The All Aussie Sega CD Model 1 Repair Adventure

I feel we are close to found the issue. Some question :

Pins 1, 2, 8, 9 of ICs 5, 6, 7 8 are all multiplexed ? I have these results :

Signal between pins 1, 2, 8, 9 and IC 5/6/7/8 and IC10's pins 20, 22, 26, 27
- Pin 1 (OE) : IC5 > ok < IC6 > ok < IC7 > ok < IC8 >> IC10 pin 22 --> ok
- Pin 2 (CAS) : IC5 > ok < IC6 > ok < IC7 > ok < IC8

- Pin 8 (WE) : IC5 > ok < IC6 > nok < IC7 > ok < IC8 I don't know if this one is ok or not. CLWE lines between IC5/6's pin 8 and ASIC's pin 170 are ok. But I don't know if these lines must be conect on IC7/8's pin 8 since the signal is CUWE and goes to pin 169 of the ASIC chip.

- Pin 9 (RAS) : IC5 > ok < IC6 > ok < IC7 > ok < IC8

 

https://console5.com/techwiki/images/f/fe/PC_BD_MEGA-CD_MAIN_BD_171-8105C-Schematic-1_of_4.png

https://console5.com/wiki/File:PC_BD_MEGA-CD_MAIN_BD_171-8105C-Schematic-2_of_4.png

Use those as reference for your board. On Export units (non Japanese) the 4 DRAM chips were replaced by a single 16bit chip.

 

I already use these shematics for dignostics.
WE lines are OK :

- IC5 & IC6 pin 8 CLWE goes to IC10 pin 27 and ASIC pin 170
- IC7 & IC8 pin 8 CUWE only goes to ASIC pin 169

Here, the tests on 68k and the tests on ASIC. In blue, all tested lines.

 

Does D15 failed between one of the DRAMs and the 68K CPU? That's enough to keep it from booting, since it's program will be put on that RAM by the Mega Drive 68K, unpacked from the MD side BIOS.

 

Yes, line between pin 4 of IC8 and pin 58 of 68k. I fix it, but without result.

Found another "broken" trace. Between SCD<->GEN edge connector. It's a trace between line B20 of the edge connector and pin 190 of the ASIC. But the motherboard is an early version (172-6105b) and I don't know if it's normal. The trace of B20 line goes to the test point "M8".

 

B20 is VA17 (Address line 17 from MD 68K.) Without that the MD 68k cannot access the CD RAM memory properly. It won't be able to reach the IC registers either. Mend that up, too.

 

Confirmed ! Trace between B20 and VA17 (ASIC's pin 190) is broken. This trace is near capacitors C10, c12 and c13 there are badldly leaked and corroded this trace. I must was scratch the motherboard to find where goes the trace. The layout of this mottherboard is totaly diferent from shematic that I used. Luckily, componement are the the sames.

 

I think it will revive once you mend the trace.

 

Great work guys. Hopefully another Sega CD raised from the dead

 

Yes, Mega-CD 1 resurrected ! A lot of work to repair it. Found 3 broken traces, obviously near these crappy bad quality capacitors !

On this model it's very difficult to directly fix the trace. First because trace is under coated epoxy you must scratch before fix it. Second, this model has specific part like interface board between Megadrive and Mega-CD. You can't test your's with others interface coming from latter version of Mega-CD 1, there is not compatible with this model.

Sensitives traces are those near capacitors. Especially c9, c11, c13 and c37. If you are in this case :

- black screen, but Mega-CD is powered
- Mega-CD Power board is OK
- ribbon cables are OK
- you have 5v and 9v from mainboard

You'r likely in the case, as I_oliviera said, where something prevents the booting process and Genesis can't initialize the Mega-CD side. I checked all connections. In fact, this is not requiered. Check before all, lines between 68k and his RAM

- IC5, 6 7, 8, 10 on the old board (171-6105b or 171-6105c)
- IC 5 and 16 on the new board (171-6326b)

And after, all connections between edge connector (MD/MCD connector) and ASIC/68k. All lines related to boot process.

On mine, there are 4 issues.



---> broken trace between IC8's pin 4 and 68k's pin 58. This trace has been vanished by bad caps C37 and C9.

FIX :
Fixed under the board.

---> broken trace between edge connector B20 and ASIC's pin 190. Guilty cap C13.

