Jump to content

Eckhard Stolberg

Members
  • Content Count

    953
  • Joined

  • Last visited

Everything posted by Eckhard Stolberg

  1. Bit Corp was from Taiwan. As for 2600 games made by UK companies: In the early 1990s Salu Ltd released some games in Europe. I think they were from the UK. I'm not sure if any of their games were actually programmed in Britain though. Ciao, Eckhard Stolberg
  2. To lock out unlicenced games, the NTSC 7800 tests the games for an encrypted signature key. If no valid key is found in the ROM, then the console wouldn't start the game. When the PAL 7800 was released some years later Atari had already given up on the lock out thing. Instead of the key test code they build an Asteroids game into the PAL console. That means that PAL 7800 games don't need to have a valid signature key and therefore won't work on a NTSC 7800. But even if the PAL games had signature keys, they probably still wouldn't play well on a NTSC console. On the 2600 it's the game that controls the frame rate. So a TV with a v-hold knob could be tuned to play 50 Hz PAL games from a 2600. On the 7800 it's the console that controls the frame rate. So a 50 Hz PAL game wouldn't have the same processing time for a frame on a 60 Hz NTSC console. Therefore the game code would probably run out of sync with the display and the game would crash or at least be unplayable. If you want to play PAL 7800 games you would need a PAL 7800 console and a PAL or multi-standard TV. Ciao, Eckhard Stolberg
  3. Both games need to be started with the RESET switch on the VCS. The emulators simulate the console switches with the function keys on the PC keyboard. In z26 for example F1 is RESET and F2 is SELECT. I think in Stella it's the other way around: F1 is SELECT and F2 is RESET. You said you tried every key on your keyboard, and neither z26 nor Stella reacted to the funktion keys? This is strange. What keyboard are you using, and do the arrow keys work for the games that you were able to start? Ciao, Eckhard Stolberg
  4. What kernel version are you using? According to the manual you'd need a 2.6 kernel or a 2.4.19 kernel (or newer) to use -T4 timing properly. If you have an older 2.4 kernel, you'd have to run z26 as root to make -T4 timing work. Ciao, Eckhard Stolberg
  5. Are there any emulators that are getting it right? If you are looking at the emulator screenshots on the Fractalus page here on AtariAge, you'll notice that only 8 double-scanlines at the top and bottom of the landscape display are there. The lines in the center are missing completely. This is because the extra RAM wasn't emulated when the screenshots were taken. If you enable the extra RAM support in MESS for the Fractalus ROM, you'll probably get the same display with every second line missing that you are getting on the CC2 in 78SC_R16 mode. The 8 double-scanlines at the top and bottom are displayed correctly on a real 7800 as well no matter if you have extra RAM at $4000-$4FFF or not. I think this part of the landscape display gets stored in the 7800's internal RAM. But we'd have to have a closer look at the disassembly to find out for sure. Ciao, Eckhard Stolberg
  6. Think of the extra RAM in blocks of 256 bytes each. In 78SC_R16 mode you get 16 blocks for a total of 4K RAM. On the Fractalus cart you only get 8 blocks for a total of 2K RAM. The mapping on the CC2 would be like this: $4000-$40FF = Block 1 $4100-$41FF = Block 2 $4200-$42FF = Block 3 $4300-$43FF = Block 4 $4400-$44FF = Block 5 $4500-$45FF = Block 6 $4600-$46FF = Block 7 $4700-$47FF = Block 8 $4800-$48FF = Block 9 $4900-$49FF = Block 10 $4A00-$4AFF = Block 11 $4B00-$4BFF = Block 12 $4C00-$4CFF = Block 13 $4D00-$4DFF = Block 14 $4E00-$4EFF = Block 15 $4F00-$4FFF = Block 16 I believe the mapping on the Fractalus cart is like this: $4000-$40FF = Block 1 $4100-$41FF = Block 1 $4200-$42FF = Block 2 $4300-$43FF = Block 2 $4400-$44FF = Block 3 $4500-$45FF = Block 3 $4600-$46FF = Block 4 $4700-$47FF = Block 4 $4800-$48FF = Block 5 $4900-$49FF = Block 5 $4A00-$4AFF = Block 6 $4B00-$4BFF = Block 6 $4C00-$4CFF = Block 7 $4D00-$4DFF = Block 7 $4E00-$4EFF = Block 8 $4F00-$4FFF = Block 8 So the game would write the graphics data for one scanline to $4F00-$4F27 and for the next to $4D00-$4D27 for example. Then it would set up the MARIA chip to read it's data from $4F00 down to $4000 in a 16 scanline DLL region. The MARIA decrements the high byte of the graphics data addresses between lines. So on the Fractalus cart the MARIA would read two consecutive lines from the same block of RAM and display two identical scanlines. With the CC2 in 78SC_R16 mode the MARIA would read two consecutive lines from different blocks of RAM. The game never writes anything to the second block since it's identical to the first block on the Fractalus cartridge anyway. Therefore in 78SC_R16 mode the MARIA will only find uninitialised garbage data in the RAM for every second line. This will lead to the effect that you described. Ciao, Eckhard Stolberg
