The PowerPak has the potential to run any commercially NES game/ROM, and runs about 99% of them perfectly (I am not including Famicom here). Testing the ROMs with the latest combination of the PowerPak (Boot ROM 1.11, Mapper Pak 1.30 & Resistor Fix), here are some of my results.
I have only tested NES and a few Famicom (Disk) Games
Dirty Harry - The War Against Drugs (U) [!p].nes
Paperboy (U) [!].nes
No input recognized from NES controller. The ROM Paperboy (U) [f1].nes works fine.
Karnov (U) [!].nes
Severe background graphical glitching.
Bandit Kings of Ancient China (U) [!p].nes
Castlevania III - Dracula's Curse (U) [!].nes
Gemfire (U) [!].nes
Laser Invasion (U) [!].nes
Nobunaga's Ambition 2 (U) [!p].nes
Romance of The Three Kingdoms II (U) [!].nes
Uncharted Waters (U) [!].nes
MMC5 is not fully implemented, shows a black screen.
Batman - Return of the Joker (U) [!].nes
Shows just a solid grey screen. Worked (mostly) with Boot ROM 1.00 and Mappers 1.00.
Hot Slot (Unl).nes
Severe graphical glitches past title screen.
Peek-A-Boo Poker (Unl) [!].nes
Severe graphical glitches in female full-screen portraits.
Metroid v1.01 (1986)(Nintendo).fds - Metroid v1.02 (1986)(Nintendo).fds
Any version of Metroid for the fds will either refuse to load a game, or will load to a severely glitched screen, with Samus unable to move. The graphics glitches start to show themselves when it asks the player to set side B.
Interestingly, Low-G-Man did not display any sound glitches at all during the first boss battle. Both Punch-Outs!! worked without glitches, as did the rest of the FDS games I tried. Games like Alien Syndrome and Death Race worked fine as far as I could tell. Games that Nestopia assigns Mapper 206 also worked with the exception of Karnov.
In evert case, I used NTSC versions of the games, as that is my region. All ROMs were from the GoodNES 3.14 set.
Let me make a wild guess...
Someone stuck a 4 page limit on CNROM games?
Dwedit wrote:
Let me make a wild guess...
Someone stuck a 4 page limit on CNROM games?
If that is so, then I would guess that similar problems would also be encountered in Bubble Bath Babes (Unl) [!].nes, even though it is assigned Mapper 3. Four pages would be the canonical Nintendo limit.
Assuming the CNROM mapper used is the one in the sample mapper, yes it has a four page limit.
Official CNROM games may have a 4 page limit, but iNES Rom files using Mapper #3 still expect up to 256 pages.
Dwedit wrote:
Official CNROM games may have a 4 page limit, but iNES Rom files using Mapper #3 still expect up to 256 pages.
How 256? I thought the maximum CHR length was 255 pages, and value 0 represented something like CPROM but with 8 KiB CHR RAM switching instead of 4 KiB.
As I understand it, the difference between official CNROM and clone boards is as follows:
CNROM
- runs CPU D0-D1 and D4-D5 to 74HC161 inputs D0-D1 and D2-D3
- uses Q0-Q1 from the 161 as the bank number
- may run Q2-Q3 through diodes as a copy protection measure
Clone boards
- run CPU D0-D3 through the 161
- use Q0-Q3 as the bank number
Well I guess mapper #3 is limited to 8-bits, but the limitation in iNES header itself would imit it to 7-bits.
I hacked both Paperboy and Dirty Harry at some point to fix them. Basically some issue with left over value on the bus or something to that effect is expected or not and the opposite happens causing the game which expects the same situation as the original board to occur to believe no buttons are pushed. I do believe this problem was supposed to have been fixed.
MMC5 is known to be unsupported so far. Would be nice if it worked atleast partially.
Metroid on FDS worked for me at some point. Never had any problem as you describe.
Crystalis seems to have some graphical issues as well.
crade wrote:
Crystalis seems to have some graphical issues as well.
Yes, I can observe a small amount of tile garbage here and there when walking on the map outside the first town.
Are you certain that isn't normal? I haven't played the game in awhile but it seemed to run ok to me.
I am certain it doesn't happen with the regular cart. The game actually isn't really playable if you try to go too far I think talking to the characters ends up illegible sometimes (and I happen to have the actual cart handy to compare)
I just threw the powerpak to test again.. It's the shops where the early more serious glitches happen... you cannot tell what is for sale in the shops due to graphical glitches. Same dump works correctly in nestopia. I'm only on boot 111 and mapper pak 1.2 at the moment though
MottZilla wrote:
I hacked both Paperboy and Dirty Harry at some point to fix them. Basically some issue with left over value on the bus or something to that effect is expected or not and the opposite happens causing the game which expects the same situation as the original board to occur to believe no buttons are pushed. I do believe this problem was supposed to have been fixed.
. . .
Metroid on FDS worked for me at some point. Never had any problem as you describe.
Could you post an IPS-type patch file for each game?
As for Metroid, I get this with half of the images out there:
http://www.youtube.com/watch?v=jO8m75qxLokThe other half I get a horribly garbled Brinstar starting screen and I can't move Samus around anywhere.
crade wrote:
I just threw the powerpak to test again.. It's the shops where the early more serious glitches happen... you cannot tell what is for sale in the shops due to graphical glitches. Same dump works correctly in nestopia. I'm only on boot 111 and mapper pak 1.2 at the moment though
I did not test extensively, but I do not experience anything like that on the PowerPak. The small graphics glitches haven't really affected gameplay, but I can also confirm they don't exist with a real Crystalis game pak. [/code]
Perhaps a bad dump of Metroid?
-Rob
Great Hierophant wrote:
MottZilla wrote:
I hacked both Paperboy and Dirty Harry at some point to fix them. Basically some issue with left over value on the bus or something to that effect is expected or not and the opposite happens causing the game which expects the same situation as the original board to occur to believe no buttons are pushed. I do believe this problem was supposed to have been fixed.
. . .
Metroid on FDS worked for me at some point. Never had any problem as you describe.
Could you post an IPS-type patch file for each game?
As for Metroid, I get this with half of the images out there:
http://www.youtube.com/watch?v=jO8m75qxLokThe other half I get a horribly garbled Brinstar starting screen and I can't move Samus around anywhere.
crade wrote:
I just threw the powerpak to test again.. It's the shops where the early more serious glitches happen... you cannot tell what is for sale in the shops due to graphical glitches. Same dump works correctly in nestopia. I'm only on boot 111 and mapper pak 1.2 at the moment though
I did not test extensively, but I do not experience anything like that on the PowerPak. The small graphics glitches haven't really affected gameplay, but I can also confirm they don't exist with a real Crystalis game pak. [/code]
I had that problem. Loopy fixed it for me by adding some delays
I am running I think mapper files 1.20 or 1.30 I can't recall and Metroid along with every other FDS game runs perfectly. Do your other FDS games work? By sure you took the original BIOS and got the patch applied to it or just get the patched BIOS for use with powerpak. The next thing would be to make sure that image works on an emulator and if all that checks out you should try another CF card or don't delete the copy of the game that is not working but instead leave it there and copy another file on and see if the new one works as maybe your CF card has some bad sectors.
Like I said it's most likely timing problems. I had the exact same problem (exact) and loopy fixed them by adjusting the timing. Seems certain powerpaks differ from others.
Great Hierophant wrote:
I did not test extensively, but I do not experience anything like that on the PowerPak. The small graphics glitches haven't really affected gameplay, but I can also confirm they don't exist with a real Crystalis game pak.
Would you mind testing going the armor or items shop in the first town? In the shops, the large images representing the items available for sale are mangled and all look the same. You can still play though. The items become clear when they are individually highlighted (didn't notice that before).
I upgraded to the 130 mappers as well to make sure that wasn't the issue.
So MMC3 still isn't right?
