SC-3000H Help (repeating 2 beeps)

[skate323k137]

Hey all,

A long shot, however, I am hoping someone has some pointers for me in fixing my SC-3000H.

I received it with no accessories / untested and so assumed some issues would be present. I'm using a nice PSU, and built an AV cable and this cartridge:

SC-3000 Sega Basic Level IIIA - Share Project - PCBWay

Sega Basic Level IIIA cartridge for Sega SC-3000. It allows 2 rom layout.

www.pcbway.com

This thread let me know the 2 repeating beeps may be the on-board RAM. I replaced it, but I'm still getting the same. https://www.smspower.org/forums/14114-ReplacementEdgeConnectorForSC3000CartSlot

5V checks out at 5.01 at the 74HCT IC on the cartridge.

Is it possible the wrong ROM image could cause this? if I press "reset" it will at least display the BASIC Boot screen, and reports something silly like 500 free bytes. Trying to RUN anything though results in just garbage scrolling on the screen for a few moments. If I run 'NEW' and then RESET it only reports about 2K of free RAM. I tried a couple good SRAM IC's on the BASIC Level III Cartridge, so I don't think it's them.

Any pointers or ideas appreciated

-Alex

 

[skate323k137]

In an interesting development, I burned BASIC Level II to a ROM, and... it works! However, I'm now confused why III A will not work, of course. I have tried both HC and HCT logic on the cart.

I am using HM62256 SRAM; The cart is labelled for AS6C62265; Could this be my issue? With BASIC II operating, this is my next suspect.

 

[skate323k137]

I still have not solved why BASIC III won't use extended RAM, however, I got a fair number of games working on EPROM in the mean time. The majority of 32k ROMs worked as is and I doubled some 16k ones successfully too.

All the pictured titles work fine. I've ordered some ENIG PCBs to replace the HASL Prototypes.

 

[seat safety switch]

I wrote a small memory-testing program for the SG-1000/SC-3000 as part of my own SG-1000 clone project. I can't attach the ROM here, but here is the source: https://github.com/barbeque/sg1000/tree/master/software/ramtest

It can be built with zasm, let me know if you need some help – I've only tested it on my clone and not on my SC-3000 which is sort of packed away right now. You can ignore the page-switching test at the end, as that's for my clone. On an SC-3000 (2K of RAM) it should start mirroring the ram at $c000 + $0800 or $c800. On an SG-1000 (1K of RAM) it will detect a mirror at $c400.

I think the "A" BASICs do not have onboard DRAM on the cartridge, whereas the "B" BASICs do. There is also an SG-1000-specific SK-1100 BASIC which has SRAM on the cartridge, as the original SG-1000 doesn't break out the DRAM refresh, etc lines onto the cartridge edge like the SC does.

The way the cartridges implement RAM is a little unusual, as there's a current-limiting resistor coming off of the '139 decoder (or in the case of the SC-3000, the MITEC LSI gate array.) The cartridge pulls up this decode output line, and prevents the SC-3000 from selecting its own internal RAM, effectively overriding the memory map. Notice the R3/1kΩ resistor here and the cartridge pinout schematic later in this document: http://www43.tok2.com/home/cmpslv/Sc3000/EnrSC.htm

I also made a IIIB cartridge board that uses um61512 SRAM, although I haven't written any code in BASIC to test to make sure all that RAM is actually legitimately there yet.

 

[VERAULT]

Thats a pretty nice system and a pretty easy way to get games working on it. I have yet to take the plunge on Japanese systems; Maybe after I get more UK/EU systems under my belt.

 

[skate323k137]

I wrote a small memory-testing program for the SG-1000/SC-3000 as part of my own SG-1000 clone project. I can't attach the ROM here, but here is the source: https://github.com/barbeque/sg1000/tree/master/software/ramtest

It can be built with zasm, let me know if you need some help – I've only tested it on my clone and not on my SC-3000 which is sort of packed away right now. You can ignore the page-switching test at the end, as that's for my clone. On an SC-3000 (2K of RAM) it should start mirroring the ram at $c000 + $0800 or $c800. On an SG-1000 (1K of RAM) it will detect a mirror at $c400.

I think the "A" BASICs do not have onboard DRAM on the cartridge, whereas the "B" BASICs do. There is also an SG-1000-specific SK-1100 BASIC which has SRAM on the cartridge, as the original SG-1000 doesn't break out the DRAM refresh, etc lines onto the cartridge edge like the SC does.

The way the cartridges implement RAM is a little unusual, as there's a current-limiting resistor coming off of the '139 decoder (or in the case of the SC-3000, the MITEC LSI gate array.) The cartridge pulls up this decode output line, and prevents the SC-3000 from selecting its own internal RAM, effectively overriding the memory map. Notice the R3/1kΩ resistor here and the cartridge pinout schematic later in this document: http://www43.tok2.com/home/cmpslv/Sc3000/EnrSC.htm

I also made a IIIB cartridge board that uses um61512 SRAM, although I haven't written any code in BASIC to test to make sure all that RAM is actually legitimately there yet.

First off, thanks for all of this! Incredibly helpful and as I work through this system again / next time I will give your code a go.

I thought the Basic III A had 16k and the III B had 32k, which is why I thought perhaps this cart design was mis-screened for III A since it is using a 32K SRAM. I don't know if I've found or tried a III B image but I probably should.

@VERAULT this system popped up untested, but it had the mechanical keyboard, so I could not help myself. To this point the only "real work" was socketing and replacing the onboard RAM IC, so I consider myself lucky to only have had to do that, source a PSU, and whip up an AV cable from a MIDI cable.