I've recently started a business and have to dedicate a lot of time to it, thus some of my more complex projects are lagging behind. I'm more or less testing the waters to find something a little less complex (compared to my Xbox Live emulation project) that i can still work on realistically for the time being. The Sega channel has always interested me, from what i can tell i don't think it actually uses a data standard like DOCSIS. From what i gather it just downloads ROM data from an actually channel signal, rather than an actual data connection. I picked up one of these devices for a pretty good price and plan on dissecting it to figure out how the ROM data is decoded. I just figured i would ask around here to see if anyone has any prior knowledge before i go head first into this thing.
You'll probably want to grab some old CableTV headend equipment. You'll need to find out what format the cartridge is expecting, then inject that into the cable signal. An agile-modulator can be picked up off of eBay for peanuts. I have one for HAM Radio TV that I got for around $35 shipped. Hopefully it's as simple as an audio signal, sort of like how old cassette games worked.
I'm currently looking into a lot of hardware to grab. Im guess the Sega Channel is going to expect a digital cable signal and a specific digital channel just repeats the initial ROM image over and over again. Other channels would repeat game rom images.
Amateur TV channels work just like any other TV broadcast, albeit in our own allocated frequencies. It's nothing special, just a video signal pumped into a modulator that spits out the proper frequency to be picked up. You may be thinking of Slow-Scan TV, which is done by HAMs. Slow scan is most easily though of as a line printer. The data slowly builds up line by line, using a buffer. That may be a good direction to think though, every game gets a little chunk transmitted one after the other. Like SONIC-PT1 then FLICKY-PT1 then TETRIS-PT1, followed by SONIC-PT2 and FLICKY PT-2 and TETRIS-PT2, on and on until the game is fully downloaded. If you only wanted Flicky, you would ignore Sonic and Tetris information.
Yes that's what I was thinking. I've been thinking about getting a ham license to teach the merit badge.
I picked up a 550mhz RF modulator for $30. I don't think cable companies broadcasted at anything higher at the time. I really don't want to have to drop a couple hundred on an 860mhz model.
Someone at Sonic Retro used to own a Sega Channel "transcoder", I suppose what was broadcasting the signal that the modems were connected to. No idea if there's anything of interest in its hardware, but thought you might like to know that hardware is out there.
It looks the signal was delivered using QPR modulation. I can't seem to find any information on QPR modulators. All i can seem to find is whitepapers describing how QPR is supposed to work. I hope it wasn't using specialty hardware sega created.
QRP is low-powered HAM radio stuff, not related to this at all. QAM for cable evolved from the DOCSIS standard, which was pushed out in 1997, so probably not related either. I think the best way to figure out exactly how this worked is to find some old Sega Channel equipment or manuals. Maybe someone could look through patents related to it? Edit - Someone more intelligent that I, should look at https://www.google.com/patents/US6874161 https://www.google.com/patents/US5632681 http://www.google.com/patents/US4905280 https://www.google.com/patents/US5935004
QAM would simplify everything, but there are some online sources that quote QPR by name. According to some diagrams i found it was uploaded to a satellite in QPSK, downloaded in QPSK, de-encoded, and re-encoded into QPR before being broadcast. My Sega Channel came in a few days ago, i should have some time this weekend to pull it apart. The RF modulator was damaged in shipping so i will have to grab another. I've looked through those patents before, and they mention both QAM and QPSK by name, but no QPR.
After taking a Scientific Atlanta model apart i found the following chips. 1x Sharp - LH5168N-80L (64K SRAM Chip) 4x ??? - ZR4040DJ (Unknown purpose, perhaps power related?) 2x NEC - 424260-07 (Unknown purpose) 1x Scientific Atlanta - XCV38122FG15A (Unknown purpose, suspected CPU of some type, may be a motorola chip) 1x Motorola - ACT32 (Perhaps quadruple or gate?) 1x Scientific Atlanta - I83426 (Suspect main ROM chip, contains version information "V 0.25") I would like to figure out what the big Scientific Atlanta CPU is, it also looks like the device may be frequency tunable via a pot.