FIX :
It's almost impossible to fix directly on the trace. Too tricky. Fortunately you can use some pads under the board. You must scratch them before.
Look under the board, M8 test point and go to a pads near IC20. You can find this pad on componements side, near pin 31 of the 68k.

---> broken trace on line LRCKB between IC20's pin 4 and ASIC's pin 75. Capacitor C11 broke the trace and his pad.

FIX :
I must take the pad under the board near C15 and link to pin 4 of IC20 chip.

---> Finaly, the cap C11 "eats" his own pad. Impossible to add a new cap here.

FIX :
I need to place it under the board.

After all, this Mega-CD is able to boot and run games fine. If you must scratch the motherboard, don't forget to put some varnish after repairing the broken traces.

Now, goes for a defect EU Mega-CD 1...

 

Told you it was going to be problem with the sub CPU DRAM.

Glad to hear you have it working now.

EU MEGA-CD is different in the sense newer version board uses 1x 16-Bit DRAM instead of 4x 4Bit DRAM. You can use MEGA/SEGA-CD 2 schematic to trace newer board DRAM pins down.

 

Yes, this information is very usefull to diagnose the issue. Before that I don't know where to look. Thank !

I have a question, the darlington transistor on the power board is fried. It's the folowing transistor.

Can't find the same as datasheet. But, found a counterpart. Someone can telle me if this one can works ?

 

Try with the BD652.

 

Thank ! Have you tested this BD652 ? I don't find the reference circuit in the BD652's datasheet.

 

Not tested by me but I know that one guy in an old spanish forum repaired his MegaCD1 with this one

Here the info (in spanish)...

https://dl.dropboxusercontent.com/u/99052134/MEGACD/MEGA CD 1 Frita !SOLUCIONADO¡.pdf

 

Alternatively, you can purchase this https://console5.com/store/sega-cd-tr4-replacement-transistor-for-no-power-condition.html. It's an upgraded replacement.

 

some news after trying to repair an euro MCD1. It's always down, maybe the CDC chip or his RAM... I have the BIOS background but without rotating logo...

But I found an interesting thing. Some BIOSes ar not compatible with Megadrive/Genesis 2. I made my test with a Megedrive 2. I remembered this EU Mega-CD was always stuck on background with no music or logo. But for these tests I only have a black screen.

I suspected a data rot in the BIOS ROM and changed with a region free BIOS. Result the background displays again. To be sure, I tested with another BIOS, official Sega-CD 1 BIOS. Again, black screen. I take my first ROM that I suspect to be corrupt and put it in working Mega-CD and it working like a charm. Definitively it's not a corrupted BIOS issue.

But the test on the working Mega-CD has been made with a Megadrive 1. I take this one, plug with the defect Mega-CD and his orriginal BIOS. Result, same issue which I remember, the background without rotating logo.

Megadrive 2 or some of them are not compatible with Mega-CD 1 orriginal BIOS. Otherwise, it works with a region free BIOS or Mega-CD 2 BIOS.

 

There are three kinds of CDC:

SONY
Hitachi (on JVC drives which are based on Hitachi chipsets, there are JVC ones using SONY chips too)
NEC (on "FUNAI" MEGA/SEGA-CD 2 or Wondermega2/X-Eye)

BIOSes meant to work on SONY CDCs will work on everything. BIOSes meant for FUNAI units won't work elsewhere.
Wondermega2/X Eye BIOS actually work on a Funai MEGA/SEGA-CD 2 but the CD muting circuitry is inverted so it won't have CD-DA audio.

Stick to the BIOSes meant for your unit and you won't have any trouble. That's why I region free hacked every single one I seen floating around and put on a single zip which mmonkey currently has, hosted publicly.

Also, some original unpatched CD BIOSes will black screen with a console configured as DOMESTIC (Japan) and 50Hz video mode.

 

It's not frequency related. The BIOS that I tested it's the first japanese one with no frequency control. It's works with both 50 and 60hz. On BIOSes with frequency control there is a blue screen with a message and I also tested with these ones. These BIOSes give a blue screen when frequency difere of thier region, not a black screen. I only have a black screen on working Mega-CD 1 when I plug a Megadrive 2 on it.

I think that I have found the issue. IC12 (74hc86a) seem to be fried. Some letters on his surface has been whitetened. This component is directly linked on LC8951's SRAM.

 

It's not frequency related but the output of JP3/JP4 is sent to the gate array IC and can be read back by the 68000. That is used on region detection.