I wanted to play
Quest Forge at my leisure in an emulator, but I'm not sure what mapper it is (or if it's a known mapper). Due to the simplicity of the game, I had presumed it to be NROM, but quickly discovered that not to be the case after attempting to dump it.
I saw some write pairs to $8000 / $8001 near the startup code, so I thought it might be MMC3, but it doesn't appear to be standard MMC3 either, unfortunately.
Does anyone happen know about this game's mapper? (It was made by Piko Interactive, if that's any help.) I've attached some pictures of the board, if they're of any use to anybody.
If I figure it out on my own, I'll report back.
I've asked one of the developers on Twitter to chime in.
https://twitter.com/koitsu2009/status/8 ... 7460035584
Searching the board for "pocketgames" lead me to
this post, and
this aliexpress listing. I think I'm looking at an MMC3-style multicart mapper.
The Quest Forge game might actually be plain NROM once the multicart loader is finished booting.
Here's what the boot looks like:
Code:
$A000: $00 ; MMC3 vertical mirroring ?
$6000: $C1 ; ?
$6001: $80 ; ?
$6002: $00 ; ?
$6003: $00 ; ?
$8000,$8001: $06,$00 ; MMC3 CPU $8000-$9FFF = $00
$8000,$8001: $00,$00 ; MMC3 CHR $0000-$1FFF mapped linearly
$8000,$8001: $01,$02
$8000,$8001: $02,$04
$8000,$8001: $03,$05
$8000,$8001: $04,$06
$8000,$8001: $05,$07
; Here it loads stuff linearly from $8000 and writes to CHR
$8000,$8001: $06,$0C ; MMC3 CPU $8000-$9FFF = $0C
$8000,$8001: $07,$0D ; MMC3 CPU $A000-$BFFF = $0D
$6000: $C0 ; ?
$6001: $82 ; ?
$6002: $00 ; ?
$6003: $90 ; ?
; Shortly after this my dump and/or emulation begins to fail.
So, probably the next thing I will try is to artificially go through the boot sequence and then try to dump it as NROM once complete.
rainwarrior wrote:
Code:
$6000: $C1 ; ?
$6001: $80 ; ?
$6002: $00 ; ?
$6003: $00 ; ?
[...]
$6000: $C0 ; ?
$6001: $82 ; ?
$6002: $00 ; ?
$6003: $90 ; ?
Looks sorta like the COOLBOY MMC3 variant?
lidnariq wrote:
Looks sorta like the COOLBOY MMC3 variant?
Ah! That was a helpful tip. It might exactly be
COOLBOY!
Knowing that last write to $6003 would "lock out" the outer bank registers was important, I was having trouble getting the CHR data dumped without that piece of information.
Anyhow, I think I have now successfully dumped it as NROM! Yay!
CopyNES plugin/src attached in case it's useful to anybody. (I dunno if this would help for anything but this specific game; it's not a COOLBOY dumper, but just a hard-coded way to dump this particular game as NROM the way it happend to be set up.)
Attachment:
File comment: Title screen emulating!
QuestForgeTitle.png [ 2.12 KiB | Viewed 11791 times ]
Out of idle curiosity, what's the Spansion Flash that's used there? I can't make out a part number.
(I'm kinda curious just how ludicrously much extra memory is on the board.)
Yeah, it's hard to find an angle where it's visible. It reflects strangely. Looks like it says:
Code:
S29AL016D90TF102
822RBP94 H
(c) 04 SPANSION
From what I see, at the time of producing the game, have pulled Aliexpress. He misses me a lot.
Is not the editor afraid that the ROM, once in the hands of Chinese, will be massively copied?
Hey everyone! Sorry I'm late.
The game is written for an NROM256 but I actually have no clue what Pico's solution is to get in on the cartridge.
So I guess I'm not much help after all. = /
I was just gonna say, I'm pretty sure it's NROM cause I've now seen the end.
I'm just now seeing this thread, but this is crazy... These are the exact boards used by aliexpress sellers for the MMC3 capable 100 in 1 carts. My guess is piko contacted the chinese suppliers and asked them to provide boards with the rom he provided. I've seen mentions in some posts about this being an available service. And this is just their go to board. So an NROM256 game gets put on a 2MByte 3v flash chip with no level translation...
