dsmall

Here are some ideas which I hope are useful: (1) The Motorola 68xxx processors have a little-known feature: If RAM for the stack isn't set up, they cannot multiply or divide. They use the stack for holding temporary results. It'll probably double-bus-error because it can't write the crash stackframe to RAM. Double bus errors can be a bitch to figure out. If you're using those instructions in the custom part of your board, look out. (2) Hang a scope on the cartridge selects (FAxxxx and FBxxxx) and see if it gets pinged early on, by TOS looking at the diagnostic cartridge port. That's very early in system startup, should give you more information about how far the machine gets. If it gets that far, then look at the next thing -- setting up chips in the order the system does it. See how many of the chips get initialized. You may also want to check the glue chip and make sure it's generating decodes for the big chips. (3) You could make a custom cartridge for the cartridge port. It just needs to have the right magic number and address to jump to in it. The system jumps there early on. That would give you a chance to see if the CPU is working. Say, do a loop inside the cartridge, and check things with an oscilloscope -- is it hammering on RESET, or hammering on the cartridge? You could also write a baby debugger to single-step through the Atari ROM code. (3) Timing after RESET? Generally, whenever a chip gets a RESET, it's a good idea to wait a few cycles after to give the chip time to wake up, wonder why it has a hangover, and such. When the ST is spun up, it takes awhile for the RAM controller to be initialized. I'd say give it more than 2 microseconds to reawaken. (This is why "reset-proof" Ramdisks aren't. When the RAM controller is down, it's not refreshing the RAMs, which leaves holes in the RAM data.) (Note: The floppy controller chip needs a hard RESET to recover after changing frequencies from low-density to high-density floppies. This causes random errors and drove a lot of people like me into "The Cliffs of Insanity" when switching between low and high densities. Finally, we ran a seperate wire to the floppy controller chipand manually kicked its RESET pin, then waited for it to wake back up. There's a good oscilloscope picture of the glitch that happens here.) Since you're trying to get the machine basically running, I'd use a hard coded delay loop, and wait a bit after each big chip gets its RESET. Doing shift instructions, especially long shifts, are a pretty good method; it's something like 4 cycles + 2*N where N is the number of shift places. With the better barrel shifter in the 020 this may not help as much. If I remember right, MOVE.L (A7),(A7) eats a lot of cycles. I had to do this to give the SCC (serial chip in Mac) lots of time to wake up after RESET. (5) Early on, when the machine is turned on, it does a RESET instruction. That makes the 68xxx send out a RESET to the big chips. If it crashes early on, it goes back and does another RESET. You could take a look at the RESET pin with a 'scope and see if it's happening once, or in a loop, triggering away. It tries to tell you what chip would not initialize but that's very sketchy. (6) Reset on the ST has some interesting quirks... Have a look at the very start of the BIOS code. In many places, if something doesn't work, it jumps back to the resethandler, which may issue a RESET. Let me know what results you're getting, and I'll try to help. Thanks, David Small

Hmmmmm. Interesting thread. I guess I really should give this "emulator" thing a try. Sounds cool. Thanks, David Small

It's been awhile since I've seen a TT, but I'll try to help. "I set up the expansion correctly for 16 MB and have 4 x 4 MB SIMMs (but 3 chips on each). But while the first TOS test indicates 16 MB of TT-RAM is detected and fine, later when using machine can I randomly get 2, 4 or even more bombs with TT-RAM attached. I've also run a test-program for TT-RAM (attached by someone on this forum) and it screams (in German! ) with thousands of errors..." This kinda sounds like something which is failing from heat. (And where did you find nybble RAMs these days? :-) If it were me, here's how I'd try to find the problem: Take your TT apart to where you can get the TT RAM card. Take the TT RAM card and put it in your refrigerator (not your freezer). Once it's well cooled down, put it back in your TT. Does it run pretty well? It probably will if this is a thermal (heat) intermittent. Now let it warm up. A memory tester whacks away at the RAM chips and heats them up really well. (I recently found out that the MemTest 86 in PC's can heat up RAMs to where they do fail.) If it starts failing in a few minutes, you know there's a thermal problem. So keep the TT RAM in the memory tester, take a can of compressed air, turn it upside down, and spray the chips one at a time on the TT RAM card with the cool-it-down stuff from the air can. That'll take them far below 0 degrees very fast. (It doesn't take much.) When the errors go away, you've found the chip that's heat-touchy. You can also use a hair dryer to heat up the chips so you can also test. What's usually a "thermal"? Well, connectors that get oxidation. (To clean off, use the eraser off a pencil exactly like the erasers you had in school. Or a Big Indian Chief eraser. ) This is around 20 years old so it's not surprising. You may need to pry up and put back down all the chips, which helps scrape through the oxidation. Electrolytic capacitors can die of old age, look for a capacitor whose "lid" up top has been forced open. If the heatup/cooldown cycle doesn't work, sometimes I've been known to look for suspicious solder connections and resolder them. That's all the ideas I have. I do wish you luck! -- thanks, Dave

