Steven N. Hirsch wrote: >Quadrajet1 wrote: >> I do recall the AE Z-80 Plus had 2 different models. One had a red LED >and >> one had a green one. The green LED model was the fastest of the two. > >Must be placebo effect. These were a complete knockoff of the Microsoft > Softcard and they all ran at 2x (IIRC) motherboard clock speed: 2Mhz. Yes--the absence of a real-time clock on the Apple II engendered a lot of subjective speed comparisons which were often misleading. >Given the way it interacts with the host processor, there's not a lot of >flexibility here. No doubt Michael Mahon will have the details on the >top of his head . Ha! ;-) I never owned an AE Z-80 Plus, so I wasn't sure if it was one with on-card memory or if it used the Apple memory. If it used the Apple memory, as the Softcard did, then you are absolutely correct--there is very little flexibility. The Apple bus allows memory accesses no more frequently than once per (Apple) cycle, or about 1MHz. The Softcard used a very clever syncopated clock derived from the 2MHz Q3 clock and the divided-down 7M clock, so it ran with "peak" speed of 3.58MHz and average speed of 2MHz. This timing allowed the 6502 to "keep alive" during the DMA accesses the Softcard made. The details on the Softcard were published in Electronic Design in the late 1980s, but I can no longer lay my hands on the article. Because the Z-80/8080 processors are much more CISCy and firmware-intensive, they take about four clocks to get done what a 6502 can do in one, so Softcard-derived designs always seemed somewhat sluggish compared to native Apple speed. (Unoptimized screen I/O that switched to the 6502 on each character compounded the sluggishness.) On-board memory allowed broad clocking flexibility, as you note below... >CP/M boards which communicated using port IO were generally faster, >since they needed no lock to the 6502 clock. I was involved in the >development of the CardZ180, which ran a Hitachi HD64180 at 9Mhz. One >of these in an accelerated IIgs was quite snappy! An easy rule-of-thumb to tell the speed of a Z-80 card is to look for a crystal oscillator on board. If there is one, that's the clock speed. If there isn't, it's 2MHz. -michael Check out parallel computing for 8-bit Apples on my Home page: http://members.aol.com/MJMahon/ Moll wrote: > You know, that makes me wonder if a nearly all-native CP/M 2.2 BIOS > could be written, that only uses the 6502 for drive access. Our CardZ180 board was heavily influenced by the PCPI Applicard (aka Micropro Starcard), which offloaded most of the IO handling to the host 6502. By means of some clever hardware and software tricks, we had an 80-column text driver capable of keeping up with the fastest modems of the day (9600baud was considered blazing) for BBSing. The CardZ180 host support was actually a ProDOS SYSTEM file. ProDOS stayed resident during operation and gave us simple access to the native filesystem. When you were done with the Z180, you just 'quit' back to the BASIC.SYSTEM prompt. Rod Pederson wrote a nice little program in Turbo Modula-2 which acted as a point-n-shoot file manager for moving data back and forth between CP/M <---> ProDOS. I think I may have added support for propagating timestamps, too. It was ahead of its day. Steve