On 9/29/2011 7:38 AM, Chuck Guzis wrote: Mulling the question of which firmware of several is installed and what boot disks are available for each triggers a thought perhaps appropriate for a different thread (unless there's an easy answer that I'm unaware of): There are several repositories for boot disks of all makes sizes but is anyone doing the same for the multitude of firmware PROMs that are slowly decaying in all these machines? Steve

On Sep 29, 2011, at 2:07 PM, "Chuck Guzis" <cclist at sydex.com> wrote: For PALs with purely combinatorial logic, this is of course doable. But what about sequential logic? That'd be pretty tough. -Dave -- Dave McGuire Port Charlotte, FL

On 29 Sep 2011 at 16:01, Alexandre Souza - Listas wrote: My intent in the next couple of weeks is to blog it on Erik's VC site. But as Dave says, combinatorial logic is pretty easy. Digging inside of a GAL can be very challenging indeed if the security fuses are blown (and on most commercial gear, they are). --Chuck

I built one years ago, initially for the CoCo, I later used in on an MS-DOS machine. It consists of 16 output lines (notionally address, but could be used for enables/clocks where nexcessoar), 8 input lines (used to read in the data), and +/-5V, +/-12V power lines. The CoCO version used TTL latches and buffers, the PC version used an 8255 PCB from Maplin. A manual toggle swithc tri-states all ouytput lines and turns off the PSU rails so you can safely insert/remove ICs. These signals all go to one row of pins of a 64 pin DIN 41612 socket. The other row of pins is wired ot the 32 pins of a ZIF socket. I then have little boards with a DIN41612 plug wired appropriately to conenct the inputs, outputs, and power lines ot the right pins of the socket. I suspect that for me, soldeirng up wone of those boards takes less time than modifying a configuration file, so it's actually easier for me to do it in hardware than to have thing that's configured in software. -tony

I used the DIN 41612 socket becyuase you can either solder a plug to a bit of squarepad board and wire-wrap or Verowire the pins together , or you can stick jumper wires into the holes in the socket. Thr former is what I do if I think I am going to use that pinout again (e.g. it's fro a common PROM), the latter is handy for a 1-off (e.g. reading a PAL where some of the 'output' pins are used as inputs only). Or a pin-programmable PIA chip (6821. 6522, etc) on the bus of a larger computer system (as an aside, I hate the 8255...) Not so. The problem is the power and ground pins. Not many (if any) microcontrolelrs can source/sink the current for some of the older EPROMs. And AFAIK none can output +12V or -5V levels which are needed if you want to be able to read 2708s and the 3-rail 2716s (that was a design requirement of my unit, I wanted to back up the firmware in a old Sanders dot matrix printer). Indeed. I think I am much like you here. I can wire up TTL and have it work first time. Programamble logic is less certain :-) -tony

All the PALs/HALs/GALs I'ev used (in DIL pacakges)_ have corner power pins. However I was looking at a more general solution because I have some very odd machines with very odd programamble devices. _Most_ EPROMs have corner power and ground pins. There are exceptions (and even nore exceptions when you look at mask ROMs). I have an old Logical Devices universal programmer somewhere. It's pretty useless because I've never managed to finf the PC interface card or software fro it. Anyway, it has a 40 pin ZIF socket and 5 pin driver boards, each driving 8 og the pins (with one exception, see below). From what I rememebr, there are 3 or 4 programable PSUs inside (set by DACs), the pin driver can connect a pin to any one of them or ground. Said pin drivers are stuffed with ULNxxxx buffer ICs, not suprisingly... Oh yes, they scratched hte numbers off the ICs, but that puzzled me for at least an hour. Anyway, the exception. Pin 20 -- the bottom left corner -- of the ZIF socket is directly connected to ground. Apparently getting a solid ground is harder than it looks, and pretty important. And I guess all the devices this programmer could support had ground in that corner. -tony