] Re^2: PAL's vs gates

cormack@crg5.UUCP (Mike Cormack) (02/07/90)

aez@Data-IO.COM (Adam Zilinskas) writes:

>Reasons for using PALs:
>1. Inventory. Many of the larger PIC (Programmable IC) chips can emulate
>  any TTL functions that you could use in a design. Its your choice to

Think real estate here too. 1 PAL uses much less space than several
glue logic devices. This can be a crucial in a board design. A reduced
parts count is always a nice idea and PALs can help out immensely.

>2. Board changes (only matters when making PCB, only slightly matters
>  if doing wirewrap). A new PIC and be programmed and patch a circuit
>  within hours/minutes compared to cutting traces on boards and patch

I know this one personally.  Reworking a board stinks and the
reliability decreases quickly once wires are added.  In some cases you
can't add wires.  A primary DFM rule is no wires at FCS and a board
turn if > 4 mod wires need to be added.  It's a competitive world
out there and reliability is very important.  Pals can be much easier
to use for design changes and sometimes are the only choice.

>Reasons for not using PALs:
>1. You have to program them. For productions lines of making thousa
>2. You have to program them. Most of the programming languages for
>3. You have to program them. A hobbyist can go to Radio Crack and get glue

O.K.  we get your point but I think that your not giving H.W.  guys
enough credit (you sound kinda biased ;-).  Gang programmers are great
looking but they're kinda spendy.

Another thing you don't point out is that when a person has to program
a PAL that means they have to touch it.  That's a handling issue which
can cause all kinds of problem for assembly folks.  The first is ESD
and right behind is orientation problems.  At least with glue logic
you can have a machine stuff them in your PCB and a person doesn't
have to touch it.  (That's providing you have an auto-inserter:-)

>I have tried to keep out any overt advertisements about what Data I/O provides

You've given some good advice and a little advert is o.k.  It's good
stuff so some advert is fine.  Some other things.  If your volume is
large enough, you could get a PAL vendor to burn the fuses as they
make it.  They use laser etching for that.  You can solder PALs into
your PCB once your confident that there are no bugs.  (That may take
some time).  Pals can be reliable (less than a 100 PPM).

mike
per yr you could probably have a pal vendor progra

elliott@optilink.UUCP (Paul Elliott x225) (02/09/90)

In article <19023@crg5.UUCP>, cormack@crg5.UUCP (Mike Cormack) writes:
> [...]
> Think real estate here too. 1 PAL uses much less space than several
> glue logic devices. This can be a crucial in a board design. A reduced
> parts count is always a nice idea and PALs can help out immensely.
> [...]

One thing that bugs me about this is that I can't seem to find any of the
20-pin PALs in surface-mount packages (CMOS would be especially usefull).
In the days of DIPs, squeezing 3 or 5 16-pin DIPs into one 20-pin PAL made
good sense, at least from a space standpoint.  Now, I find it hard to 
justify using a 20-pin DIP PAL when I can fit the same logic in about the
same space using SMT (SMT = Surface Mount Technology) HCMOS.  On a SMT board
we want to avoid through-hole DIPs where possible as additional manufacturing
steps are required, so the PAL has two strikes against it.

Yes, I am aware of the larger PALS in the 28 PLCC (and up) packages, and we
do use them where applicable, but for smaller glue-logic jobs the SMT logic
chips have become more attractive again.  Another reason that random-logic
chips remain a choice is the proliferation of simulation tools in the
workplace.  When PALs hit the world, often a PAL design could be simulated
much more easily than the equivalent discrete logic design.  More than once,
I designed a function with a PAL just to check out the logic, then implemented
it with the equivalent logic chips, because my power budget demanded it.

The die sizes in the erasable PALS (as viewed through the window) look pretty 
big; is this the limiting factor in package size?

If anyone out there knows of an EP310-equivalent functionality part in a SMT
package, let me know.  Even a low-power CMOS 16L8/16R8 would be nice.


-- 
Paul M. Elliott      Optilink Corporation     (707) 795-9444
           {pyramid,pixar,tekbspa}!optilink!elliott
                  "The dog ate my disclaimer."

phil@pepsi.amd.com (Phil Ngai) (02/10/90)

In article <3120@optilink.UUCP> elliott@optilink.UUCP (Paul Elliott x225) writes:
|One thing that bugs me about this is that I can't seem to find any of the
|20-pin PALs in surface-mount packages (CMOS would be especially usefull).

I can assure you that 20-pin PALs in surface-mount packages are
available. Try talking to the bigger PAL makers. CMOS is also available,
as are quarter power high speed (15 ns) PALs which are electrically
erasable. (since they are EE, you don't have to pay for a window, it
comes in a regular low cost plastic package)


--
Phil Ngai, phil@amd.com		{uunet,decwrl,ucbvax}!amdcad!phil
When guns are outlawed, only governments will have guns.