MottZilla wrote:
I am running I think mapper files 1.20 or 1.30 I can't recall and Metroid along with every other FDS game runs perfectly. Do your other FDS games work? By sure you took the original BIOS and got the patch applied to it or just get the patched BIOS for use with powerpak. The next thing would be to make sure that image works on an emulator and if all that checks out you should try another CF card or don't delete the copy of the game that is not working but instead leave it there and copy another file on and see if the new one works as maybe your CF card has some bad sectors.
Yes, other FDS games work. I use the BIOS contained within the 1.30 Mappers. The images work just fine with Nestopia. I tried the following images:
Metroid (v1.01).fds
Metroid (v1.2).fds
Metroid v1.01 (1986)(Nintendo).fds
Metroid v1.01 (1986)(Nintendo)[a].fds
Metroid v1.01 (demo) (1986)(Nintendo).fds
Metroid v1.01 (demo) (1986)(Nintendo)[a].fds
Metroid v1.02 (1986)(Nintendo).fds
Metroid v1.02 (1986)(Nintendo)[a].fds
All suffer from one of the two problems described above, so I sincerely doubt my CF card has bad sectors.
crade wrote:
Would you mind testing going the armor or items shop in the first town? In the shops, the large images representing the items available for sale are mangled and all look the same. You can still play though. The items become clear when they are individually highlighted (didn't notice that before).
Something like that happened to me when I played it at one point, but I played it twice and could notice any graphical glitches, even the ones I described above. I think it may be an issue with the resistorpak fix.
I also have never been able to get FDS metroid to work on the powerpak. I used Loopy's fds mapper.
Well I never did anything special with Metroid. It just works for me. Nothing strange. Pretty sure its version 1.02 or something like that.
powerpak_fds_alt.zip
If you're having problems with FDS, give this one a try.
loopy wrote:
powerpak_fds_alt.zipIf you're having problems with FDS, give this one a try.
The FDS emulation doesn't work at all with this one, it won't even load to the BIOS screen, nevermind a game.
Great Hierophant wrote:
The FDS emulation doesn't work at all with this one, it won't even load to the BIOS screen, nevermind a game.
Doh... I thought that one was working. My powerpak has problems right now, so I'll have to fix it some other time.
All the games that do not register input work correctly now. See here for the fix:
http://nesdev.com/bbs/viewtopic.php?t=3698
So, we are down to the MMC5 games, the Panesian games, Batman: Return of the Joker, Karnov and Metroid FDS. All but Karnov have to be mapper fixes.
I think I know what the issue with Karnov is. Karnov is set to Mapper 4, but really uses the Namco-109, which only supports a very small portion of the MMC3 functionalities. When Karnov tries to switch banks, it must be using a value with a 1 in bit 7. In MMC3, this causes a character bank invert, which should explain the graphic errors. In Namco-109, this bit does nothing. So, the fix is to find the value which Karnov is using and change it to have a 0 in bit 7. Unfortunately, I cannot find the value myself.
Batman RotJ and Metroid FDS work fine for me. I have no idea why they don't work for you.
FDS DiskChange Report
Japanese Official Old Famicom
1.20:OK 1.30:OK
Japanese Official New Famicom
1.20:OK 1.30:OK
Japanese Official Twin Famicom
1.20:OK 1.30:OK
US Official Nes
1.20:OK 1.30:OK
Japanese Unofficial Famicom (
Familator)
1.20:OK 1.30:NG
dosirouto wrote:
Japanese Unofficial Famicom (
Familator)
1.20:OK 1.30:NG
I guess that being japanese doesn't make clones any better than the standard NOAC.
I wish I knew Verilog so I could write an open-source NES clone in it. Then the famiclone makers would have no excuse.
Does anyone know if mapper 2 (UxROM) games with 512KB or PRG-ROM work on the PowerPak? Is there a page limit like there seems to be for CNROM?
Great Hierophant wrote:
Metroid v1.01 (1986)(Nintendo).fds - Metroid v1.02 (1986)(Nintendo).fds
Any version of Metroid for the fds will either refuse to load a game, or will load to a severely glitched screen, with Samus unable to move. The graphics glitches start to show themselves when it asks the player to set side B.
FWIW, I got a new powerpak and I'm seeing similar symptoms now... so maybe we can figure this out.
First, the FDS mapper in 1.31 (F.MAP) will work better than the old one (1.20). So make sure you have that.
If that doesn't work, you can give this one a try.
http://home.comcast.net/~olimar/NES/powerpak_fds_alt.zip I made some small changes to it that that might make it work better.
Is Metroid the ONLY game that doesn't work? For me, when Metroid isn't working, Zelda also fails (freezes after you enter a name and start the game). So if it's only Metroid, maybe we're looking at a different problem.
anyone got FELIX THE CAT to work ? just got a white screen with music
Fab wrote:
anyone got FELIX THE CAT to work ? just got a white screen with music
Works fine for me... Check if it's a header thing (download a new copy, fix the header, whatever).
well it does work on a AV Famicom with AV output but not on the one that is RGB modded... other games work fine on it.
Nintendo's NES-compatible RGB PPU, used in PlayChoice, Famicom Titler, and some rare brand of TV with a built-in Famicom chipset, turns the screen white if the game turns on all the tint bits in PPUMASK. That's why Just Breed's packaging carries a warning about using it on that TV.
loopy wrote:
Is Metroid the ONLY game that doesn't work?
For me, every version of Metroid Great Hierophant tried (except the ones with parenthetical version numbers, I couldn't try them) refuses to load a save, with both the "regular" and alternate FDS mapper versions. All the versions of Zelda ("regular," [a], and [a1]) I've tried seem to work fine with both mappers.
Other FDS issues I've seen (all of these happen with both mappers for me):
Puzzle Boys (1990)(Atlus Ltd.) - freezes when saving.
Lutter (1989)(Athena) - Freezes when trying to save from the menu with some really weird momentary horizontal blue artifacts on the whole screen.
Namida no Soukoban Special (1986)(ASCII Corp.) - After a brief flash of "ERR 04," Keeps cycling between "Please wait" and "Flip to side A (or whatever), ERR 24" when attempting to return to the menu. Also freezes (on "please wait") when returning to the menu from edit mode.
Silviana (19xx)(Pack-In-Video Co.) - Black screen after FDS BIOS.
Kalin no Tsurugi (1987)(Xtalsoft - DOG) and [a] - Both versions seem to freeze whenever I try any writing to the disk, whether deleting a save game or saving a game already on the disk. Strangely, the translated version works fine.
I tried all these games on Nestopia 1.4 and saw none of the above described issues, though I couldn't try saving in Lutter because the B button wasn't working.
Whew. Well, I hope this information is of some use
tepples wrote:
Nintendo's NES-compatible RGB PPU, used in PlayChoice, Famicom Titler, and some rare brand of TV with a built-in Famicom chipset, turns the screen white if the game turns on all the tint bits in PPUMASK. That's why Just Breed's packaging carries a warning about using it on that TV.
what ? Are there more games like that? Why a waring ?!
I had to remodify my rgb modded famicom to make it work with the Powerpak. Don't know what my modder had to change...
A warning cause it won't work. There's 1 or 2 games that shouldnt work iirc. A final fantasy has this problem iirc.
Was playing Final Fantasy III recently and there is a glitch that prevents me from progressing forward in the game. I thought it might have something to do with the translation patch, but when I used the original ROM I ran into the same issue. The first airship disappears when I come back from a battle in a town or castle. This doesn't happen when playing on an emulator.
Quote:
Was playing Final Fantasy III recently and there is a glitch that prevents me from progressing forward in the game. I thought it might have something to do with the translation patch, but when I used the original ROM I ran into the same issue. The first airship disappears when I come back from a battle in a town or castle. This doesn't happen when playing on an emulator.