  7. I think it's more a problem with accessing the graphics in the RAM due to the way how the RAM chip on the cart is mapped into the 7800 memory. If you look closely, you'll see that only the landscape display has every other line being empty. The metal frames around the windscreen show up fine in every scanline for example. The landscape must be created dynamically and therefore has to reside in RAM. Looking at the disassembly, it seems that the game is accessing RAM in the area at $4000-$4FFF. But according to Curt's description there are only 2K of RAM on the board. I suppose the guys at GCC mapped the RAM chip in such a way that each 256 bytes block gets repeated twice in the memory map. Due to the way how the MARIA chip reads graphics data between two scanlines in the same DLL entry, that mapping method would allow for an easy way to generate a 2-line display. Each scanline of landscape data would be repeated twice. The programmer would only need half the amount of RAM and only half the time to calculate the data for the next frame. The extra RAM support in the CC2 and in MESS has a full 4K at $4000-$4FFF. So every other 256 block of RAM doesn't get filled by the game program. Therefore the MARIA will display empty lines or garbage, depending on how that RAM was initialised. Ciao, Eckhard Stolberg
  8. Thanks for the picture. It clearly shows that all the data and address lines are going straight between all eight ports in a cartridge port row. Only the two topmost pins on the left side of each port have an individual connection to the other side of the board. These pins are ground and +5V, so they must be used for selecting the active cartridge. Ciao, Eckhard Stolberg
  9. Activision itself changed Decathlon to use Atari's 16K F6 type bankswitching for the PAL version that they licensed to HES. But that's probably not going to help you much. Ciao, Eckhard Stolberg
  10. The EPROM can't be damaged from being plugged backwards into the socked on the game board alone. You'd also have to use game, so that an inproper voltage could be applied to the wrong pin. I suppose Bruce only damaged his EPROM, because he tried to program some data onto it. For that you'd have to apply a much higher voltage to a special pin. If that accidentally goes into one of the address lines, that might cause some damage. A scenario like this seems highly unlikely in your case. But if one of the address lines was really dead, then you wouldn't be able to properly address all the data on the chip anymore. Half of it would be lost. If you suspect such a problem anyway, you could test the address pins with an ohmmeter like Bruce suggested. Do you have a multimeter? Ciao, Eckhard Stolberg
  11. Most of the games that don't work in the Imagic selector either have a timing-sensitive bankswitching scheme or are writing to an extra RAM chip. Since the ribbon cable is pretty long, the signals between the cartridge and the cartridge slot on the 2600 would be delayed a bit. This is probably why these games don't work properly. Maybe plugging them into the selector cartridge slot that is closest to where the ribbon cable connects to the board might improve the situation for some of them. Looking at the board, the only line that is seperated on all cartridge slots seems to be the +5V line. So I suppose the games are selected by only providing power to the corresponding cartridge slot. Ciao, Eckhard Stolberg
  12. The 7800 KLAX prototype has some code in it that should have no effect on a real console, but for some reason makes it crash sometimes at startup on certain 7800s. I think Mitch has a fixed version of the ROM on his site that has this code removed. Ciao, Eckhard Stolberg
  13. BTW, what is the problem with the 7800 games that has been holding up their release for so long? I know the KLAX prototype had some problems with starting up on certain 7800s. That turned out ot be a problem in the game's startup code. The first thing a 7800 game has to do is to lock the console into MARIA mode. For some reason KLAX writes a different value to the mode register right after it has locked MARIA mode. The second write shouldn't have any effect, because the mode is already locked at that point. But apparently on some 7800s the console gets switched into a mode where the game would crash. When you overwrite the second write to the mode register with NOP instructions, KLAX seems to play fine on all 7800s. If your 7800 protos have problems with starting on some 7800 consoles as well, it might be worth looking for similar instructions in their startup code too. Ciao, Eckhard Stolberg
  14. What controller are you using with your Stelladaptor? The Stelladaptor tries to autodetect paddles and standard joysticks. So, if you are using a joystick with more than one fire button or some other extra functions, that might cause problems. Ciao, Eckhard Stolberg
  15. The picture is the same as "Bobby is going home" by BitCorp. Since the game is called "Niky Wanders", that would fit too. The cartridge looks like it might be one of those shorter "S.S." cases. So this game is probably one of the generic "Cooper Black" pirates of "Bobby is going home". Ciao, Eckhard Stolberg
  16. IIRC the two games that didn't work with my dumper program were Espial and River Patrol. So you should be able to read out the Miner games, if you can fit them into your 7800. Ciao, Eckhard Stolberg
  17. Actually the UK uses PAL-I whereas most of continental Europe (including Norway) uses PAL-B/G. The main difference is the sound encoding. If your TV can handle the different types of PAL (most modern PAL TVs can), then you should have no problem using a VCS from ebay.co.uk. Otherwise you might have more luck with a VCS from ebay.de. Ciao, Eckhard Stolberg