I ordered a copy of QF a couple days ago after hearing about how great it was on the assembly line podcast. Fingers crossed things have been improved and this isn't the board I end up with. If it is, thinking I'll dump the cart and program it onto one of my own legit 5v NROM boards..
Are you saying that there are people out there selling homebrew games on those chinese "time bomb" PCBs? That's... terrible!
It's beyond disgusting when people risk tarnishing an entire market just to make a quick extra buck. If reports of homebrew carts randomly stopping functioning or damaging consoles start spreading, that could really hurt the community.
infiniteneslives wrote:
I'm just now seeing this thread, but this is crazy... These are the exact boards used by aliexpress sellers for the MMC3 capable 100 in 1 carts. My guess is piko contacted the chinese suppliers and asked them to provide boards with the rom he provided.
I know (from experience) that that's what he does with GBA carts, so I wouldn't be surprised if that's what he does with NES carts also.
tokumaru wrote:
Are you saying that there are people out there selling homebrew games on those chinese "time bomb" PCBs? That's... terrible!
Sadly yes, I've attached photos of the aliexpress multicart I bought awhile back. Has same "pocketgames" branding and looks nearly identical aside from the connector pin count.
tokumaru wrote:
It's beyond disgusting when people risk tarnishing an entire market just to make a quick extra buck. If reports of homebrew carts randomly stopping functioning or damaging consoles start spreading, that could really hurt the community.
Indeed, it's rather disappointing to see it pop up in the homebrew community. Although based on piko's track history I guess we shouldn't be too surprised.. In his defense though, I would give him the benefit of the doubt that he didn't understand or could tell what was going on from a hardware perspective. I doubt it was intentional aside from cost motivations.
It's even gone so far as retrobit buying enough from 'pocketgames' to put retrobit's branding on the boards used for the
Data East officially licensed multicart (board images attached). I purchased it and the thing doesn't even fit in a front loading NES because the connector bay is too shallow for the ZIF style connector. I guess they only tested in retrobit clones when making that mold...
It was one thing when you could tell what was junk based on the fact you were buying on aliexpress or ebay. But one would hope the retrobit carts purchased from Amazon would be legit, and having a homebrew only available in this fashion is pointing things into a disappointing direction where you have to try and verify the hardware prior to purchase..
I have ordered custom carts from Aliexpress in the past for personal use, but they were always cheap EPROM+discrete based genuine NROMs and CNROMs. At least those were safe.
gauauu wrote:
I know (from experience) that that's what he does with GBA carts, so I wouldn't be surprised if that's what he does with NES carts also.
GBA is 3v though, so using 3v chips would not hurt the console; the biggest risk with those would be the cart itself failing due to bad quality, no?
calima wrote:
gauauu wrote:
I know (from experience) that that's what he does with GBA carts, so I wouldn't be surprised if that's what he does with NES carts also.
GBA is 3v though, so using 3v chips would not hurt the console; the biggest risk with those would be the cart itself failing due to bad quality, no?
Yeah, it's not problematic on the GBA the way it is on the NES. But just that since he's comfortable doing it for one system, that's likely what he does for others.
Thank you for alerting me about this game being dangerous. I've bought it quite a while ago but still haven't gotten around to actually playing it (as a kid, I never imagined I'd ever buy more NES cartridges than I had time to play). So guess it looks like I will make sure to dump it first and never run the original board in my NESes.
Out of curiosity, I did open my version, and it looks just slightly different.
Attachment:
IMG_20180308_145850.jpg [ 2.14 MiB | Viewed 6592 times ]
Attachment:
IMG_20180308_145832.jpg [ 2.47 MiB | Viewed 6592 times ]
But I'm guessing it's having the same problem with missing level translation?
Ironically, Dorke & Ymp - the other cartridge I bought from Piko at the same time as Quest Forge - I also needed to get a dump for, but for a different reason: The game is very frustrating and only gives passwords when worlds are completed. So when the game suddenly crashed with a garbage screen near the end of world 2, I just didn't wanna replay it all.
When contacting Piko about it, Eli blamed my SNES console, saying the game was "bug free", and wouldn't provide me a dump. So after creating a
post about how to dump it myself on this forum, Eli sent me a mail about how a "post like that can get you and the forum in trouble".