This is just a suggestion (and may well just reflect my inexperience with atariage.com ! ) Awhile ago, on the PLATO network, we had something similar to a topic called "Where are they now?". It was for getting in touch with people that are "lost" or who have not checked in here for a long time. Say, like me (blush). Is there already such a topic? I'd sure like to find Dan Moore, Doug Wheeler, Barb Hahn, and others involved in the Mac Emulator project, for example. Be fun to talk to Charles John and John Eidsvoog again. And where's Jim Capparall? Many thanks, Dave Small p.s. This is just a suggestion. If you've got something like this, I just missed it. Feel free to delete.

One thing you could always do is drop by the main emulator site and ask the question there. I don't know if you're using MAME or which under Linux. MAME is pretty popular and works pretty well, I've heard. You could always approach trying to understand emulator software by trying to understand the hardware underneath. You *may* find this helpful. Maybe not, too -- some video stuff is, err, complex. The XBOX uses some extremely complex video hardware. If memory serves, it uses an Nvidea chip and a Conexant video encoder. What I would do, if I were you, is go over to, oh, xbox-scene.com, or the XBOX Linux Project at Sourceforge (sourceforge.com) and look up the datasheets on those parts. They're not that easy to find but they are findable there. After that, I'd have a look at the emulator source code and see precisely what those options *do* to the hardware. All in all, I'd say this is going to be a learning experience ... *grins* Best of luck on this. If I knew the answer directly, I'd tell you; the best I can do is point you to how to learn the answer. Dave Small yeah, that Mac emulator guy

Many thanks for the kind notes! I may be able to supply back Antic and/or START issues for scanning. I'll have to look in storage. But would be happy to help. The Integrator was a nice OS board for the Atari 800 series. What it did was to decode the "SIO" vector, in hardware, and if it found that, force a new address (to yet another Dave ROM) onto the bus. It made everything as seamless as possible... Corus, Axlon RAMDISK. I did it mostly for a friend in Austin. Only now I know that using the Atari code for simple hex->dec conversion could foul up some things. Ah, well, live and learn. The whole question of how long magnetic media "lives" is of interest to me, so I'll probably be dragging the ATR-8000, an SB-180 (non-Atari), and the AT&T 6300 out of storage and seeing how that works. Tell me: Is there any source for 810 drives? Question: I simply do not know the law on posting "pasted up" pages from ANTIC nor START, nor the START disks. Has Jim C. given an ok? If so am happy to help. (I still have many of the START disks). How is the OCR software these days? I will *try* to get a website running (mine is notably lame) with kid pics, etc. Please note my dsmall@well.com address is still up, but, at last look had 1.5 megs of spam in it ... just junk mail. I hate digging through that sludge. Finally ... while it may sound like trials and tribulations, it wasn't. Whenever I got to feeling down, I would go read warranty cards. (Yes!) And what people wrote made me feel better, perked up my day, you name it. *I had fun *. Did you know that the "Saga of Yap" finally, somehow, made it into a device that emits ultrasonic energy as a "dog trainer"? Too funny... I hope it isn't misused. Just to let you know, in the other room, are a couple of IC's flip-flopping in a timer circuit that has ... *heh* *heh* ... "interesting uses." "Hacking is a survival trait" -- movie "Hackers" which, I think, was Angelina Jolie's first. Hey ... if anyone has schematics on how to supply a deglitched, good +5 from a car environment, I would appreciate it. The Camaro lives! Someone sent them a long time ago. Many thanks, Dave Small davetracer@aol.com (p.s. Please don't overwhelm that signon or it'll overload the lame aol mail buffer, ok?) thanks...