I just noticed the -exact- same glitch. I was playing it on the powerpak, I go into the first cave and come out to go back to the town, and the airship is gone! This has to be something involving the WRAM. I have to test it out, but I think it happened when I didn't load a .sav file created when I first opened the game on my computer. This -could- have an effect on it. I'll have to see if it happens when I open the .sav file associated with the ROM on my CF. Also, I've noticed another VERY annoying glitch. Perhaps it's just me, but you know how you sometimes want to shut off the NES and save your game? Yeah, that never works for me with FF3 on the powerpak. I save my game, hold reset for 2-3 seconds, and it asks me if I want to write to a .SAV file. I do so, and shut the system off there. When I open it back up with that SAV file, no data was saved. Works in FF1, but not FF3. Has anyone else experienced problems with this on the powerpak? Am I saving my games correctly?
I noticed with the game Base Wars, the saving doesn't work. Not sure why, I didn't look at the header. I'm pretty sure there was a SAV file.
I've also seen glitchy scanlines on Zelda 2 sometimes, I have no clue why that would be. It doesn't happen very often.
Celius wrote:
Quote:
Was playing Final Fantasy III recently and there is a glitch that prevents me from progressing forward in the game. I thought it might have something to do with the translation patch, but when I used the original ROM I ran into the same issue. The first airship disappears when I come back from a battle in a town or castle. This doesn't happen when playing on an emulator.
I just noticed the -exact- same glitch. I was playing it on the powerpak, I go into the first cave and come out to go back to the town, and the airship is gone! This has to be something involving the WRAM. I have to test it out, but I think it happened when I didn't load a .sav file created when I first opened the game on my computer. This -could- have an effect on it. I'll have to see if it happens when I open the .sav file associated with the ROM on my CF. Also, I've noticed another VERY annoying glitch. Perhaps it's just me, but you know how you sometimes want to shut off the NES and save your game? Yeah, that never works for me with FF3 on the powerpak. I save my game, hold reset for 2-3 seconds, and it asks me if I want to write to a .SAV file. I do so, and shut the system off there. When I open it back up with that SAV file, no data was saved. Works in FF1, but not FF3. Has anyone else experienced problems with this on the powerpak? Am I saving my games correctly?
I ran into the same glitch where the airship disappears. I think I still have the before and after disappearance .SAV files, if that helps.
So today, I had a bit of extra time to myself so I thought I'd mess around with Final Fantasy 3 on my PowerPak and try to figure out what the heck was going on with it. I haven't had much success.
I don't know besides the airship glitch if anybody else is having any problems, but I'm having a major problem with FF3 relating to SRAM. It started out when I noticed my FF3 saves weren't saving or loading. So that's when I tried out a couple of other things to see exactly how it was behaving.
When I don't load any .SAV file, and continue to play the game for a while until I can save, I try to dump the save afterwards to a blank .SAV file. The data that is dumped into that file is not even related to FF3; it's data that was in $6000-$7FFF from the PowerPak menu. When I load an existing FF3 SAV file that I created on my computer, FF3 doesn't load/recognize the data, but when I play until a save point, dumping afterwards to a blank .SAV file dumps the data from the .SAV file I attempted to load previously.
So it may be that WRAM read/writes aren't working. I tried to see if for some reason the copy I had was bad and wasn't enabling reads or writes, but it wasn't the case. I overwrote the first couple executed lines of code in the ROM to:
Code:
lda #$80
sta $A001 ;Enable SRAM chip, allow reads/writes
sta $6000 ;Write to SRAM
loop:
jmp loop
Dumping the contents to a blank .SAV file showed that $80 was not written to $6000. It behaved the exact same way as FF3 did without that code.
I'm pretty confident whatever problem I'm experiencing here is directly related to the airship glitch. That information about the airship's location is undoubtedly stored somewhere between $6000-$7FFF. If it can't read/write data properly to that memory location, it's no wonder there are issues with this sort of thing.
Also, I'm able to load/save my game just fine with FF1 and some other games, so it seems specific to this mapper (or with the emulation of TNROM).
Does anybody have this issue as well? Am I the only one who's not able to load/save with FF3?
Okay, so an interesting update here. I made a test ROM on the PowerPak that uses MMC3, battery enabled, 32k of PRG with 8k of CHR ROM. With that, I was still unable to read or write to SRAM. I was able to using various emulators. I concluded there must be some sort of issue with my MMC3 emulation on my PowerPak.
That's when I attempted to -downgrade- my PowerPak mappers. What do you know, as soon as I did that, I was able to load my FF3 save! I am also able to dump SRAM data to a SAV file. However, there are graphical issues that I believe were fixed in updates. I don't believe the mapper-controlled mirroring part of it was working properly before the updates, because the ROM I have is set to horizontal mirroring by default, though FF3's code switches it to vertical through register writes. When I turned it on, it was glitching like it was set to horizontal. After switching the header info to be set at vertical, everything worked fine.
I have yet to test the airship out to see if that glitch is still occuring. But for now, it seems downgrading was more of a solution than upgrading (which I had tried previously). I'm curious if anyone has issues with MMC3 games that use battery with Mapper versions 1.30+. Downgrading to 1.20 solved the problem I was having with the battery, so unless I'm not loading it correctly, I think there's a bug with the updated mapper version.
EDIT:
I have indeed verified the airship glitch is not present when playing FF3 on Mapper Version 1.20. Going in and out of Jinn's cave proves that the airship is still there. So I'm pretty positive there is a problem with the mapper update that prevents proper access to/from SRAM, which would provide adequate explanation for the issues I was having and the airship glitch.
Have you tried with MMC3 from loopy's site?
Hmm, no I haven't. I'll go ahead and give that a try a bit later.
EDIT:
Confirmed that it does not work with loopy's MMC3. So far the only version I've found that successfully loads/saves information is 1.20.
I had the same problem with FME7 when I was testing my game out. I eventually wrote up a test rom like yours and verified that WRAM was not being mapped in (which also breaks Batman:RotJ). I downgraded just that mapper file to the version in 1.20 and it worked.
Off the top of my head, I remember Tecmo Super Bowl saving correctly.
I think I found the bug (thanks to bunnyboy). Apparently loopy's MMC3 implementation maps a different 8K WRAM bank in (there's 32K WRAM) based on what PRG bank is mapped in at that time, so unless the write is done in bank 0, 4, 8, ... it will go to a wrong bank. Unless I'm mistaken
I informed loopy about this in IRC so hopefully he can fix it.
Yeah, remember how I said the airship glitch doesn't happen in 1.20? I lied. With version 1.20, it seems that after returning from a battle, the airship is gone. Going in and out of towns seems fine, but fighting a battle causes the airship to disappear. Very disappointing
.
Hi everybody, I just purchased the powerpak a while back in the hope to play some RPG classics on the NES. However, I encountered the same problem while playing ff3 (the airship glith).
thefox wrote:
...I informed loopy about this in IRC so hopefully he can fix it.
I was wondering if you guys already found a solution to this problem?
J-Aap wrote:
I was wondering if you guys already found a solution to this problem?
There is a solution (if I was right about the problem in the first place.. I never verified it with code). It's up to loopy or bunnyboy to fix it.
Meanwhile you can try if it works with my mappers:
http://kkfos.aspekt.fi/projects/nes/pow ... e-mappers/
Anyone got a savegame so I can test this airship bug more easily?
http://home.comcast.net/~olimar/NES/powerpak_mmc3test.zip
Give this a try. I fixed a careless bug thefox pointed out to me.
I also changed the IRQ code to make it more accurate. I don't know if it makes a difference in any games (probably not)
..the problem still occurs
J-Aap wrote:
..the problem still occurs
Did you start the game from the beginning with the new mappers or did you load a game that was saved with old version of the mappers? The only surefire way to test it is to play it from the beginning with the new mappers unfortunately...
nothingnew wrote:
http://www.mediafire.com/?yeba5cdcfemqxfh
3'rd save slot.
The cart only has 8KB of WRAM, so your 32KB save may already have something wrong.
Wouldn't it be faster to make a test ROM that just changes banks after writing known values to WRAM and testing them? Seems like it would be faster than playing a game like FF3 to some point that I'm sure isn't near the beginning.
ibeenew2 wrote:
nothingnew wrote:
http://www.mediafire.com/?yeba5cdcfemqxfh
3'rd save slot.