But my post was successful in that I learnt I could get a dump just by paying an additional $5 to download the game on Steam. So a "happy" ending to it all. But I actually haven't played the game since. Getting threats like that - as a paying customer who's just trying to work around a pretty bad game bug - left me with a sour taste that I've rarely experienced in the homebrew community.
Anyway, bit of a rant here, but point is I personally wouldn't give Piko the benefit of doubt about knowingly using dangerous components to cut costs, seeing how Piko treats customers. And while I realize that some precious lost games can only be obtained from Piko, I would recommend people to stay away from dealing with Piko when there are alternatives.
Bananmos wrote:
But I'm guessing it's having the same problem with missing level translation?
Yes. All the board appears to do is try to convert the 5V power supply to 3V for powering the 3V parts. It doesn't account for the different voltage levels of all signals like address and data lines.
Bananmos wrote:
I personally wouldn't give Piko the benefit of doubt about knowingly using dangerous components to cut costs ...
Crazy thing about it is the parts on those boards is significantly more expensive than if they would have used 5v parts (2x 1Mbit roms trimmed to NROM256). But I assume the Chinese already have those multicart boards in large volumes so it's cheaper/easier for them to use the boards on the shelf instead of designing a dedicated 5volt NROM (and/or discrete logic mapper) cartridge.
Hi all - Derek from Gradual Games pointed me to this thread via NintendoAge and I wanted to post a picture of the board that I just got when my game arrived today. Looks like it's basically the same as the earlier board. I'll let the hardware experts have at it though.
Thanks,
Mike
Xelius wrote:
Looks like it's basically the same as the earlier board.
Yes it's basically the same non 5v tolerant design. I still haven't got any info on the shipping of the copy I just bought, but I'm starting to loose hope that changes were made with the recent batch..
I guess one way to use these boards would be to build an adapter akin to an fc-nes adapter or game genie, with a proper 5v -> 3.3v level shifter, but ugh
My recommendation would be to simply play them on a cheap clone as it will likely be running at 3v itself. The fact that clones typically run near 3v may be part of the reason the Chinese designs don't bother with 5v tolerance.
It is under my understanding that the voltage regulator (U7) on the PCB would shelter the PCB from any damage. I don't think the PCB would cause any damage to the console, since the console is the one with higher voltage.
I test all our games my self, beta test that is, so I have to play through them for hours on both NES NTSC and PAL consoles (I don't care about clones, most are fine with exception of Retron 5).
Im not a hardware engineer, the factory we used is not the one directly selling bootlegs on Aliexpress, they serve all china, taiwan, and hong kong. And are the only ones that really give quality components and gold pins on PCBs (thus why Retrobit uses them) The guy that designs has been an electronics engineer for over 40 years and has developed games for NES and Genesis. He has done clone consoles, NOACs, and now he is working on clone consoles with upscalers He is pretty smart and we trusted his judgement, on top of that we tested the PCB to work.
It is a shame that he does sell the PCBs to other trade companies in china that do bootlegs but oh well, we just buy the PCBs, and cannot control anything else. We might be moving to get the games done by other company that has done legit hardware products sold world wide; we'll see how it goes.
We'll look into this issue and stress test to see if the PCB causes any damages to the console or to itself by being played on the console.
Also, I am going to reach out to Kevtris to see if he can put the rumors to sleep about the whole 5v >3.3V chips killing retro consoles.
piko wrote:
I Don't think the PCB would cause any damage to the console, since the console is the one with higher voltage.
That's unfortunately not how it works. If you feed a 5v data signal into a 3.3v device, you have excess voltage. To protect the 3.3v device (from ESD spikes), you have clamping diodes, and if done right, a level shifter (for continous excess voltage. not just on the power line but on the data lines as well). However if the latter is not present, a data high means the clamping diodes will dump the excess to the 3.3v-gnd line just like it would dump an ESD. Unlike ESD, this happens frequently which means you have a more or less continous excess. This demands excessive current output on all logic outputs, console-side, everytime it outputs a high. This draw of current is well outside specs and it is probably the explanation to reports of dead consoles as a result of using pirate multicarts of the particular design that is being debated.
