cjl@ecsvax.UUCP (05/17/84)
To summerize my findings re PAL programmers: I still have not found a working <$1000 programmer available NOW. Data I/O is the hands-down favorite. With Logic-Pak (includes PALASM 1.?) around $4-5K; needs dumb terminal for PALASM. Almost any distributor worth their salt will program PALs *IF* you buy/bought the parts from them, plus a fee... What has anyone heard about the new MegaPALs? 32R16s? Talk of ">1500 programmable gates"? Does the new PALASM 2.0 support the new Megachips? (Vax and IBM software is always more than micro software. It costs more to develop and support, even if it is essentially the same program.) -Charles J. Lord cjl@ecsvax
sct@lanl-a.UUCP (05/18/84)
We evaluated the PALASM module for the DATA I/O and were not impressed. There were a lot of problems with it. I still believe that the best bet is to have PALASM on a host machine. MMI is sampling the new 32R16 now. This is a 1500 gate equivalent PAL with 16 inputs and 16 outputs (128 product terms) with a max prop delay of 35 ns. It comes in a 40 pin dip. We will try to get a sample but are not sure how we will program it as DATA I/O does not have a module for it yet. The 64R32 should be out later this year (?) and have a 5000 gate equivalent. MMI will also have a CMOS version. PALASM ver 2 is supposed to be out now but our salesman has not seen it yet.
phil@amd70.UUCP (Phil Ngai) (05/24/84)
VAX software is more expensive for the following reasons, I think:
1) sales volume will be lower, there are more PCs than VAXs
2) several people can use it on a VAX, only one per PC
3) people who have VAXen are used to paying high prices
Re the 32R16, I am left with a great "so what?". It sort of looks
like two 16L8s but in the space of 3 or 4 16L8s. And I bet the
price will be $50. Less functionality (half the functions per square
inch of PC board) at 5 times the price. Oh, and it's half as fast too.
A real winner!
(by the way, I could be biased in my opinions about MMI)
--
Phil Ngai (408) 749-5286 {ucbvax,decwrl,ihnp4,allegra,intelca}!amd70!philsct@lanl-a.UUCP (05/25/84)
The MMI 32R16 "super pal" may be a bit slower than a 16L8 (Tpd of 40 ns max) but to compare it with a 16L8 I think is unfair. To quote the spec sheet of the 32R16: The PAL/32R16 is a synchronous state-machine that implements the familiar state equation, Y := F(X,Y) where Y, the state vector, "is replaced by" on the rising edge of the clock, a Boolean function, F, of Y and the input vector X. It has 16 SMAC's which stand for State Machine Atomic Cell which consists of a registered sum of products with feedback forming a one-bit state machine. The 32R16 features product term sharing between output pairs, programmable output polarity and register preload. Registers can be bypassed in banks of eight leaving combinatorial outputs. There are two separate clock pins for independent clocking of register banks. End quote. From this it seems to me to be a very flexible and potentially powerful PAL.