The cart only has 8KB of WRAM, so your 32KB save may already have something wrong.
Game doesn't save when using a 8KB save-file. Haven't tried using a 16KB one, but I don't think the savefile should matter as I've used them for 8KB games without problems.
EDIT: And are you sure it uses 8KB? I know Kirby uses at least 16 and it uses a similar save system.
MottZilla wrote:
Wouldn't it be faster to make a test ROM that just changes banks after writing known values to WRAM and testing them? Seems like it would be faster than playing a game like FF3 to some point that I'm sure isn't near the beginning.
Only takes an hour or two to get to that point IIRC. I'm going to get back to testing thefox's mappers as soon as i can. I'll test loopy's new one if the glitch still occurs.
nothingnew wrote:
Game doesn't save when using a 8KB save-file.
Odd. It saves on an 8KB file for me.
Edit:
It seems the airship glitch is fixed on thefox's mapper. I entered the ship, exited, entered a battle, exited the battle, and the airship was still there. I'll test this on loopy's mapper later.
nothingnew wrote:
I know Kirby uses at least 16
What? All my Kirby's Adventure saves are 8KB.
Use bootgods data base at
http://bootgod.dyndns.org:7777 Kirby also only has 8KB on cart.
joe1088 wrote:
MottZilla wrote:
Wouldn't it be faster to make a test ROM that just changes banks after writing known values to WRAM and testing them? Seems like it would be faster than playing a game like FF3 to some point that I'm sure isn't near the beginning.
Only takes an hour or two to get to that point IIRC. I'm going to get back to testing thefox's mappers as soon as i can. I'll test loopy's new one if the glitch still occurs.
But it only takes a couple seconds to run a test ROM to find out the same thing. But fair enough, so long as the glitch gets fixed we can all be happy.
Option A:
- Play through the game.
Option B:
- Make a test ROM.
- Verify its behavior in Nestopia and Nintendulator.
- Verify its behavior on an MMC3 devcart.
- Try it on PowerPak.
I'd prefer option B, and I've started on such a test ROM myself, but option A might not take as long.
EDIT: I made a test ROM, and I'm not entirely sure what the PowerPak is doing wrong.
ibeenew2 wrote:
Use bootgods data base at
http://bootgod.dyndns.org:7777 Kirby also only has 8KB on cart.
That's weird, I could never get Kirby to work on my Powerpak with 8KB files. I stand corrected.
joe1088 wrote:
Only takes an hour or two to get to that point IIRC..
Actually, it can even be done quicker. When you go straight to the airship, it shouldn’t cost you more than 45 min. Because of this, it wasn’t a problem for me to start from the beginning and continue playing until the glitch occurred (without using a savegame).
To illustrate how/when I encountered the glitch:
When I left the first cave (the cave where you start), I stocked up in the first town (UR) and then continued to the second town (KAZUS) to talk to Cid. After talking to Cid you are able to get the airship in the dessert (which is right next to KAZUS). So after talking to Cid, I picked up the airship and flew back to the first village (UR), where I parked the airship right next to it. In this first village I fought my way through the grass area to get the “cure” spell. After getting the spell I left town and…*surprise* the f**ng thing is gone (yes I took the keys with me
).
thefox wrote:
Did you start the game from the beginning with the new mappers?
To illustrate what I have done concerning the mappers:
I downloaded v1.34 of the mappers on the Retrozone site and then copied loopy’s mappers to the directory. So to be clear, I only copied the document below to the standard mapper directory.
loopy wrote:
There was glitching on WRAM /WE. I added a filter to M2 to fix it. This seems to have fixed Crystalis as well.
http://home.comcast.net/~olimar/NES/powerpak_mmc3test2.zip
Tepples I hope you still make your test ROM to compare the old mapper code to the new one loopy just posted.
Strange that such a bug didn't cause problems with more games. I thought alot of MMC3 games used WRAM. By some coincidence they didn't trigger this glitch?
I've made a ROM that fills PRG RAM with a pattern and then hex-dumps it. Does
$A001 bit 6 (read-only PRG RAM) exist? Nestopia, Nintendulator, and whatever version of mappers are installed on my PowerPak's CF card don't appear to honor it.
Doesn't Low G Man or something like that break if WRAM Protect isn't implemented?
The saving in
Downtown Special never worked for me, I didn't notice any side-effects though (other than PowerPak menu doesn't ask to save). I don't know what board it uses (Famicom-only) but I just assumed it's a normal MMC3/battery/WRAM because it seems OK in other emulators.
And Base Wars is another MMC3 game that refused to save (edit team or start a season - it's supposed to save everything automatically). Both of those, I mean on the old mappers, I don't have my NES hooked up for test the new ones at the moment.
I don't know if that's related to the airship bug, that is really strange.
loopy wrote:
There was glitching on WRAM /WE. I added a filter to M2 to fix it. This seems to have fixed Crystalis as well.
I tried the title screen of M.C. Kids, and it was broken. I installed your fixed mapper, tried it again, and it was fixed. Thank you.
Could you elaborate on what kind of conditions would have caused such a glitch in old MAP04 versions, so that I can trigger these conditions in my test ROM to make sure this doesn't happen again?
tepples wrote:
I've made a ROM that fills PRG RAM with a pattern and then hex-dumps it. Does
$A001 bit 6 (read-only PRG RAM) exist?
Yes, I was using it a few weeks ago and verified that it write-protects WRAM, but still allows reading.
loopy wrote:
There was glitching on WRAM /WE. I added a filter to M2 to fix it. This seems to have fixed Crystalis as well.
http://home.comcast.net/~olimar/NES/powerpak_mmc3test2.zip
This seems to work!! Played ff3 again and the airship glitch did not occur. Fantastic! Thnx Loopy
tepples wrote:
Could you elaborate on what kind of conditions would have caused such a glitch in old MAP04 versions, so that I can trigger these conditions in my test ROM to make sure this doesn't happen again?
PRG /CE (which is used by the mapper to get A15) changes slightly after M2. Writes to E000-FFFF would briefly assert WRAM /WE, causing ram corruption.
I applied the same fixes to my other mappers, and made a few other changes. It's interesting that thefox didn't have the same problem with his MMC3 mapper, I wonder why.
MMC3: improve IRQ timing
MMC3: add "open bus" read for 4016/4017 (fix Dirty Harry control problem)
VRC6: added mapper 26 (Esper Dream 2, Madara)
N106: IRQ bugs fixed
http://home.comcast.net/~olimar/NES/
If anyone has other requests, now would be a good time to ask them, while I'm in the mood
loopy wrote:
If anyone has other requests, now would be a good time to ask them, while I'm in the mood
What would be great is if there was an easier way for me to create PowerPak version of the Verilog modules I've made (for a CPLD cart). Some kind of Verilog template mapper would be great. I asked thefox for a similar type of template also and maybe I'll see that sometime, but I don't think it would hurt to have another example.
loopy wrote:
MMC3: add "open bus" read for 4016/4017 (fix Dirty Harry control problem)
Please elaborate! I've recently run into some issues with open bus bits not always being zero for $4017, and wonder whether this is related.
Memblers wrote:
What would be great is if there was an easier way for me to create PowerPak version of the Verilog modules I've made (for a CPLD cart). Some kind of Verilog template mapper would be great. I asked thefox for a similar type of template also and maybe I'll see that sometime, but I don't think it would hurt to have another example.
http://home.comcast.net/~olimar/NES/powerpak_loopy_src.zip
Not a template, but here's the source to the mappers I've made.
I'll try to get a decent template for you tomorrow.
Can anyone implement these mapper variants, if needed?
Implement Mapper 206? This should fix Karnov and possibly some Tengen games. Ditto for Mapper 144 and Hot Slots/Peek-a-Boo Poker.
Mapper 158 - Used for Alien Syndrome, but the game seems to work fine in the PowerPak using Mapper 4. Ditto for Death Race and Mapper 144/79.
These numbers are used primarily, if not exclusively, in Nestopia.