Cartridge-side, this will cause excess heat on any clamping diodes used. I'm not too worried (note: i'm a layman) about diodes not being designed for continous conduction simply because they're used this way in control voltage circuits of comparable voltages without deteriorating (i've never seen a diode fail this way). I think it's more important to be on the lookout if the excess heat is within specs of the diode/other affected components or not. But anyway...
edit: edited in hopes of better clarity.
To be clear, trying to start an argument here. I'm simply trying to educate and provide the easily obtainable facts that aren't really subject to interpretation. It's very clear the datasheet specs are being violated.
MX29LV160ABTC datasheetI've attached page 21 Absolute Maximum ratings spells it out pretty clearly. The design exceeds the 3.3v + 0.5v = 3.8v maximum rating by ~1 volt when connected to original 5v consoles. As the datasheet notes this can cause permanent damage to the device. The permanent damage results from the high current condition that's created on the pin's ESD diode this is the effective circuit:
5v placed on flash pin ---|>---- 3.3v flash chip internal supply.
This forces a 1.7volt differential across that ESD diode which will result in a high current condition as it greatly exceeds the typical 0.7v forward biased voltage drop. This is precisely why the absolute spec in the datasheet is listed as Vcc + 0.5v to keep the ESD diode from turning on. Note 1 in the datasheet lets you exceed the Vcc+0.5v spec but only for a very short period of time 20nsec which isn't long enough for the diode to conduct a significant amount of current assuming that was even enough time to turn on.
Now in the case of putting this flash chip directly on the NES's CPU bus, you won't actually ever get the flash chip's i/o pins to 5v. The 5v CPU, PPU, & memories on the main board will try all they can to drive their outputs to 5v, but once the 3.3v chip's ESD diode turns on ~4v the console's drivers won't be able to drive the voltage any higher. They will try as hard as they can which equates to sourcing as much current into that line as possible. That high current is a load all the chips in your console weren't designed to supply, it results in heat which will only make the problem worse. Sure, maybe exceeding these ratings will never result in damage to your console. But it's a significant risk and burden I would rather not put on my beloved consoles, nor expect others to do without their consent.
Quote:
It is under my understanding that the voltage regulator (U7) on the PCB would shelter the PCB from any damage.
The funny thing is if you instead removed that regulator and supplied the flash chip directly with 5v it would remove all of these burdens from the console's chip set. It would clearly be violating the flash chip's Vcc specs, but this is mostly only concern for the flash chip which I care much less about than the chip set in my console. The flash chip supplied with 5v will obviously consume more current than when supplied with 3.3v, but without measuring it, I would guess it's insignificant compared to loads we typically ask of our console's power supply.
infiniteNESlives wrote:
The funny thing is if you instead removed that regulator and supplied the flash chip directly with 5v it would remove all of these burdens from the console's chip set.
Fwiw, the sram on the photographed mindkids boards is actually rated to tolerate both 3.3v and 5v on all its pins (2.4-5.5v)
Part number: BS62LV2001
Current consumption:
Code:
Vcc = 3.0V C-grade : 20mA (Max.) operating current
I- grade : 25mA (Max.) operating current
0.1uA (Typ.) CMOS standby current
Vcc = 5.0V C-grade : 35mA (Max.) operating current
I- grade : 40mA (Max.) operating current
0.6uA (Typ.) CMOS standby current
edit: typo - sram not srom
Quote:
Fwiw, the srom on the photographed mindkids board is actually rated to tolerate both 3.3v and 5v on all its pins (2.4-5.5v)
Yes, assuming it's supply is powered with 5v. IDK what supply the SRAMs use on the design. If they supply them with 3.3v then it's the same story/problem as the flash.
Anyway, it sounds like good news that it sounds like you've had a personal exchange with the pcb designer, piko. Maybe these concerns can be taken to them in hopes for a future revision? Even better if that hypothetical revision would be what they passed on to other pcb buyers as those multicarts, however unlawful, are quite popular/widespread and thus potentially harmful to legacy consoles.
Ready for some Engrish?
