Hello - I'm not sure if this is the correct part of this forum to post this question, but here we go.
A few weeks back I got a Famicom Disk System. It works almost perfectly, but - in some games worse than others - the sprites have garbled, glitchy areas. It's worst in Mario 2, where half of Mario/Luigi is garbled.
I've discussed this issue at length over at Famicom World (
here is the thread for the full background), and it was recommended that I try asking about the issue here since it seems like it might be a microchip issue and this is a more technical forum.
I'd appreciate any advice. Please keep in mind that I can be a bit of a tinkerer but I'm no engineer or programmer - I'm kind of a fish out of water here.
Most things I can think of that would cause obvious corruption in one place (graphic data) would cause corruption in many (crashes)...
The guesses I have:
- Something's wrong with your disks
- All the games you have just happen to store the graphic data in the same area on disk, and the drive head is tracking poorly causing incorrect data when reading graphics but not program data
- Something's wrong with the FDS adapter's graphics RAM
- Something's wrong with the FDS adapter's BIOS
Unfortunately, without the ability to run arbitrary code on it, I have no idea how to go about debugging which of these things could be wrong. I can't really tell what's going wrong from the picture on your thread at FW... could you provide a picture of Mario (or anything else corrupted) against a solid background?
Are the sprites garbled when you switch on the FDS with no disk (the BIOS intro with Mario and Luigi jumping) ?
Does normal Famicom carts works fine ?
lidnariq wrote:
Unfortunately, without the ability to run arbitrary code on it, I have no idea how to go about debugging which of these things could be wrong. I can't really tell what's going wrong from the picture on your thread at FW... could you provide a picture of Mario (or anything else corrupted) against a solid background?
Affirmative. I'll take some more pics of SMB2 as well as the other disks when I get home tonight.
Bregalad wrote:
Are the sprites garbled when you switch on the FDS with no disk (the BIOS intro with Mario and Luigi jumping) ?
Does normal Famicom carts works fine ?
The sprites are not garbled on the BIOS screen; normal Famicom games work perfectly.
Then this implies the only problem can be your disk drive or the cable that connects to it. The RAM adaptor can only work perfectly if the intro work perfectly.
However, there is CRC checks on disk loads, and any bad loads should result in a disk error instead of corrupted data being loaded. However, the RAM adapter's character RAM is fine, and I don't see how it's ROM could not be fine considering ROM is one of the most solid component ever by it's "read only" nature.
This is really a strange problem indeed.
The problem could be with your RAM adapter. The BIOS doesn't use all of the Character RAM. Maybe there is a problem with an address line on the chip. It could be that a solder joint went bad or the chip somehow is damaged or not working now. I would try another RAM adapter if that is a reasonable option for you.
MottZilla wrote:
The problem could be with your RAM adapter. The BIOS doesn't use all of the Character RAM. Maybe there is a problem with an address line on the chip. It could be that a solder joint went bad or the chip somehow is damaged or not working now. I would try another RAM adapter if that is a reasonable option for you.
I am considering buying a second RAM adapter to see if that's the problem. I wanted to see if I could rule out other possibilities before actually spending any money, though.
Anyway, here are some more pictures.
The BIOS screen, free of garbling:
Moero TwinBee, which shows the worst garbling of all of my disk games. You can see it in the clouds, the score count, and TwinBee's shots.
It's just barely visible on Pidgit here in Doki Doki Panic. This game has the least issue with garbling.
Here are a couple more shots from SMB2, hopefully it's clearer in these ones. You can see it on the Koopa shell too on the second pic.
It's pretty bad on Baseball, which is on side B of SMB2. You can kinda tell from this pic.
Metroid has it in a few spots. Sometimes Samus' cannon is garbled or part of her sprite disappears, and sometimes little graphical elements pop up around her while she's moving. Not as bad as Mario or TwinBee though.
And finally, here's a pic of the innards of my RAM adapter, just for funsies.
Zycrow, do you live in the United States? If so, I'd be willing to send you (temporarily) my FDS RAM adapter (which I know works, as does my entire FDS for that matter) to see if replacing that fixes your problem. You would need to send it back to me (on your dime) after confirming/rejecting. Let me know.
So, here's the thing that makes this Really Weird:
In SMB (LL), sometimes the right lower half of his body is made of a reflection of the left lower half. Same exact tile. Coming from the exact same place in memory. Which is what seems to on screen in both the original photo you posted and at least the latter of the two you have here. And yet, Mario's right half is distorted, and his left half isn't.
In Moero TwinBee, I haven't even the foggiest idea where that "half bubble" image is coming from: I don't see it in the pattern tables. It's supposed to be a simple 5x5 sprite (padded into an 8x16 cell, hence "half" bubble).