Despite the previous work, FDS Metroid still does not work in my Famicom AV. Can anything be done? The best I can get is to load to a Brinstar screen loaded with tiles all over the screen and unable to move.
loopy wrote:
If anyone has other requests, now would be a good time to ask them, while I'm in the mood
Mapper 90? MMC5? VRC7? Might as well reach for the sky!
Jeroen wrote:
loopy has an early version of mmc5 on his site..........it works with cv3 but has some glitches in other games.
I can't seem to find such anywhere, and CV3 shows a black screen with the current set.
Crystalis still seems to have issues even with your latest mappers. I play up the pont of findind the village elder in the cave where he tells you to get the lighthouse working, exit the cave & game crashes.
Edit 12/18/2010:
Just popped back on to mention I re-downloaded the updated Loopy map files & now Crystalis & others working fine for me.
MottZilla wrote:
loopy wrote:
If anyone has other requests, now would be a good time to ask them, while I'm in the mood
Mapper 90? MMC5? VRC7? Might as well reach for the sky!
I tried mapper 90. There are still some problems with it. You can download it with the others.
blargg wrote:
loopy wrote:
MMC3: add "open bus" read for 4016/4017 (fix Dirty Harry control problem)
Please elaborate! I've recently run into some issues with open bus bits not always being zero for $4017, and wonder whether this is related.
Probably not. All I did was set D7 and D5 low on 4016/4017 reads.
Thanks, loopy! This is pretty cool to see.
I'll try compiling one of those when I get home later. I'm presuming I still need the constraints file or something from the PowerPak schematic, so the pinout can get assigned, and all that. I'll find out soon enough.
OldNESJunkie wrote:
Crystalis still seems to have issues even with your latest mappers. I play up the pont of findind the village elder in the cave where he tells you to get the lighthouse working, exit the cave & game crashes.
I was able to get past that point just fine. However, the status bar is jumpy on the inventory screen and very occasionally when walking.
As soon as I saw your posts, I went to get my flash card reader and test out the fix, only to find my flash card reader had slipped off of my desk into the trash below, which I had taken out a few days prior to that! Needless to day, I was a bit disappointed.
But as soon as I get another, I'm going to test this out. I'm anxious to see the results, as I also have a test ROM of a polygonal rendering engine that uses MMC3 + WRAM that didn't work previously, as it uses 5k of WRAM. Thanks so much for the fixes!
Just popped back on to mention I re-downloaded the updated Loopy's mapper files & now Crystalis & others working fine for me.
loopy wrote:
I applied the same fixes to my other mappers, and made a few other changes. It's interesting that thefox didn't have the same problem with his MMC3 mapper, I wonder why.
MMC3: improve IRQ timing
MMC3: add "open bus" read for 4016/4017 (fix Dirty Harry control problem)
VRC6: added mapper 26 (Esper Dream 2, Madara)
N106: IRQ bugs fixed
http://home.comcast.net/~olimar/NES/If anyone has other requests, now would be a good time to ask them, while I'm in the mood
I just got a PowerPak for Christmas and the main mapper I was sad to see not working was the alternate VRC6 used in ED2 and Madara. I was surprised it hadn't been written because CVIII worked so well and I thought this varient was not a large change. So, I decided to see if I could request that it be worked on, and low and behold it had already been done! Thanks loopy!
Well, I just tried the AG translation of ED2 and it seems to work perfectly.
But, when I tried the japanese version of Madara it had serious graphical glitches, wrong tiles shown in seeminly random places, starting with the title screen and wierd music from the intro onward.
Has anyone else seen these same issues with Madara?
And just in case you are still in the mood, a few less pressing requests:
1) All of our completed translations at DvD Translations work great on the Power Pak except surprisingly, Valkyrie no Bōken: Toki no Kagi Densetsu. It is unplayable from the point where you start to walk around. I checked and it has the same problem with the original rom too so it isn't an artifact of the translation.
2) We're currently working on a translation of a two disk FDS game. It would be neat to see it work on a real NES, going somewhere past the title screen, but I'm sure getting an FDS game with more than two sides to work will be a challenge to figure out how the user will pick the proper side. I guess an automatic way for the NES to do it that might work is you could have it just try every side in order until it doesn't get a load error? I've done extensive work disassembling the FDS BIOS so I would help on this one if I can.
3) Lastly, we are working on a game called Pocket Zaurus who's mapper, 152, hasn't been implmented yet. No biggie.
I bought a Front Loader at a Goodwill that looked to be in OK shape, but the first cartridge booted on the first try, so did the second. Results from this with loopy's latest mappers demonstrate:
Crystalis seems fixed of its jumpy menus.
Batman : ROTJ is now working properly.
Metroid FDS shows no change from before.
I have yet to mod my NES for Powerpak audio, need that 47kohm resistor.
Great Hierophant wrote:
I have yet to mod my NES for Powerpak audio, need that 47kohm resistor.
They carry them at Radio Shack. Just ask the guy at the counter. Or I could mail you one
qbradq wrote:
Great Hierophant wrote:
I have yet to mod my NES for Powerpak audio, need that 47kohm resistor.
They carry them at Radio Shack. Just ask the guy at the counter. Or I could mail you one
Or go into the back. There usually is a small section of drawers that contain basic electronic components.
Kind of a pity how the store is mostly cell phones and junk now, rather than electronic parts.
Edit: Also, I found a bug in the PowerPak's interface.
If you have a battery-backed game that you played, the next time you turn on your PowerPak, if you hit start (to access the last game you played), it won't load the SRAM. Instead, you have to go to the file browser and load the game from there.
I'm using the default PowerPak mappers/firmware along with Loopy's updates.
Hello there, I'm a regular poster at NA. I tried loopy's mappers and saw that Batman ROTJ works... at first. I encountered serious glitches in level 3-2, second section (a roof of spikes that goes up and down to quickly, makes impossible to pass through).
So, just wanted to let you know, I hope loopy reads this.
Sounds like an IRQ problem of some kind.
Hmm.. Sunsoft mapper. I'll take a look. Do you have a password to the buggy stage?
loopy wrote:
Hmm.. Sunsoft mapper. I'll take a look. Do you have a password to the buggy stage?
http://www.gamewinners.com/nes/BatmanReturnOfTheJoker.htm
joe1088 wrote:
loopy wrote:
Hmm.. Sunsoft mapper. I'll take a look. Do you have a password to the buggy stage?
http://www.gamewinners.com/nes/BatmanReturnOfTheJoker.htm
heh, you beat me to it joe. Thank you for the attention given loopy.
EDIT: with the password for stage 3-2 you don't get to the buggy stage directly. You have to play through the first part of the stage (which is normal) to get there.
buttheadrulesagain wrote:
Hello there, I'm a regular poster at NA. I tried loopy's mappers and saw that Batman ROTJ works... at first. I encountered serious glitches in level 3-2, second section (a roof of spikes that goes up and down to quickly, makes impossible to pass through).
So, just wanted to let you know, I hope loopy reads this.
Fixed it. (changed the IRQ behavior)
Hello, I noticed that Joshua (Wisdom Tree) gets stuck with a grey screen of death after completing the first level. This doesn't happen with Exodus, which has the same engine. If someone has time to look into it, it'd be nice. Thanks.
hey I tried installing your powerpak mappers but on kirbys adventure the stats bar (health, current power, etc) started going up and down like crazy. Although I was playing it on a rgb modified av famicom but when I use other mappers I don't have this problem. Have any idea what about your mappers causes this?
At this point, the nesdevcompo multicart is a problem PowerPak game. It's a mapper 34 (BxROM) game with 256 KiB PRG ROM. After I discovered that it wouldn't run on a PowerPak, I tried changing the header to mapper 7 and got the menu and four out of five games running. Then I made a test suite demonstrating that mapper 34 on the PowerPak can see only 128 KiB, while mapper 7 can see 256 KiB.
Download BNTestNestopia and FCEUX:
Both mappers allow access to the entire 512 KiB, as could be supported by a clone board.