Thanks! I'm going to setup a mega CD dumping system to dump the cartridge to see if i can find any clues as to what architecture the main CPU chip may be. Hopefully it isn't fully proprietary and is based off of a standard architecture. It could possibly be a SPARC chip, considering Scientific Atlanta has licensed the architecture. Although this licensing could have happened after the sega channel was developed.
From what I remember, the Sega CD transfer cable will not dump the sega channel at all. Good luck with this! I have always been very interested in the Sega Channel.
My guess would be that there isn't a real ROM for the genesis to read on the Sega Channel. What most likely happens is that the Sega Channel loads ROM data into it's ram and then tells the genesis to execute it after the fact. There is probably a preprogrammed error message in the Sega Channel firmware that gets pushed into RAM if it detects no signal. My guess is that the genesis itself is completely unaware of the Sega Channel hardware. The most useful thing i can think of is to dump the firmware ROM chip, but it has been re-branded which makes finding a pinout very difficult.
There is necessarely a real ROM on the cartridge because this is from where the Genesis boots and run code. It contains the Sega Channel Menu program and data that appears when you power the console on. There are actually a few dumps floating around online and some people already studied the hardware, see this thread for example: http://forums.sonicretro.org/index.php?showtopic=25935&view=findpost&p=757692 Off course, Genesis hardware itself knows nothing about the hardware on cartridge port but that's true for every existing cartridges. Also, cartridge cannot "tell" the Genesis hardware to do "things", only way to "discuss" is through software that must first be running on Genesis CPU. Software (in the boot ROM) which is running on the Genesis hardware knows about the cartridge hardware and can communicate with it either through shared RAM for data (mapped in both 68k and ASIC address space) or hardware registers mapped in 68k address space for commands (I think I remember seeing the program reading/writing from/to $A13xxx range) Very likely, what happens is that the big ASIC is managing cable communication protocol (data acquisition and commands) and also does address decoding for the cartridge hardware acccess from the Genesis. It is driven by Sega Channel program running on the Genesis and when no signal is detected, it probably just reports an error status acknowledge to that program when asked to initiate communication, then the program displays the error message. Downloaded ROM is probably decompressed and stored in one of the on-board RAM which would be accessible by both ASIC and Genesis and run from here once the Sega Channel program jumps to it. Downloaded data (text, graphics, sounds, ...) are probably stored similarely, maybe in other shared RAM location. If you want to understand how this works, your best bet is to dump the boot ROM (or get one of the existing dump) and disassemble it / run it through a debugger / analyse it to figure the protocol used to communicate with Sega Channel hardware, where the different on-board RAM are mapped in 68k address space and the format expected for data. The cable protocol used to communicate with the outside would be more complicated to figure I guess since you can not really know what the ASIC is doing internally on that side.
Thats pretty much what i plan on doing one way or another, but i am unfamiliar with any dumping method other than the Sega CD method. Either way i am going to attempt it, but if the Sega CD method doesn't work i will have to figure something else out. Information on this hardware is scarce and scattered across the net, but from what i had found it seemed like the menu system was broadcast as well. The only real way for me to figure this out is to dump it myself. It is possible that the Sega Channel has no boot rom, because it has everything it needs to operate on the cartridge. This is purely speculation, but it is possible that the Sega Channel ASIC / CPU has a RAM chip that is shared with the Console port. The Genesis would see this RAM chip just like a normal ROM and attempt to boot from it. All the Sega Channel would have to do it put a small piece of code that halts the Genesis at boot, then load a bunch of data into this shared RAM chip and resume execution. This data could be the menu program, a game, or an error program. The Sega Channel could very well be it's own computer in this sense. My reasoning is that there is only one chip that could be a ROM chip on the cartridge. None of those traces appear to connect to the cartridge port, but i would have to do a trace test with a multi meter to be sure. The ASIC /CPU could simply relay the ROM to the genesis, but that seems like a waste. Again, it's all speculation at this point. I would like to trace the RAM chip to see if it connects to the cartridge port in any way.