Buh. I have really no idea what could be causing this ... it's definitely not simply the adapter's CHR RAM, if that's even to blame at all. I don't suppose you might have access to another famicom to test on, also ?
lidnariq wrote:
In SMB (LL), sometimes the right lower half of his body is made of a reflection of the left lower half. Same exact tile. Coming from the exact same place in memory.
SMB and SMB2 (J) do this to save 3 tiles on the sprite side of the pattern table, 1 for small Mario and 2 for Super Mario, even though it twists Mario's legs into a painful
turnout. This was fixed in Doki Doki Panic and SMB3, and the versions of SMB and SMB2 (J) in Super Mario All-Stars use the SMB3 graphics for Mario.
Since that many games has graphical problems I wouldn't suspect the disks. I suspect the problem lies in either the Famicom or the RAM Adapter (or in the drive head as Lidnariq said). As can be seen in that Famicom World thread, reports of faulty CHR RAM chips in RAM Adapters has appeared before, so I wouldn't rule out that either.
As was said before, it would be best to first try a working RAM Adapter and/or try your RAM Adapter on another Famicom or NES.
Well, at the very least I'm learning some things about how the FDS ticks!
Seriously though, thanks for everyone's input so far.
lidnariq wrote:
Buh. I have really no idea what could be causing this ... it's definitely not simply the adapter's CHR RAM, if that's even to blame at all. I don't suppose you might have access to another famicom to test on, also ?
Unfortunately I don't. I do have another Famicom, but it's definitely not working (pretty sure I fried it by using a NES AC adapter many years ago).
koitsu wrote:
Zycrow, do you live in the United States? If so, I'd be willing to send you (temporarily) my FDS RAM adapter (which I know works, as does my entire FDS for that matter) to see if replacing that fixes your problem. You would need to send it back to me (on your dime) after confirming/rejecting. Let me know.
I do live in the states, and I'm happy to take you up on this offer, so long as you're willing to risk sending your hardware to a stranger!
Zycrow wrote:
koitsu wrote:
Zycrow, do you live in the United States? If so, I'd be willing to send you (temporarily) my FDS RAM adapter (which I know works, as does my entire FDS for that matter) to see if replacing that fixes your problem. You would need to send it back to me (on your dime) after confirming/rejecting. Let me know.
I do live in the states, and I'm happy to take you up on this offer, so long as you're willing to risk sending your hardware to a stranger! :mrgreen:
I do as long as I have your full name, address, and a telephone number for contact in case anything odd comes up, I'm fine with it. I tend to operate on the notion that most people here on nesdev are good-natured and not malevolent.
There's the inverse option too: you could send me your RAM adapter and I could try it out on my FDS + Famicom (both of which are confirmed working). I don't have the same FDS games you do, but I do have Metroid. My FDS collection is limited to Kid Icarus, Metroid, SMB2 (Japanese version), and Falsion. As for trustworthiness, I think pretty much every person here could vouch for me on that.
Either option is fine by me.
The added bonus to the latter choice is that if it does turn out to be your RAM adapter, I might be able to figure out what part is broken in it (maybe some bad RAM inside of it, for example) and try to repair it for you for free (including cost of parts; yeah, I'm nice like that). I'm good with a soldering pencil and "so-so" with a multimetre, and do have access to an oscilloscope if needed, but I'm not good with EE in general, so would probably need help from the amazing number of EE-savvy folks here. This is all optional of course, not mandatory in the least.
Think about it and let me know -- as I said, either option is fine with me.
koitsu wrote:
Think about it and let me know -- as I said, either option is fine with me.
Cool! I'll send you a PM and I'll update the thread if more is figured out on this issue. Again, I appreciate everyone's help.
So I got Zycrow's FDS RAM adapter today...
The issue does not appear to be with the RAM adapter. I tested multiple games (specifically Kid Icarus, Metroid, SMB2, Falsion, and Twinbee) and I see absolutely no graphical corruption of any kind. I tried with two different Famicoms as well (Famicom AV and original Famicom). I also tried gently wiggling the FDS RAM adapter cable while disk I/O was going (thinking maybe bad connection in cable might cause data corruption).
If cartridges don't have this problem, then that would mean the issue is either with the Famicom itself (are there edge connector pins which the FDS uses that cartridges don't? I'm unsure), or with the FDS itself. The latter has my vote, but I suppose it could be the former as well.
I'm betting there is some particular chip inside of his FDS which is busted -- I think we can (mostly) rule out the disk drive itself since CRC errors would get detected, but I am certain there is other circuitry that is used ""past"" that point, i.e. before or during the data is loaded into the RAM adapter, which could be corrupting something.