  18. My dumper can partly handle 3F bankswitching, but I don't think I ever released the version of 7800CTRL where I was experimenting with it. The code would only work with four of the six Tigervision games, and of course there is the problem with fitting the Tigervision cartridges into the 7800 cartridge slot. Ciao, Eckhard Stolberg
  19. Thanks for the pictures. They are really cool. H.E.R.O. looks a bit pale though. Could you post a complete list of all the special SECAM Activision games that you have, please? And did you try them on a PAL VCS too? I'm wondering if Activision really made special SECAM-only games, or if they made dual-standard PAL/SECAM games, that adapt the colours depending on the state of the colour/bw switch, like Atari did. And did you find any double games that you'd like to trade yet? Ciao, Eckhard Stolberg
  20. You are using a Linux emulator, right? By default there is a threshold around the center position for generic joysticks. This will go away, if you calibrate the Stelladaptor with the paddles plugged in. Use a command like "jscal -c /dev/js0" and set minimum to 0, center to 128 and maximum to 254 when asked. Then the paddles should go smoothly from one end to the other in a VCS emulator. I think there also is a way to record the calibration results so that you don't have to do this everytime you turn on your computer, but I don't use Linux myself, so I forgot what it was. In Windows it's usually better not to calibrate the Stelladapter to get the best results, BTW. Ciao, Eckhard Stolberg
  21. Yes, z26 enables the Pitfall II bankswitching and extra chip depending on the size of the ROM image file. There should be no need to mess with the overdump though, because except for the unnessessary extra byte it should be exactly the same as the normal Pitfall II binary. Ciao, Eckhard Stolberg
  22. z26 should automatically use the first two joysticks in the Windows controller settings dialog. Gameport joysticks don't get automatically detected by Windows though, so you'll have to install your's manually. To do so you need to open the Windows Control Center and find the joystick settings. There you chose "Add..." and select a controller from the list that matches your joystick. Then the joystick should appear in the list and you should be able to calibrate and test it. If it works properly in the test dialog, it should work nicely in z26 too. Ciao, Eckhard Stolberg
  23. If you have a SECAM VCS (the UHF sticker on the cable indicates this) and your TV is set to SECAM mode when you search for a channel, and you can't find anything but an orange screen, then this is probably comming from your VCS. On my SECAM VCS I sometimes also only get a fuzzy orange screen when I accidentally touch the cartridge after the console has been turned on or when I don't flip the power switch all the way. Maybe trying a different cartridge or cleaning the carts and the cartridge slot or the power switch might fix the problem for you. Ciao, Eckhard Stolberg
  24. The person at Atarisupport probably didn't quite understand how the Stelladaptor works. It's a normal USB joystick, so it will work fine with DirectX. Otherwise it shouldn't work with the Atari 80-games-in-one CD at all. This CD was created by Digital Eclipse, so you might have more luck with contacting their lead programmer for this project, Jeff Vavasour. He has done a lot of commercial emulators in the past and occasionally reads RGVC, so he might be more likely to listen to user suggestions than the support people at Atari. Maybe he'll program a patch for better Stelladaptor support. On your problem with Circus Atari: This game is one of the most sensitive games on a real VCS as well, so it is supposed to play like this. But in the control settings for this game you can change the sensitivity of the joystick axes a bit. You might be able to find a setting that suits you more. In arcade Asteroids you are probably not able to assign certain functions to the joystick axes, because they need to be available independent of other functions and there might be problems if you assign them to opposite directions on the same axis. With a Stelladaptor this would not be a problem though, because it sends a special signal when Up and Down are pressed at the same time for example. this is nessessary for supporting the driving controller. It requires special handling in the program though, so you'd need a patch for the Atari disk to make these games work with the Stelladaptor and your homebrew five-button Atari controller. Ciao, Eckhard Stolberg
  25. This is correct. Some of the undocumented inteructions on the 650x type processors will crash the CPU. These are generally referred to as 'JAM' instructions. The 6507 can only address 8K or memory, so address $1822 could also be $F822, depending on where the code was ORGed to. This error message in z26 usually means that you are using the wrong bankswitching format for the game or that the reset-vector was not set correctly in the ROM. Then z26 might try to execute some data as code and encounter a 'JAM instruction' in there. Ciao, Eckhard Stolberg
×
×
  • Create New...