Yes, but does anybody have one of his systems running? What is it running? How well/reliably does it work? How compatible with "real" S-100 hardware is it? Dick ----- Original Message ----- From: "Sellam Ismail" <foo(a)siconic.com> To: <cctalk(a)classiccmp.org> Sent: Friday, June 14, 2002 8:59 AM Subject: Re: IMSAI News? Festival -- http://www.vintage.org *

Back when Todd Fischer was starting the development of his Imsai-II I had reservations about his plan to use a Z80 scion rather than the "real-McCoy, and using an SMPSU that provided regulated supply rather than the traditional unregulated supplies. Whe he asked me to become involved in the development I had a full plate and couldn't do it, but I certainly didn't like the "smell" of the notion of trying to compete in the rabidly competitive PC market, which clearly what it looked like he was wanting to do. Hoping to capitalize on the availability of low-cost PSU hardware, that having been a major problem with the early S-100 hardware, and that being a major weight/shipping-cost contributor, I think he was seeing the Z800, or whatever it was, as a PC-competitive vehicle, which it wasn't. In fact, I don't believe Zilog has anything capable of competing in that market at all. If one wants a product that is both CP/M-80 capable and also able to run the on-card firmware of ISA card peripherals, the V-40/V50 from NEC, and some of their scions, though the later scions no longer have the 8080 core, which is a problem, offer an alternative. The V40/V50 contain the serial I/O, timers, DMAC's, interrupt hardware, etc, that readily support executing 8086-code, thereby making the use of ISA boards straightforward, at least as part of the CP/M BIOS. They also offer speed advanatages that can't be overlooked, but generating the S-100 timing from their signals is not trivial and requires the sacrifice of some performance. If one wants to "have is cake (in the form of the 8080-compatible hardware execution unit) and eat it too (in the form of the 8086 instruction set compatibility)," the V50 certain would be my choice. I've built up a board that appears to generate the required timing from a V50 to run a basic S-100 interface compatible with late-70's (pre-IEEE 696 standard) S-100 hardware. Since I don't have an Imsai or Altair front-panel, I have no way of ensuring that such a device would work with this board, and since I've not had the time or motivation to get it done, I've not worked out the details of swtiching back and forth between the 8080 execution unit and the 8086 execution unit, which would be required to make all this stuff work properly. One of these days ... Dick ----- Original Message ----- From: "Bob Shannon" <bshannon(a)tiac.net> To: <cctalk(a)classiccmp.org> Sent: Saturday, June 15, 2002 8:04 AM Subject: Re: IMSAI News?

see below, plz. Dick ----- Original Message ----- From: "Tony Duell" <ard(a)p850ug1.demon.co.ukTo: <cctalk(a)classiccmp.orgSent: Saturday, June 15, 2002 12:30 PM Subject: Re: IMSAI News? Sadly, I don't think you'll find anyone on this list who believes it is. I'm not sure "superior" is an issue, however. If it were just, in some way, compatible with what it seems it pretends to be. The obvious solution, here, is to get a mono card. CP/M software didn't support graphics unless the graphics were associated with a specific display adapter anyway, so the use of a mono card, particularly one of the "short" mono cards with two serial ports and parallel port on it would probably be just the thing. Mono monitors are getting a mite scarce, however, but I still have one or two, as I suspect many others do. The mono card is well documented and does text at least as well as the best of terminals from back in the '70's. I doubt you're alone there, Tony. That's why I didn't make time to become involved with the Imsai-II development. I personally would have gone hog wild if he'd simply taken up the Z80C0020 and run it at full speed on a CPU card complete with peripherals (NOT Z80 peripherals!!!) that could easily be run without requiring that the CPU never run faster than the slowest of them. An SCN 2681 and a simple MSI parallel port would have worked fine, but an LSI like the 37C6-whatever-series would have worked fine with it as well. That would have handled the need for an FDC, 2 serial ports, a parallel port, and some other stuff one didn't have to use. In fact, that combo with some sort of on-board MMU hardware would quickly have made a whole and fully useable computer on a single S-100 board. Off-board I/O would simply have required one slow the CPU down whenever an off-board address was detected, which is simple enough. Using a PC keyboard and a PC mono card (normally has no ROM code required for it, aside from what's in the PC's system BIOS ROM, for which listings are readily available and easily translated) would have handled the console interface readily enough and making up a BIOS (remember, the BIOS normally is software under CP/M, not firmware) that supports these as well as an IDE HDD would have been VERY straightforward.

While it does make sense to put a cheap short mono card or the like in an S-100 box, together with a simple interface to a PC Keyboard, which is pretty cheap, the objective being to avoid the extreme expense of a dedicated terminal, or a PC used as one, I find the whole business of making a PC live on the S-100 silly, since a PC can simulate a CP/M environment at 50-100x the speed of the original with no trouble at all. If you have to hack the working original gear just to run it in the new box, I doubt it's worth it. More below ... Dick ----- Original Message ----- From: "Tim Shoppa" <classiccmp(a)trailing-edge.com> To: <cctalk(a)classiccmp.org> Sent: Saturday, June 15, 2002 9:03 PM Subject: Re: IMSAI News? the If you did that for every operation that required it, you'd be looking at a 1 Hz PC. The Z80 is probably a mite faster than the old 4.77 MHz 8088, but it would have to go some to match the already slow speed of the 6 MHz PC/AT in operations like running a VGA. display Unfortunately, the S-100 attempts to emulate the functions of a PC were sadly inadequate and actually cost a mite more than the same hardware on a PC. The stuff that worked "just like a PC-Clone" didn't work as fast, since the business of generating timing adequate to the task ran the S-100 at speeds considerably slower than the PC/AT. The one useful thing that the '696 standard could have done for us was to relieve us of the burden of having to generate the i8080 signals that were initially put on the bus by folks not intending to make a "real" computer out of the system. They still left in the old phase-1 clock and Sync, together with those stupid, and in some cases, redundant, status signals that simply allowed the slipshod designer to do things wrong in a wider range of ways. Once the 8080 was no longer a player, they should probably have gone with a signal subset that looked more like the signals on the Multibus-I, which, BTW, looked a lot like the ISA in several meaningful ways. They were just attempts to get between a fool and his sheckels ...

If you look at the spec's in detail, with an eye to what the "real" S-100 hardware required, you'd see that the Imsai-II comes up short. It wants to be something, but the decision as to what was not made before it was designed. Dick ----- Original Message ----- From: "Geoff Reed" <geoffr(a)zipcon.net> To: <cctalk(a)classiccmp.org> Sent: Saturday, June 15, 2002 8:34 PM Subject: Re: IMSAI News?