Nintendulator:
Mapper 34 allows 512 KiB, but mapper 7 allows 256 KiB. This is consistent with allowing 4 bits (one 74HC161) worth of state.
PowerPak mappers 1.34:
Mapper 7 allows 256 KiB, and mapper 34 allows only 128 KiB, to the maximum supported by those cartridge boards manufactured by Nintendo.
I've added a fixed mapper 34 to my mapper pack.
Link
Works. Thanks.
EDIT: Can someone celebrate with rule 34 of Deadly Towers?
I don't know if this is an issue or not, I searched on Google and your site. Is there a known issue with megaman3's pause/weapons menu being transparent with NES Powerpak? The only thing that shows up is the "E" and mega's head. It makes the game un-enjoyable and almost not playable. I have tried 2 NES's and 4 different NES roms(including the b's and o's). I own the game, but it would be nice to include it on my Powerpak.
Evotistical wrote:
I don't know if this is an issue or not, I searched on Google and your site. Is there a known issue with megaman3's pause/weapons menu being transparent with NES Powerpak? The only thing that shows up is the "E" and mega's head. It makes the game un-enjoyable and almost not playable. I have tried 2 NES's and 4 different NES roms(including the b's and o's). I own the game, but it would be nice to include it on my Powerpak.
See this thread:
http://nesdev.com/bbs/viewtopic.php?t=7879
And bunnys fix from his site:
http://www.nespowerpak.com/powerpakmod.html
You sure it's the powerpak?
Evotistical wrote:
Is there a known issue with megaman3's pause/weapons menu being transparent with NES Powerpak?
I never heard anything about this... This is really weird! Can you post some pictures?
The Powerpack I bought is brand new (1 week old) would it still effect it?
This is normal play:
Pause menu:
I have no issues with any other game, (Ive played like 50). And as you can tell by my other NES, I'm not b'sing. Does anyone else have this issue?
Nope, it should be fixed. But that is very strange. You try other ROM's?
All of the good tools roms pack for it, including the b's and the one o'. BTW I'm not a fan of rockman3. I need my mega!
That menu needs scanline IRQs to be rendered. For some reason the IRQ doesn't seem to be firing. Did you try different mapper files? Maybe the MMC3 code in your POWERPAK folder is broken. BTW, it has nothing to do with the sprite issue that was mentioned.
Just tried a mm3 rom from D*PER0Ms, and still no-go.
tokumaru wrote:
That menu needs scanline IRQs to be rendered. For some reason the IRQ doesn't seem to be firing. Did you try different mapper files? Maybe the MMC3 code in your POWERPAK folder is broken. BTW, it has nothing to do with the sprite issue that was mentioned.
So I should re-download the powerpak folder?
I'd recommend redownloading the POWERPAK folder from retrousb.com and then applying
Loopy's latest PowerPak mappers on top of it.
I just tried MM3 with no problems on my PowerPak (not using loopy's latest set of mappers yet either).
Worked!!! I think It just shipped with corrupted mappers. Hope this doesn't effect anyone else.
On a different note:
Now that I have all up to date mappers, Is there any other official or Tegen games I should be worried about not working or having glitches (not counting Castlevania III). Is there a work around for karnov? Any other fixes I need to implement? Thanks so far!!!
BTW wierd quirk now that my first issue was fixed....MM3 flickers all over on the bottom 1/3 of the screen on 1 NES but not the other..... weird.....
Glad to have 2
It's really weird that the PowerPak works as expected in one system but not the other... Did you try the actual MM3 cart you said you have?
Evotistical wrote:
Worked!!! I think It just shipped with corrupted mappers. Hope this doesn't effect anyone else.
On a different note:
Now that I have all up to date mappers, Is there any other official or Tegen games I should be worried about not working or having glitches (not counting Castlevania III). Is there a work around for karnov? Any other fixes I need to implement? Thanks so far!!!
BTW wierd quirk now that my first issue was fixed....MM3 flickers all over on the bottom 1/3 of the screen on 1 NES but not the other..... weird.....
Glad to have 2
You can also try my save state enabled mappers if you're adventurous:
http://kkfos.aspekt.fi/projects/nes/pow ... e-mappers/
thefox wrote:
You can also try my save state enabled mappers if you're adventurous:
http://kkfos.aspekt.fi/projects/nes/pow ... e-mappers/What do these do specifically? Sorry I'm not very dev oriented. I just plays the games
tokumaru wrote:
It's really weird that the PowerPak works as expected in one system but not the other... Did you try the actual MM3 cart you said you have?
Yea, same issue. Like I said I'm glad I have 2
. I wonder what I need to do to fix it? Might just build a nes PC
.
Evotistical wrote:
thefox wrote:
You can also try my save state enabled mappers if you're adventurous:
http://kkfos.aspekt.fi/projects/nes/pow ... e-mappers/What do these do specifically? Sorry I'm not very dev oriented. I just plays the games
They allow you to save/load at any point during the game. BUT they are also complete rewrites of the mappers, so some things like MMC3 IRQ *might* work better with these than other versions. I say might because I've never tested them with an NTSC NES.
Wow, that's pretty awesome. How are you managing to save internal PPU states and RAM? Are you just doing this on NMI or are you using some interrupt hooking?
qbradq wrote:
Wow, that's pretty awesome. How are you managing to save internal PPU states and RAM? Are you just doing this on NMI or are you using some interrupt hooking?
Yup it simply hooks the NMI when it detects a certain button sequence on the data bus. It doesn't save the entire PPU state, like address registers, but this doesn't seem to be a big deal for most games, since they rewrite those registers almost always in the NMI handler anyways. Everything else is read back like any other application would, using $2006/$2007.
I just connected up the expansion audio in my NES using a 100k potentiometer. I then adjusted volume levels so that a 2a03 square at vol 15 would be equal to a vrc6 square at vol 15 (same as many emulators.. hope this is correct).
Now some fun stuff.
1) With loopy's mapper pack installed, Gimmick's NSF doesn't play the sunsoft channels. However, the actual game, Gimmick!.nes, DOES play the expansion audio. What's up with that??
2) MMC5 isn't implemented yet, etc. Is there a way that a stand-in could be made for NSFs that use MMC5? Either that or have the NSF code use 2 vrc6 squares instead.
3) Namco106 is incredibly quiet compared to all of the other expansions. Why is this? I can use the potentiometer to correct this, but I do not want to have to change volume every time I change expansions.
loopy wrote:
I applied the same fixes to my other mappers, and made a few other changes. It's interesting that thefox didn't have the same problem with his MMC3 mapper, I wonder why.
Turns out I did. Now that I tested with an NTSC NES (which I got about week ago), I noticed WRAM corruption in Super Bat Puncher and M.C. Kids at least. Fixed it the same way you did. Funny that it worked on PAL NES without a hitch.
I summon the attention of the gods...
Just wanted to note that the following games run, but are glitchy:
Guerrilla War - lines flash on the screen, very minor glitches
Kirby's Adventure - Wrong palette
Cybernoid - The Fighting Machine - impossible to play, all the graphics are glitchy garbage
Karnov - Glitchy backgrounds, impossible to play.
Fixing these would reduce the list of incompatible games almost only to MMC5 games. I hope you have time to look into these. Thanks!
Guerilla War and Kirby's Adventure don't sound like PowerPAK problems.
For Kirby for sure has incorrect palette issues, for instance the final boss battle looks incorrect. I thought this was known already though.
All of the games I mention have glitches. Guerrilla War has glitches that are noticeable, but it's still playable. I have the original cart, and it definitely is different in the Powerpak.
Quote:
Guerrilla War - lines flash on the screen, very minor glitches
I didn't notice anything wrong with this (when compared to emulators).
Quote:
Kirby's Adventure - Wrong palette
Also didn't notice anything wrong, but I haven't played the game very much.
Quote:
Cybernoid - The Fighting Machine - impossible to play, all the graphics are glitchy garbage
This is a known problem that could be fixed by emulating bus conflicts in CNROM. Game Genie code SXZNZV can also be used to fix it.