I had a quick chat with the engineer, here is what he told me:
"The Flash and sram is 3.3v, We use LDO for flash's VDD, but other pin from cartridge into the flash, so, the fact voltage is 4.3v"
I think he is telling me that there is less voltage actually going into the flash and sram than 5v?
Anyways, he told me he will do a new revision for the PCB. Still, I left the console today on for about 6 hours with a game on and no problem.
I'll do that every day to see what dies first, The first NES console I've ever played (it has been in my family since 1989) or a game with this PCB.
A thermal camera could probably be of use in this situation?
Quote:
I left the console today on for about 6 hours with a game on and no problem.
I'll do that every day to see what dies first, The first NES console I've ever played (it has been in my family since 1989) or a game with this PCB.
I think it's cool and appreciable that you do this. I'm not sure though that doing a wear test on a single unit is quantifiable? Somebody more experienced might want to chime in on this but it is my understanding every part may have small quality/tolerance differences that may be a factor in a function over time. I'm a little worried you may wear a unit of apparent nostalgic value for naught?
A similar (not identical) situation: the filter section in the EDP
Wasp synthesizer and later clones is known to break down with wear. The theory has been contested but it is popularly believed that it is because the way it is designed (which gives it a certain musical character) is (ab)using digital inverters as if they where opAmps (digital inverterters pass little current when in a certain state (hi or lo) but pass a lot more than they're rated for when held in-between). Some users have needed to replace the part quite quickly, others have never faced any problem.
I'm not really certain what goes wrong when too high of a supply voltage is used. Maybe it's thermal after all—MOSFETs are tuned to have a specific on voltage and you'll end up dissipating more power if your supply voltage is higher than designed.
But as far as what goes wrong when you're running at different logic thresholds—there's a good writeup
here (stackexchange).
Piko wrote:
I'll do that every day to see what dies first, The first NES console I've ever played (it has been in my family since 1989) or a game with this PCB.
Man if there was ANY risk at all to my childhood NES I would not do this; I'd just buy some junky one off of ebay to test. Or better yet, a clone console so no original NES is harmed. I am not going to be inserting Quest Forge into my NES with what I know about the pcb; instead I am going to dump the ROM and play it by other means. The game will sit on my shelf as a collectible. I think I may even remove the PCB so there is no chance anybody will ever accidentally grab this cartridge and use it.
Also thanks to everyone in this thread for being awesome and paying attention to this; I learned about this problem BEFORE using this cartridge. THANK YOU!
Your experiment is a bit like saying you're going to start smoking a pack of cigarettes everyday to try and see if smoking causes cancer. We're supposed to wait around indefinitely to find out if you get cancer or not? Or perhaps an experiment where you aren't going to change your car's oil for 30k+ miles to prove changing your oil frequently isn't necessary. Many people cherish their old consoles. If they were to be compared to vintage cars, they would happily change the oil every 3months/3k miles. With no concerns about how a couple dollars could have been saved by waiting to 10k miles to change their oil.
It's unlikely that your one test case is going to result in anything being damaged to the point where you will notice anything. It would only show that complete failure isn't certain. We already knew it wasn't going to cause certain failure. The extra load these improperly designed carts are placing on the consoles are more likely to show failure if there's already pre-existing damage which has not yet failed completely, or the driver is already weak due to a manufacturing defect.
Nothing the designer says, nor experiment is going to change the fact that data sheet specifications are clearly being violated. [EDIT: we're not talking about some minor like a timing spec either, these are the "ABSOLUTE MAXIMUM RATINGS" which the datasheet explicitly states "may cause permanent damage" if violated] There are many ways to create a design which doesn't violate these specs, or places to buy well designed boards. There really is no good excuse to use these boards (especially for an NROM game). The people buying the games have already spoken up and explained what they would like. Now that you're aware of this issue with the QF boards the choice is up to you on what you'd like to do moving forward.
What about the opposite --- will licensed 5V-expecting Nintendo cartridges take any damage from, or cause damage to, a Famiclone using 3.3V connections?
NewRisingSun wrote:
What about the opposite --- will licensed 5V-expecting Nintendo cartridges take any damage from, or cause damage to, a Famiclone using 3.3V connections?
Generally no it shouldn't be an issue in terms of causing damage. The console generally supplies the cart with the same supply it's running off of. If the cartridge has problems running at the lowered voltage it may not play, but that will result in a CPU crash or graphics glitching generally. It's not going to cause any wear damage, it just may not run.
