katsumbhong wrote:
mbc1 + flash chip...
= 512Kbytes ROM capacity.
I made these for mGB carts but they will of course work with any game up to the capacity.
Offline
Michigan
Offline
Minneapolis
I know if this thread happened on Neo-Geo.com forums it would be called the "NASTY BOOTS ZOMGWTFBBQ" thread. Hehe.
I much prefer CM.O's approach. 
Offline
Alive and well in fucksville
The most reliable cart ever
There. fixed name.
Last edited by bitjacker (Jan 25, 2014 8:06 pm)
Offline
Taichung, Taiwan
bitjacker wrote:
The most unreliable cart ever
There. fixed name.
Worked 100% of the time all the time for me. You may have just had bad luck perhaps 
Offline
Taichung, Taiwan
kitsch wrote:
there were quality control issues
Problems with every component or just one in particular?
Offline
Alive and well in fucksville
$75 was a bit much to pay for a cart that only worked 10 times. the problem with mine was the memory. now I have a derp cart though.
Offline
Alive and well in fucksville
i still use mine for pounder, so its not a total loss. actually it was sort of a blessing to force myself to use other softwares than lsdj. it worked fine with carillion for awhile.
Last edited by bitjacker (Sep 24, 2013 3:16 pm)
Offline
What's wrong with it
Pm me so as to not derail thread more
#backonthetracks
Last edited by 12ianma (Sep 24, 2013 3:17 pm)
Offline
matt's mind
katsumbhong wrote:
kitsch wrote:there were quality control issues
Problems with every component or just one in particular?
the main issues were failing MM1134A ICs.
essentially... ~25% of the ones we got had some kind of issue. the 1134A bug was easy to spot because it filled LSDJ with 00 39 00 39 on every line (or, it was something like this). but, there were all sorts of other problems too. lifted IC legs was another big one. the 'quality control/assurance' step following assembly was skipped, basically (it appeared, or else whoever was doing this simply didn't check the right things). pretty lame.
Offline
Sweeeeeeden
00 39 is a different issue, that's when the Gameboy crashes. When the MM1134A fails, the SRAM chip is never instructed to output any data. This gives the value 1A or 2A repeated everywhere. This is the value of the opcode that is used to read from SRAM, which stays on the data bus because of bus capacitance. (The traces and metal parts in the connectors are acting as capacitors which store this value for a short period of time, but long enough for it to be picked up again.)
Offline
matt's mind
yeah, thats it!
the "1A" bug
i knew it was a repetition of something... sets the time to 26:26:26 or something too, from memory. among other things (chains phrases)
that was the main overall problem with them. thankfully easy to spot and fix.
Offline
matt's mind
(if anyone's lightbulb above their head is going off now)
pull the ICs from Nintendo carts. the new ICs which are available seem to have a decently high rate of not working or failing over time. since i was left to swap these ICs out on carts myself, I found pulling the Nintendo ones to be worth the extra effort rather than using the new part because of their poor quality. even if the new ICs worked when replaced, it wasn't that infrequent a customer would right back saying, 'hey man, when i turned on my cart today, it is filled with all this 1A crap".
eh, i can't recall how they are marked. not 1134A though... something else. i'm really having some brain farts this morning, meh
edit -- new ones are marked 6735. ones Nintendo was using were '134A' on the bottom line. just from my experience with that, if you can pull an original for the replacement, i'd suggest doing this to save a future headache. maybe its not a quality issue, it could be something like ESD or simply those carts used bad parts sourcing. or were designed incorrectly. don't know, but that singular part was my bane for a while. and putting the nintendo ones on instead solved nearly all those issues.