Quote:
Karnov - Glitchy backgrounds, impossible to play.
This I can confirm. A strange MMC3 problem.
Quote:
All of the games I mention have glitches. Guerrilla War has glitches that are noticeable, but it's still playable. I have the original cart, and it definitely is different in the Powerpak.
Would be helpful if you could provide screenshots or something. Also you should tell the exact version of the mappers you're using.
EDIT: FUCK. I accidentally edited my old post out when I meant to make a new post. Oh well.
EDIT2: Was able to stitch it back together from another tab...
Quote:
Quote:
Karnov - Glitchy backgrounds, impossible to play.
This I can confirm. A strange MMC3 problem.
I have fixed this in my mappers. The game was relying on the mirroring register to be set to horizontal mirroring by default (never initialized it).
EDIT: Actually seems like Karnov doesn't actually use MMC3, but mapper 206, functionality of which is a subset of MMC3. That explains why they never write to the mirroring register. Seems to work fine though when the mirroring register is initialized to horizontal, just have to hope there's no other mapper 206 games marked as MMC3 that rely on other initialization values. Maybe a better fix would be to initialize the register to the value from the iNES header...
Quote:
Quote:
Cybernoid - The Fighting Machine - impossible to play, all the graphics are glitchy garbage
This is a known problem that could be fixed by emulating bus conflicts in CNROM. Game Genie code SXZNZV can also be used to fix it.
Fixed as well.
New version of save state mappers:
http://kkfos.aspekt.fi/downloads/save-s ... rs-1.4.zip
Just a head up about Super mario bros + Duck Hunt multicart having sound issue. The music play too fast in Super Mario. Really not a big issue as the regular mario bros rom play fine but tought I would still report it as some other game might suffer from a similar bug. I use loopy's mapper with TheFox save state mapper over it.
SkinnyV wrote:
Just a head up about Super mario bros + Duck Hunt multicart having sound issue. The music play too fast in Super Mario. Really not a big issue as the regular mario bros rom play fine but tought I would still report it as some other game might suffer from a similar bug. I use loopy's mapper with TheFox save state mapper over it.
Sounds normal to me. Maybe it's a PAL vs NTSC issue?
So sorry
I just checked again and turn out that I had the wrong regional version of the rom on my powerpak and didn't notice it as the name was too long to display on the powerpak menu. I transferred the USA version and it is indeed working properly. But I'm not sure if it's known but Bubble Bath Babe and Maxi-15 have their GFX all messed up.
Nintendo's NES-CNROM board supports only four 8 KiB banks of CHR ROM. Bubble Bath Babes uses a CNROM-compatible board that supports more memory. I seem to remember old PowerPak mappers not supporting more than CNROM's 32 KiB for mapper 3, but I'm under the impression that one of the later versions of PowerPak mappers emulates the other board as well. I know BNROM (mapper 34 with CHR RAM) got extended to a full 512 KiB upon my request.
Ok, I made a CNROM mapper with no CHR size limit. This fixes several arousing unlicensed games. It also has the fixes for Paperboy and Cybernoid (open bus issues).
[edit]
Karnov's fixed too. thefox already fixed it in his mapper, I'm just getting caught up to him
Bubble bath babe work perfectly now. Thanks, It will make the 2 time I'll ever play that game in my lifetime more enjoyable
Okay, Karnov, the Panesian games, Batman RotJ and the input issue games work with the latest loopy mappers and BIOS. I believe I am still using some hacked roms for Paperboy and Dirty Harry and other input issue games.
I have a current issue with Low G Man. The latest MAP04.MAP here causes graphical glitches on the status bar and the game crashed at the first boss. It also caused jumpy subscreens and level select screens in Megaman 3, so I tried the ones on this thread :
http://nesdev.com/bbs/viewtopic.php?t=8337
The second one was the only one that did not cause glitches in Low G Man or other MMC3 games, but the boss music for Low G Man sounded "warbly" compared to a real cart.
Great Hierophant wrote:
I have a current issue with Low G Man. The latest MAP04.MAP here causes graphical glitches on the status bar and the game crashed at the first boss.
I've heard that Low G Man requires WRAM enable/disable to work, the crashing in the boss thing might be related to that, but I haven't verified it. WRAM enable flag is currently disabled, I believe, in loopy's mappers (and in mine as well) to get StarTropics 1 and 2 to work. I don't think it's possible to fix this in an easy way without one of these: 1) implementing iNES 2.0 support 2) patching the ROM 3) checking ROM checksums.
EDIT: WRAM protection -> WRAM enable
thefox wrote:
Great Hierophant wrote:
I have a current issue with Low G Man. The latest MAP04.MAP here causes graphical glitches on the status bar and the game crashed at the first boss.
I've heard that Low G Man requires WRAM protection to work, the crashing in the boss thing might be related to that, but I haven't verified it. WRAM protection is currently disabled, I believe, in loopy's mappers (and in mine as well) to get StarTropics 1 and 2 to work. I don't think it's possible to fix this in an easy way without one of these: 1) implementing iNES 2.0 support 2) patching the ROM 3) checking ROM checksums.
Low G Man uses TLROM, it does not have WRAM. I heard it is an issue of emulating an open bus. Of course, for those games that use WRAM or SRAM, a conflict may arise.
Great Hierophant wrote:
Low G Man uses TLROM, it does not have WRAM. I heard it is an issue of emulating an open bus. Of course, for those games that use WRAM or SRAM, a conflict may arise.
Actually I said it wrong, I meant "WRAM enable" when I said "WRAM protection". Yes, Low G Man doesn't use WRAM, but it checks for its presence (or so I've heard). To be more specific, it excepts to read out open bus after disabling the WRAM. This is not true for the current mappers (WRAM is always enabled), and changing that would break StarTropics 1 and 2. The reason for this is that StarTropics 1 and 2 are often marked as MMC3 when in fact they use MMC6.
thefox wrote:
Great Hierophant wrote:
Low G Man uses TLROM, it does not have WRAM. I heard it is an issue of emulating an open bus. Of course, for those games that use WRAM or SRAM, a conflict may arise.
Actually I said it wrong, I meant "WRAM enable" when I said "WRAM protection". Yes, Low G Man doesn't use WRAM, but it checks for its presence (or so I've heard). To be more specific, it excepts to read out open bus after disabling the WRAM. This is not true for the current mappers (WRAM is always enabled), and changing that would break StarTropics 1 and 2. The reason for this is that StarTropics 1 and 2 are often marked as MMC3 when in fact they use MMC6.
I understand all that, it is a limitation of mapper 4, which lumps in virtually all MMC3/6 boards. Any game that uses WRAM, whether battery backed or not, would break if it were not present. This includes HKROM, TKROM and TSROM boards. However, Low-G-Man uses TLROM, which does not have WRAM, so why would it disable the WRAM?
Great Hierophant wrote:
thefox wrote:
Great Hierophant wrote:
Low G Man uses TLROM, it does not have WRAM. I heard it is an issue of emulating an open bus. Of course, for those games that use WRAM or SRAM, a conflict may arise.
Actually I said it wrong, I meant "WRAM enable" when I said "WRAM protection". Yes, Low G Man doesn't use WRAM, but it checks for its presence (or so I've heard). To be more specific, it excepts to read out open bus after disabling the WRAM. This is not true for the current mappers (WRAM is always enabled), and changing that would break StarTropics 1 and 2. The reason for this is that StarTropics 1 and 2 are often marked as MMC3 when in fact they use MMC6.
I understand all that, it is a limitation of mapper 4, which lumps in virtually all MMC3/6 boards. Any game that uses WRAM, whether battery backed or not, would break if it were not present. This includes HKROM, TKROM and TSROM boards. However, Low-G-Man uses TLROM, which does not have WRAM, so why would it disable the WRAM?
Copy protection most likely.
It disables it incase it's actually there I guess. If your board doesn't have it then you get open bus but if you do then you need to disable it to get open bus. Generally when emulating iNES 1.0 and most rom images don't really have a reliable way of identifying if they need WRAM or not unless it's always battery backed. So you usually allow WRAM even if games don't use it since normally it has no negative effect.
