Note from the author : please discregard the updates, as my inventory becomes more precise and I add "notes for myself". Thank you :-)

Rumaging in my "archives" I found old riches, pulled from boards of the mid-90s era, that landed in the recycle bin. Then I reached more recent landscape of tapes and reels and trays... Which of these could be used to build the prototype then the real device ?

In the "old pulls, misc samples and eBay finds" department we have :

• 18×IDT732 PLCC52: 2K*8 dual port SRAM :-) these will be handy to develop the memory boards. I will use 64K (and possibly larger) SRAM chips for the final design but a smaller version is welcome for the tests :-) (5V BTW)
• Surprise : 2×IDT7132 DIP52 ! same as above but bulky :-D
• For good measure : two samples of IDT7052 (2K x 8 FourPort PGA). Because after all, the YASEP can access up to 4 memory locations per instruction (instruction, SI4, SND, RES).
• 15×MACH210-PLCC44 : how do I program these ???
• >30× 32Ki×8 "cache" SRAM (in sort of SOIC package) like IDT71256 from old Pentium motherboards (I used to hoard the DIP versions too... they must be somewhere as well) Not sure I'd use them because of the high power draw, useless high speed and 5V supply (I have enough lower power 64K+ chips in 16 bits TSSOP)
• Flash "BIOS" chips : that's how I intend to encode arbitrary logic, such as multiply tables. Two 1Mb chips can make one 8×8 bits multiply table (with signed and non signed result). I must check the power supplies though. I had a BIOS Flash programmer but it works on the parallel printer port under Windows (2K) and... I don't want to look back to it, the system crashed hard years ago. A decent solution under Linux is appreciated :-D
• P3×AM29LV160DB : did I say "multiply tables" ? That's in tiny TSSOP and 3V only though and i have no idea how to program them. At least, the 8×8 multiply of the YASEP has input MUXes that can do the level translation.
  In the "overkill" department, that could also drive 2 7-segments digits in parallel :-P
• Wait for it... 25×ATF1502AS and 10×ATF1504AS (tqfp44) with their ZIF and LPT JTAG dongle and even CDROMs. Low power 5V with ISP. Never tried them, i got the box from a colleague but I was more interested by large FPGA than 32/64 macrocells... No idea what they are capable of but the JTAG programmability is a very welcome thing :-D it looks like some development tools are still provided by Atmel, it's in the $3 range and still distributed... Any feedback ?
• I was worried that I would use really too many 74HC374 on the bus interface. Fear not for I have some tape of sn74lvth16374. That's damned fine pitch but it drives 16 bits in half the surface of a DIP20 chip. I'm sure it will be handy but I wonder how I will route the signals...
• I found some MC74HC164D. Not at all like the '165. It's more like a '595 that was rocked too close to a wall as a baby. But it's a shift register so it could be handy, for example for the scans, to replace the '138 (but the 138 has output high as default, which is desired)... Anyway there is always a use for a shift register and it's cool that this one does not have to always copy the register to the output, like the 595 has to do.

(tbc, stay tuned)

PS: Sorry for the rants, I was only looking for 74HC138 chips :-P I found the SOIC tape, I needed 4 PDIP for the keyboard and I found 6, victory :-D

I now proceed to wiring the keyboard's control logic.

• 40× PAL22V10 in DIP24
• 10× 28F010 in DIP

Time to get a programmer !