Oh, I've been doing this and that. I've been writing a lot more on the fiction side and finding out how hard it is to deal with characters in books (as opposed to, let's say, character sets). I'll ramble across topics here and do an "update". I think Linux is incredibly neat and I use it as a tool all the time, most notoriously, to solve word puzzles that drop by from the Internet. Linux on an "old" machine makes the PC downright useful at 200 Mhz with older EDO RAM. That durn thing hums along. My fave is SuSE Linux but everyone has their own fave, and I willingly listen. I put in some less than memorable time at a few outfits... there are funny stories there. I keep up with some friends. Toad Computers is now an ISP but Dave and Jenn Troy are still my bro and sis; Dave just had him a birthday! Jimmy Hotz allowed me to record a song at his studio; I still go the Hacker's Conference. Quantum computers are a comin' ... learn about QuBits and how to crack PGP stuff very fast. I had the 800XL on .. let us see, yesterday ... o'scoping out the millisec and microsecs of video timing with son #3, Jamie. It's funny I had never done this before, but I found out each TV frame is 1/60th of a sec (0.016666 sec) (kay, I knew that), that horizontal refresh is about +/- 64 microsec (15000 Hz), and the individual pixel times are around 500 ns. I learned a lot from Don Lancaster's TV Video Cookbook so I learned what I was seeing. I am quite surprised by the _beauty_ I see on the scope just by varying the display. Heck, that 8-bit Atari is a *great* video demonstration tool; Jamie learned a lot. The Camaro still lives and still Warp Drives. Hyperweb is still a project in the making. I wish I was more comfortable with "C", but I always feel like I'm playing with something that can instantly shoot me in the foot. Oh, well, getting down and rock and rolling with machine code was always kinda my thing. Maybe I should go write boot Roms? The trick is having enough computer power to drive the user interface, which we finally have, and enough mass storage, which we finally have. The kids ... Eric (19) is up at college. Currently he's sitting behind me beating a game called "Counterstrike" on a 1.4 Ghz machine I built up. Jenny (18) is also off at college and won herself a full scholarship to the place. She looks a lot like Sandy, which I think is a good thing (Sandy looks a lot like Laura Dern). I am impressed with Jenny. Jamie (13) is still in high school, of course, and going through Argelbra or however it is you spell it. Those of you who saw them while they were little (or remember Eric breaking his leg and us having to do a shutdown) would not believe he is now 6'3". I'm sorry there wasn't better information available about what killed Gadgets. I really didn't know that people didn't know. (Yeah, there's a great sentence! :-) ) Back in the early 90's Gadgets by Small simply died because we lost a lawsuit we filed against a memory controller outfit, for a bug in their chip, which delayed shipping the 68030 SST. The thing had a bug that was nasty indeed to find ... you had to switch "banks" of RAM (4M banks) very fast, and then the controller chip crashed, killing the 68030 and RAM refresh, thus leaving no tracks. You can sorta see this is a way nasty, lay back and bite you sorta bug. We happened to trip the bug when the stack was around the 4 Meg border, I believe; it hit below, above, below, all *fast enough* to trigger the bug. And doing it with a debug monitor just was not fast enough to cause a crash. I had to go to the absolute best computer guy I know -- who was in my high school computer club -- and he ended up finding the damned thing. Whooo, that was a lot of work. We took it to trial and I learned the lesson of "Always Settle Out Of Court". I guess the jury didn't "get" what was wrong, or whatever. Anywho, they hit Gadgets with their legal fees > $250K, and it was file b-k for Gadgets time. To make it all weirder, Jim Allen was the Expert Witness for the chip guys (?), and the Judge disallowed George, the board designer, as an Expert Witness. Puhlease. (Whatever happened to Jimbo?) The stress of all this was, of course, a world of fun, and I think it played a big role in Sandy and I breaking up in 1996. Yup, we've been apart five years. We had a fine fifteen years together and three great kids. I hear she is a fine Web programmer; she always could code well and had the artistic side covered as well. I wish her well. It's a lot of fun to look at the emulator market these days. There's around five Mac emulators for the PC. It just amazes me because, who ever expected clock speeds to hit 1000 Mhz? Good Grief! At those speeds it just runs away from (say) a 33 Mhz 68030, and getting the 68030 SST running was just not easy. I have a lot of respect for those designers. There are also Apple ][, NES, Sega, and you name it emulators around there, and I really think we all had some role in proving out to the world the issues of, basically, the right to do whatever you wanted with your ROMs, including, well, sticking them in another machine! I can sure remember the non-action