Perhaps a workaround would be to use the MAP04.MAP file which did work for the game and change the mapper number in the game's header. What needs to be done for the MAP file? Looking at a hex editor it seems that more than just changing the file name is necessary.
Are you sure? There is a fair chance changing the file name is all that is needed for mapper files.
Make a iNES 2.0 Supmapper for boards without WRAM.
Great Hierophant wrote:
Perhaps a workaround would be to use the MAP04.MAP file which did work for the game and change the mapper number in the game's header. What needs to be done for the MAP file? Looking at a hex editor it seems that more than just changing the file name is necessary.
Yes, you can do that. You don't need to change anything but the filename (and the ROM header of course).
It worked, no more warbly boss music. The MAP04 file is the one included in the latest official powerpak mapper release, (POWERPAK134105.zip) and I changed it to MAP06 and the iNES mapper for this version of Low-G-Man is 06.
Great Hierophant wrote:
crade wrote:
Crystalis seems to have some graphical issues as well.
Yes, I can observe a small amount of tile garbage here and there when walking on the map outside the first town.
this is really confusing me either,wich games are compatible and wich games are not/yes compatible.
sadly retrousb has no clear list with games wich are compatible and game wich are incompatible.
yes they has a mapper list with colors to indicate wich games works good,malfunction,glitchy,corrupting or do,nt work,but the list only show some weird numbers and not the name of the mappers,why not do the same as what they dit whith snes powerpak,putting an incompatilbility list.
even games wich are stated to worh on it,can still not work becouse of damage and/or may require error correction (wich the powerpak lacks)i can,t get ducktales castlevania 1 etc,, too work,whereas i sew on youtube that on someones powerpak his ducktales rom jst work on it.
note;i have the last update,so what,s the point??
I don't observe that anymore with loopy's latest mappers. The only non-Japanese NES games that will not work or cannot be made to work with the PowerPak at present are Action 52 and the MMC5 games:
Bandit Kings of Ancient China
Castlevania III - Dracula's Curse
Gemfire
L'Empereur
Laser Invasion
Nobunaga's Ambition 2
Romance of The Three Kingdoms II
Uncharted Waters
johannesmutlu wrote:
Great Hierophant wrote:
crade wrote:
Crystalis seems to have some graphical issues as well.
Yes, I can observe a small amount of tile garbage here and there when walking on the map outside the first town.
this is really confusing me either,wich games are compatible and wich games are not/yes compatible.
sadly retrousb has no clear list with games wich are compatible and game wich are incompatible.
yes they has a mapper list with colors to indicate wich games works good,malfunction,glitchy,corrupting or do,nt work,but the list only show some weird numbers and not the name of the mappers,why not do the same as what they dit whith snes powerpak,putting an incompatilbility list.
even games wich are stated to worh on it,can still not work becouse of damage and/or may require error correction (wich the powerpak lacks)i can,t get ducktales castlevania 1 etc,, too work,whereas i sew on youtube that on someones powerpak his ducktales rom jst work on it.
note;i have the last update,so what,s the point??
johannesmutlu wrote:
even games wich are stated to worh on it,can still not work becouse of damage and/or may require error correction (wich the powerpak lacks)i can,t get ducktales castlevania 1 etc,, too work,whereas i sew on youtube that on someones powerpak his ducktales rom jst work on it.
Correcting bad ROMs should not be a responsibility for the PowerPak. Although I'm sure the PowerPak is mostly used to play pirate games, RetroUSB sells it as a "development tool", meaning you are supposed to use ROMs you make/dump yourself, instead of using the pirate shit that comes from the internet. The PowerPak has no obligation to accommodate that.
When you make the decision to use the PowerPak for piracy, you should at least have the decency of studying up why some games that are supposed to work don't and how to fix them, instead of publicly complaining that you can't easily steal games.
Quote:
note;i have the last update,so what,s the point??
Ducktales and Castlevania 1 use one of the simplest mappers ever, so it's definitely not the PowerPak's fault. Your ROMs probably have bad headers and/or are corrupted. Just download other copies until you find one that works or use an emulator that can fix headers, such as Nestopia and learn how to do it.
I have loopy's latest mappers. Kirby's Adventure screws up a lot on my powerpak.. I have the PRG0 and PRG1 versions of the ROM, both can't play correctly. Sometimes it's right at the beginning after the intro that it freezes, other times, just before level 1, kirby smacks the ground with his star and the graphics go haywire all over and it freezes, other times i can complete level 1, then it freezes. This is annoying.
RushJet1 wrote:
I have loopy's latest mappers. Kirby's Adventure screws up a lot on my powerpak.. I have the PRG0 and PRG1 versions of the ROM, both can't play correctly. Sometimes it's right at the beginning after the intro that it freezes, other times, just before level 1, kirby smacks the ground with his star and the graphics go haywire all over and it freezes, other times i can complete level 1, then it freezes. This is annoying.
Ah Kirby ... What a marvelous piece of software engineering.
Something you're doing with the ROM is kicking the copy protection in.
(Hint: It's checking if there's anything wrong with the backup memory to figure out if the cart is a pirate copy or if it's running on a copier.)
Nevermind, found
this, loopy's 4th mmc3 mapper seems to work with no issues.
The exact reason for the problem is that Kirby checks the WRAM throughly:
If it's smaller than 8KB
If it can be write protected or not (Kirby is supposed to run on a TKROM board and that one has both CE and /CE pins connected to MMC3)
And a few other sanity checks.
If anything fails, the game runs but background changes oddly, stage layout is corrupted and you can spot "HAL" written on the background of the1st stage at the furthest part you can get to. Also, if you try to do the "inhale" move Kirby get stuck doing that.
Why I know this ? When I was testing my MMC3 dev board I played around with the WRAM hookup and it only works properly when WRAM is hooked "right".
I didn't get any of those symptoms. Kirby could inhale fine. This was more along the lines of an actual freeze: music would freeze / continue doing what it was doing forever, all the graphics suddenly went BUHHHHH .. kind of like if you wiggle a cart around too much and some pins lose contact while playing a game. Again Loopy's mmc3 mapper on that page fixes the issue.
One other hilarious thing I had happen with the other MMC3 is that one time, the door to level 1 was a solid object: I could not run in front of it or go inside. Probably the strangest glitch.
RushJet1 wrote:
I didn't get any of those symptoms. Kirby could inhale fine. This was more along the lines of an actual freeze: music would freeze / continue doing what it was doing forever, all the graphics suddenly went BUHHHHH .. kind of like if you wiggle a cart around too much and some pins lose contact while playing a game. Again Loopy's mmc3 mapper on that page fixes the issue.
Then perhaps it was some glitch with the mapper logic ?
Older versions of the PowerPak MMC3 would corrupt bytes when the scanline IRQ was turned on or off. The writes to $E000-$FFFF were momentarily decoded to $6000-$7FFF because PRG /CE changes slightly later than A13-A14. This was noticeable as map corruption in M.C. Kids and Crystalis.
Maybe his newest mappers haven't been updated. Try crystalis with the old mapper, that had problems with his other older mapper too IIRC.
Having some issues with Metroid FDS v 1.02. Some doors won't spawn, high jump boots didn't show up and the zebetites and mother brain were already dead when I entered the room. I have used the same rom and powerpak before with zero problems. Rev E powerpak on a sound modded NES. I think that saving progress and continuing is causing the problem since the first time I beat it I didn't save until I finished the game.
loopy wrote:
tepples wrote:
Could you elaborate on what kind of conditions would have caused such a glitch in old MAP04 versions, so that I can trigger these conditions in my test ROM to make sure this doesn't happen again?
PRG /CE (which is used by the mapper to get A15) changes slightly after M2. Writes to E000-FFFF would briefly assert WRAM /WE, causing ram corruption.
How how to solve the problem?
Is there a powerpak mapper sourcecode or code section?