gartenzwerg

Is the VecFlash still available to buy? Someone tried my last version of Bloxorz and it didn't work with it. I could try to figure out why it doesn't work, if I would have one or alternatively a circuit diagram would help as well. The ROM for the latest Bloxorz version is here: https://github.com/FrankBuss/bloxorz/raw/master/Vide/bin/Bloxorz.bin Works without problems in emulators like ParaJVE, but not on VecFlash. I though maybe that the usage of PB6 would be a problem, because on my cartridge there is a PIC microcontroller and I implemented a protocol for it to save and load highscore etc. So this is a version without accessing PB6: http://www.frank-buss.de/bloxorz/Bloxorz-no-pb6.bin.zip but still no luck.

BTW, I published the first version of the cartridge board as open source hardware, in case someone wants to build his/her own game as well, but doesn't want to back and wait for the Kickstarter: https://hackaday.io/project/28295-vectrex-cartridge Otherwise you can still back my project, it is already 61% funded. I think the collector box looks really nice:

Nice project, I would like to buy one, too. Currently I'm running a Kickstarter campaign for a new game, and I was thinking about to store some information, like high score for lowest number of moves or best time. I've read in the help page of Vide that Robot Arena uses a DS2431 EEPROM. 1 kbit would be sufficient to store the data. Could you share the 6809 source code for accessing the EEPROM? Then I could already implement and test it in Vide and maybe use it for my cartridge as well. And then the backers of the binary version only, but who already have your cartridge, can use all features of the game with it. But I might just detect the type of cartridge and then just use my existing PIC to store the data instead of an extra chip on my cartridge. Is it possible to detect your cartridge at run time from a Vectrex program? Then I would need just one ROM file. Wouldn't it be possible with your ARM microcontroller to use the internal flash of it as well for storing and reading data from a Vectrex program?

I've started a Kickstarter for it, to implement the other levels: https://www.kickstarter.com/projects/frankbuss/bloxorz-for-vectrex-pc-mac-ios-and-android

Sorry, I lost interest in the project, might continue someday.

The PCB arrived. First I soldered the 3.3 V voltage regulator, the three voltage level translators and the FPGA, then I wrote a simple VHDL test program which just emulates a 4k ROM, and converted an Atari 2600 ROM to a Lattice memory initialization file (the FPGA is big enough to include the ROM data) with a Python script. I programmed the FPGA with a FT2232H dev kit, which is supported out of the box by the Lattice Diamond programmer. Was a bit tricky to config the FT2232 EEPROM. First I had to change it from "Virtual COM Port" to "D2XX" with the FT_PROG tool from FTDI. Then I had to change the IO pin drive setting from 4 mA to 16 mA because with 4 mA the clock signal looked more like a sine wave on my oscilloscope instead of a square because of the long jumper wires, which caused lots of JTAG programming errors, and I enabled the Schmitt trigger for the inputs. But now it works . Game screenshot from my LCD display: This is the test setup: Bottom right is the FT2232H dev board, top left a homebrew S-video converter (I should really mount this somewhere on the mainboard someday). Now I'm waiting for the WiFi module, shipping needs up to a month when I buy something from China on eBay. But meanwhile I have a nice test setup to implement all the different RAM extensions. The FPGA has a few kB internal RAM, but I'll try to use the external 128 kB RAM on my board. Much more fun to have some more RAM, which is required anyway to implement a web browser. Development turnaround time is fast, too: Synthesizing the current design needs about 10 seconds and flashing to the FPGA 24 seconds. And it can be done in-circuit, just with a mouse click (but needs power cycling after programming).

Do you know if it is still FCC certified, if you upload your own program to it? this article says there could be problems, if the software can change the way the RF part works. Maybe better to use just an external small microcontroller and use the module as bought, if I sell it someday.

It is only a prototype, I don't plan to sell it at the moment. Some people already suggested they wanted USB, so I think I'll add an additional FT232 chip for the next version (could be optional). The WiFi module is nice, because it has a programmable microcontroller (80 MHz RISC CPU) and flash (but not 512 MB ). First I wanted to add a SD card, too, but I don't need it, because all the programs for the Atari are so small, you can store a lot of them on the 8 MB WiFi module on-board flash, or at least on the external 1 MB flash, which I added just in case there are problems with the on-board flash or programming the integrated microcontroller. Parts cost are less than EUR 40. The most expensive part is the big FPGA (about EUR 10). Could be replaced by a cheap CPLD (less than EUR 3), but it makes development much easier. For example it can store a stub for the menu program which is executed on reset, then it loads a second stage menu from the serial flash in the RAM, all without the microcontroller. With a CPLD there is not enough space to store a program or to implement the load from the serial flash to the RAM, so it would need some tricks, like a NOP slide and meanwhile the microcontroller fills the RAM, then switches to the menu program at the right moment. I could use a big parallel flash chip, but this makes routing more complicated (for my Kerberos board I used a 4 layer board). Another thing which could be optimized are the voltage level translators. There are three 74lvc245 chips (8 bit voltage level translator), which cost EUR 0.50 per chip. With a CPLD with 5V tolerant IO pins (like the one I've used for my Kerberos cartridge, a XC9572) I don't need it. But maybe better to use at least one of the voltage shifters for the data bus, because the output of the CPLD would be only 3.3V. Too bad Xilinx recently discontinued the XC9572. Of course, there are still lots of these chips to buy, but not good for a new project. Is there another cheap CPLD with 5V tolerant input pins, which is still manufactured?

I'm designing a flash cartridge for the Atari 2600 with WiFi, this is a first prototype for which the PCBs are currently produced (will get them in a few weeks) : http://www.frank-buss.de/atari2600/index.html The hardware is not difficult, because basically I did this already with my Kerberos cartridge for the C64 and the Crazy Cartridge prototypes. But the software will be more work, because I plan to integrate flashing over a web interface, and a web browser on the Atari 2600 would be fun, or the ability to download and start cartridge images over the internet.