The worst thing that could happen though is loosing your save, and/or draining the battery for battery backed SRAM cartridges. If the console's supply voltage isn't high enough to beat out the ~3v of the coin cell in the cart the SRAM may be operating from the battery's power instead of the console. That would drain the battery pretty quickly. It can also result in corrupting the save even if the battery doesn't get drained. I've seen this first hand with a battery powered FCmobileII clone of mine playing Zelda. I played until the console batteries were dead a few times. And when powering up the cart afterwards all my saves had been lost. The battery still has enough charge to hold saves, but I learned my lesson on playing games which I care about the save files with such devices.
Thanks for the info Infine.
Even Watermelon made this blunder with Pier Solar.
infiniteneslives wrote:
My recommendation would be to simply play them on a cheap clone as it will likely be running at 3v itself. The fact that clones typically run near 3v may be part of the reason the Chinese designs don't bother with 5v tolerance.
Would it be safe (safer?) to just dump the ROM with Kazzo and then play in an emulator (or use another flash cart)?
Yes, remove the offending hardware from the equation and you obviously don’t have it’s concerns. Extracting the NROM game from the Chinese multicart design and playing on another cart isn’t going to be straight forward.
infiniteneslives wrote:
Yes, remove the offending hardware from the equation and you obviously don’t have it’s concerns. Extracting the NROM game from the Chinese multicart design and playing on another cart isn’t going to be straight forward.
Thanks! I wasn't clear on your answer but my question was ambiguously stated. I know the ROM would be safe once extracted. What I meant to ask was: Is it safe for my Kazzo (and me as the operator) to connect the cartridge to the Kazzo to extract the ROM? Is the Kazzo tolerant to the board's design or could it also be damaged? Moot point now though since I've already done it.
I was able to convert Brad's CopyNES script to a Anago WX script and extract both the PRG and CHR. Seems like Anago doesn't like it when you tell it to read from PRG in the CHR script so I had to dump both chips as PRG with two separate scripts and then assemble the ROM file manually with a hex editor.
Seems to work just fine. I was just getting a black background (with sprites) at first but it was because I had the wrong mirroring set in the iNES header (should be set to vertical).
Can anyone else who has dumped this ROM confirm the CRC32? This is what I have:
32KB PRG: 982C8405
8KB CHR: E1AAE1ED
40KB ROM: 0F6EA0FF (combined ROM)
Thanks everyone!!
Yes, those CRCs match my dump.
I'm really curious about the quality of retro titles that are being (re)released now.
AFAIK,
TecToy is using the same 3.3v parts also I've posted on that thread the somewhat translated statement of a repro seller who claims these worries are all bullshit.
I think in 20 years or so we'll end having a big shortage of original retro systems, and I really wish I'm wrong!
But having a
defective SNES that probably was damaged because of this, I really don't think so.
So it appears I should have kept my mouth shut until I got my copy of Quest Forge at least.. A month after my purchase I messaged piko asking when I can expect my copy. Didn't get a reply, just a refund, and no reply to my follow up message inquiring about the refund. So I guess I'm officially black listed from playing piko released games unless I buy second hand or pirate... Looks like there's a limited edition up on ebay now I'll have to keep my eye on. Appears piko sells QF through ebay as well, but I assume my purchase would get immediately refunded there as well. I just want to play the game!
BacteriaMage wrote:
Thanks! I wasn't clear on your answer but my question was ambiguously stated. I know the ROM would be safe once extracted. What I meant to ask was: Is it safe for my Kazzo (and me as the operator) to connect the cartridge to the Kazzo to extract the ROM? Is the Kazzo tolerant to the board's design or could it also be damaged? Moot point now though since I've already done it.
Oh right, the current kazzo design has 5v i/o so technically it will have some of the same concerns as 5v consoles. However considering it only takes a min or less to dump I wouldn't be too worried about it. That's not long enough for any overheating concerns like there is for playing the game for hours on end in a console.
Fisher wrote:
I'm really curious about the quality of retro titles that are being (re)released now.