by Apple and people being mystified; trust me, a lot of work went into defining how to do this thing legally. We never hit a point where the market for the GCR fell off; we hit a point where the business was wiped out by the legal stuff. Believe you me, I keep some of the Spectre stuff framed on the wall, including some registration cards, which were so kind... I am not sure if I can get my articles online to the web. There's a good chance the magnetic media they are on has died from, well, just age. I have no idea if my ATR-8000 (remember?) still works. Sure a fine machine for its time, though. I am having trouble getting my other writing platform from then, an AT&T 6300 (PC clone with 8086) to work again. Scanning the articles in would be enough to stun The Terminator; I believe I have copies of nearly all of them, but whooo, they fill two shelves worth of notebooks (just the articles, not the entire magazines). If you remember Barb, last I heard she was down in Jamaica, which she loves, with a new daughter (!). Barb was one of the big reasons Gadgets worked and I'll always be grateful. I don't know what happened to Dan when Supra was bought out by Diamond. Last I saw of him, he was in Vancouver, WA. Anyone know? Last I heard from Doug Wheeler I think he was up at MECC in Minnesota, I think. I may have gotten a more recent note and spaced it. Alas, my dsmall@well.com signon is getting spammed to death ... it is no fun to ask for mail and get, literally, 1.7 megabytes each time. I wish I had any clever ideas on that, but I don't. Spectre stopped at 3.0 externally, at about the 16th Beta revision. I worked my tail off at 3.1, especially to address SCSI drives, and we got them working. I got that working in the assembly code, and Dan Moore got it working in the "C" code, which was the frontend menu magic he did, but we could never get the "handoff" of variables from the "C" code to the ASM code to work. Dan was up at Supra by this time and coordinating with him was tough. It was tricky stuff. I also had some real unfun playing with the MMU on 030 machines, trying to get it to settle down on 12 meg Mac machines. Just never got it out before Gadgets was cashiered. (And I always wondered why I couldn't map faster, fastRAM memory into the low memory, and make things run faster. It just never worked for me, and I really hacked at it.) If someone wanted to try and release 3.1, and laugh themselves sick at some of my 68K Gadgets code, that might be okay ... you can sure tell the portions where I got mildly desperate and tried everything. It is *extremely* commented. Dan's C is pretty heavy in my opinion. I'd have to ask Dan if we could release his code under GPL or something. If it'd help one person learn some clever ways of getting impossible stuff done, I'd post it under GPL, no problem. But as my mentor Wayne Smith told me, it didn't matter how elegant 3.1 was; we never shipped it. What matters is what ya ship. I never got 3.1 out the door, but I did get the TT running Spectre off SCSI hard disks. The Falcon version just never flew; we did some work on it, the very hard "What about THIS idea?" stuff, but never found a good way to access the disk drive to read McDisks (probably have to go with 1.4 MFM format) nor really use its video capabilities. I wish I'd kept up on what other people did with the Falcon; I saw some pretty cool stuff. System 7.0 had a bug early on in the boot process that took a horrific amount of work to find. I believe it ended up being a Nil pointer *read* that returned a pointer to a data structure, well, it returned the Atari ROM startup addresses. That then plopped corrupted data all over everything. It was a subtle bug way in the middle of nowhere that caused a chain reaction crash. (The zero reads were a major pain for us because, well, they never triggered a bus error, and thus they were not fun to find and fix). MultiFinder was unstable because it was doing that. The only reason we really found it was the ZAX In-Circuit-Emulator. (This is also the only reason I found an interesting thing in 68K chips ... multiply and divide fail unless the stack is pointing at valid RAM. It uses RAM for a scratchpad!). I think I had to set up a multiple event trigger ... some line-A trap after this and that ... and then trace into some machine with a huge disk (probably a whopping 20 megs back then), then read to the end and find the disaster. It was really un-fun. I patched it by hand (what fun!) and let the machine run, but it crashed further in. If you remember your Mac history, the first release of 7.0 was not exactly a thrill. And the tragedy, I think, is the system update fixed the zero-store and zero-read. I never tried it personally but was told the patch issued to the initial release of System 7 fixed it. I don't know myself for sure. I will try and setup a non-spammed signon here shortly that ya can email; right now, the ones I have get clobbered very, very quickly. Need some work on anti-SPAM tech. Wish you all a very Merry Christmas and/or a Happy Holiday, hope you had a good 2001 (I did), and hope for a better 2002! -- thanks, Dave Small