How exactly do PAL chips get programmed?

[Teo Zenios]

So if I have the original code and know what chip it needs to be programmed
on I might be in luck, otherwise there is little hope?
Are PAL's or GALs expensive to get? Do people here have access to
programmers?



----- Original Message ----- 
From: "Tony Duell" <ard at p850ug1.demon.co.uk>
To: <cctalk at classiccmp.org>
Sent: Thursday, November 17, 2005 5:13 PM
Subject: Re: How exactly do PAL chips get programmed?


> >
> > As some of you might recall I was looking for drivers for my Tokamac II
FX
> > PDS card on my mac IIfx. Well I found the drivers from the original
company
> > founder. I read in Macworld (and had it confirmed from the company
contact)
> > that you need a PAL chip on the IIfx reprogrammed (or replaced?) because
> > Apple screwed up BUS mastering on the IIfx and nobody caught on at the
time
> > because those types of cards were rare.
> >
> > Well I have found another person who has the same card AND the PAL chip
> > needed on his unit but is having other problems getting his setup to
work.
> >
> > What I want to know is are PAL chips somewhat like eprons in getting
> > programmed or is there some other process involved? Does anybody know of
a
>
> Well, physically, a true PAL (as opposed to a GAL or similar) is
> programmed like a fusible link PROM. There are little 'fuses' on the chip
> that you burn out when you program it.
>
> However, there is one important difference between a programmable logic
> device (PAL, GAL, CPLD, etc) and a memory (PROM. EPROM, Flash, etc). And
> that is that when you read a memory, you are effectively reading out the
> programmed pattern directly. When you use a logic device you are not
>
> Therefore, PALs (etc) have a 'programming mode'. You get to it by,
> normally, 'overvoltaging' one of the pins. In this mode, the other pins
> on the chip select a particular fuse, and enable you to program or read
> it. The exact method varies from device to device and manufacturer to
> manufacturer, and often was not available without an NDA (although the
> methods for the first PALs were published in the data sheets).
>
> This, in general is the only way to read out the contects -- the
> programming -- of the chip. In nromal mode, you can attempt to solve it
> as an unknown logic circuit, but rememebr there might be feedback terms
> or internal D-types (depending on the PAL) which will give sequential as
> opposed to combinatiorial behaviour.
>
> And there is some bad news. There is a 'security fuse' in most, if not
> all, PALs. If that has been programmed, there is no (official) way to
> read out the fusr map. You can only use the device in normal, not
> programming, mode. It's a sort of copy-protection.
>
> If you know the device type and manufactuer and have a programmer that
> supposts that _exactly_, it won't hurt to try to read it out. But don't
> be suprised if you get a file of all 0's or all 1's.
>
> -tony
>