My copy of
Mystery World Dizzy arrived this week. The quality of printed materials and everything is top notch. But I was rather disappointed to find the same Chinese board inside the cartridge though. Guess it all makes sense though as piko branding was on the back label so they produced the carts per what appears standard procedure now. Safe to say fusion retro books & Oliver twins don't realize the hardware decisions/risks much like ludosity I'm sure.
I was curious about Watermelon Games' PCBs.
So I searched a little and found some pictures:
Attachment:
File comment: Back
2zqeae1.jpg [ 423.12 KiB | Viewed 6651 times ]
Attachment:
File comment: Front - Epoxy
20sf2ib.jpg [ 431.84 KiB | Viewed 6651 times ]
Attachment:
File comment: Front free
vSjowIZ.jpg [ 561.57 KiB | Viewed 6651 times ]
Is it dangerous?
I think the CPLD is 5v tolerant, but what about the straight connections to the ROM?
It looks like all the signals go through the MAX 2, so the ROM should be safe.
Thanks!
It's nice to known these games are (or at least should be) safe!
Fisher wrote:
I was curious about Watermelon Games' PCBs.
dbelectronics found his glob top WMgames cart to be excessively hot while playing and found the 5v signals were getting clipped due to 3v clamping diodes. It was one of the motivating factors for him writing his
article, but nice to see some of the carts appear to handle translation properly.
infiniteneslives wrote:
So I guess I'm officially black listed from playing piko released games unless I buy second hand or pirate... Looks like there's a limited edition up on ebay now I'll have to keep my eye on.
I missed out on the ebay auction, but I heard back from the developer and they were able to help me so I'm a happy camper now.
So mapper 268 on the wiki mentions the MindKids board. Mapper 268 appears to be specifically for Coolboy. Is the MindKids board a submapper, or will it be assigned its own mapper?
I've dumped my NES Wisdom Tree Collection cart, and I hacked up a version of FCEUX that works with it, so now I want to know what mapper to use add to the header permanently. It requires a Coolboy type mapper, but the registers are at 0x5000-0x5FFF (4 registers), not 0x6000-0x6FFF. It has work RAM that it uses at 0x6000 for game patches.
The Noah's Ark USA release is also a MindKids board, but it runs with the same mapper as the original MMC3. The MindKids mapper is technically more accurate, but doesn't really matter for that one from a functional point of view.
I originally assigned mapper 268 for the Coolboy mapper, meaning the additional registers are at $6000, then just emulated Mindkids under the same mapper number, just having the mapper emulation respond to both $5000 and $6000. I have since come across carts for which this does not work, hence, I intend to assign mapper 268 submapper 1 to Mindkids.
One of these is a Chinese RPG (楚留香新传, Chǔliúxiāng xīn Zhuàn) running the Mindkids board in GNROM mode while using WRAM --- since the "lock" bit does not work in GNROM mode, WRAM writes to $6000 would inadvertently change the mapper configuration when responding at both $5000 and $6000.
There are Chinese RPGs from Hengedianzi that use the original Coolboy board as an oversize TNROM board (i.e. using the full eight bits of MMC3 registers 6 and 7 for 2 MiB of PRG-ROM space) with WRAM, even though the ASIC registers sit at $6000, by carefully using the WRAM Enable and Lock bits. They boot in the last 8 KiB bank of the first 512 KiB block, like a regular MMC3 would, then switch to 2 MiB mode using the following sequence that must be run from RAM at $000-$7FF:
Code:
LDA #$00
STA $A001 ; Disable WRAM, so that writing to the Coolboy ASIC does not corrupt saved game
LDA #$00
STA $6000 ; clear outer bank offset and lower mask bits
LDA #$60
STA $6001 ; Enable PRG A19-A20
LDA #$80
STA $6003 ; Lock the Coolboy ASIC registers, so that WRAM writes do not change them
LDA #$80
STA $A001 ; Enable WRAM
JMP ($FFFC) ; Execute RESET Handler in the last 8 KiB bank, now in 2 MiB address space
Coolboy/Mindkids (268) and another (mapper 91) submapper proposal added to
submapper proposal page. No response after two weeks shall mean approval, after which I shall copy to main submapper page. No test ROMs will be made unless specifically requested.
Edit: Submapper added.