Taito F3 multi won't boot anymore .. Stuck watchdogging

[Tyrith]

I've edited this post several times and more recently for clarity given it's been nearly a year off-and-on of troubleshooting combined w/sporadic on-offs. It did this periodically when I first got it & went away for several months. Issue came back on occasion and got progressively worse until the board can no longer boot at all. Troubleshooting below.

 

[Tyrith]

I received some new Sandisk SDcards today & reimaged them. I took a video of whats happening & compressed it but server won't take it at 10mb -- I can put it on youtube if that's any help.

- Have confirmed voltage at 5.08v
- New & reimaged sdcard
- Confirmed the Taito F3 works w/the old Puzzle Bobble 2 game

Given the multi used to work after a reset & the button doesn't appear to work anymore, any suggestions for getting this up and running again?

 

[Tyrith]

** Issue disappeared for a couple of months.

 

[Tyrith]

This issue returned recently & did similar troubleshooting to previous w/hit or miss success. Sometimes it works, sometimes it doesn't. Most recently, it doesn't work at all.

I realized after some troubleshooting if I left it running for long enough (1.5-2min) it would load in to the options menu & start up the game. I ran it for 30 minutes this afternoon however and nothing ever loaded, just the on/off cycling of weird random stuff on the screen. Still works fine if I swap in the Puzzle Bobble 2.

Anyone have any ideas for troubleshooting? I've run out of ideas for this and as much as I want a functioning F3, this thing seems to conk out every couple of weeks when being used.

 

[Tyrith]

I was showing somebody the Taito F3 motherboard w/o cartridge as seen in the above video and had a nice ray of sunlight shine in from the window .... shining on an obvious repair I had missed. I had looked over the board before for any issues previously but IC8 (Sanyo LC321664AM-80) appears to have been previously repaired. Looks like some pads may have been lifted & some possible trace damage, along with a bit of rust on the legs.

I'm guessing since it's a DRAM chip, Puzzle Bobble 2 isn't large enough to cause issues whereas the Multi uses more? I reflowed it and waiting to test it. I'm guessing if resoldering doesn't fix it I'll have to pull the whole chip, check the traces & replace it. Anyone know where to source these?

 

[ShootTheCore]

I didn't notice your thread about watchdogging F3 issues until now - if I had, I would have suggesting checking the work RAM. Watchdogging means that the CPU is crashing, which typically means either bad CPU ROMs or bad work RAM. Glad you were able to track down the fault.

As for equivalents to the Sanyo LC321664AM, it's difficult to find equivalent RAM with the same SOP packaging. You'll likely have to buy J-hook packaging parts and straighten out the J-hook legs before soldering. Look into Toshiba TC51164BJ and Vitelic V53C664K80L.

The Vitelic RAM is available on Ebay: https://www.ebay.com/itm/161917433350

The Sega 32X uses Toshiba TC51164BJ on IC7 and IC8 on the main board - you could try buying a junk unit off of Ebay and harvesting the RAM from it.

 

[Tyrith]

No worries on response time - as indicated by the thread length, it's been a learning process. I had no idea what whatdogging was when I originally posted

The original repair has a broken trace+wire to pin 22 (A5), missing pad on 27 (N.C.) & missing pad on 39 (I/O15). Given 38 is also I/O15 and there's no visible trace on 39, it's looking safe to assume it's the ram itself. Guess I'll be on the hunt for something I can use to replace it with and will update down the road.