Random question for Zycrow: are you powering your FDS with an AC adapter or C batteries? I power mine with an official AC adapter (cost me an arm and a leg).
The best I can do in the meantime is:
1. Clean all the contacts on his FDS RAM adapter (no harm in doing that, unless already done),
2. Return his FDS RAM adapter to him,
3. Also send him my FDS + FDS RAM adapter so he can use that to troubleshoot which exact piece of hardware is busted, then send me my stuff back once he's narrowed it down.
Repairing FDS systems is not something I'm good at. Yes I've opened one up to try and repair a drive belt (I failed), but I always hate working on those things given all the sensitive parts and how timing-oriented the dang thing is; the last bad FDS I had (bad belt / bad motor / timing / something, even after FDS belt replacement) I gave to Matt Conte because I couldn't be bothered to try and fix it.
If other folks here have ideas of things to try, I'd be happy to give them a shot.
Well, that answers some of our questions!
To answer yours, I have been powering my FDS with batteries. I believe I have an AC Adaptor that would be usable, based on Famicom World's guide, but I haven't tried it as of yet.
The FDS AC adapter I have is the official one, with the following information:
Input: AC 100V, 50/60Hz, 9VA
Output: DC 9V 400mA
Polarity:
Negative polarity, i.e. positive outer, negative inner
I use this on a standard US AC outlet without any problem (i.e. no 120V-->100V step-down transformer).
Given my observation about how only some of the reflected sprites in SMB seem to be damaged ... I'm not convinced it's the disk drive.
But if it's not been tickled using ordinary cartridges also... I don't know? Could it be voltage supply sag?
Is it the same sprites every time, even between full system power-cycles?
It is the same sprites every time, yes.
For those following thread: plan of action is to send my working FDS, FDS RAM adapter, FDS AC adapter, and an FDS game (Kid Icarus) to Zycrow so he can mix-and-match working pieces with his own system in attempt to find out which piece is malfunctioning.
If the only thing that's ever corrupted are sprites (which seems to be the case), that would seem to point to an issue with the PPU (possibly some bad logic when reading sprite tile data?). However, the fact that the system works with cart games would seem to rule that out, as I'm not aware that SMB and SMB2 have too many differences in their handling of OAM that would manifest here. I would suggest checking the RAM adapter for a bad address line or similar, but koitsu has already made that moot.
Not sure where the tiles are actually stored in CHR RAM for SMB2, but if it helps in deconding the relevant addresses, the incorrect tile in place of Mario's foot looks like the Mushroom Retainers' foot tile (which I'm pretty sure is also mirrored for both feet from the same tile). Perhaps it may be prudent to specifically check cart games that use CHR RAM? Games that use CHR ROM (like many early games) probably wouldn't exhibit the same problem in such a case.
I haven't received Koitsu's FDS yet, but when I do, I'll post a report of what configurations I try and how it all looks.
OK, you guys ready for some weirdness? I got Koitsu's hardware in the mail yesterday and this morning I did some tests.
Here are the results:
- My FDS with my RAM adapter (again): Garbling
- Koitsu's RAM adapter with my FDS: No garbling (tried all games)
- Koitsu's RAM adapter with his FDS: No garbling
- Koitsu's FDS with my RAM adapter: Garbling
I haven't done any tests regarding power supplies, but I think the results show something pretty clearly here. The problem lies with my Famicom unit and my RAM adapter, since my RAM worked perfectly on Koitsu's system. Crazy!
My suggestion is, if you have any of the games on
this list (which contains all Famicom carts with CHR RAM) is to test one of those games and see if you also have corruption. The only thing to throw a wrench into that line of reasoning is that your FC worked with koitsu's RAM adapter but not your own. That should rule out the FC itself, except for the fact that your RAM adapter worked on koitsu's FCs but not your own.
Maybe it's an issue of dodgy pins on both your FC and your RAM adapter, but the pins aren't dodgy enough to fail when used with other hardware, but when used together you have a pin not making a connection at all. Sort of like how the Guitar Hero 3 guitars with removable necks inevitably get dodgy connector pins, but swapping necks can cause different behavior from the pins being worn differently.
What if the timing on the RAM in the adapter is really borderline, perhaps due to slight undervoltage, and the extra time taken in the PPU to flip a sprite is enough to get from "data arrived in time" to "didn't" ?
But does it really take extra time in the PPU to flip a sprite horizontally? I thought it'd be just a mux between the flipped version of the sliver and the not-flipped version, selected by the H bit of the sprite's attribute 2. Vertical flip is just an XOR gate between the line-within-sprite and the V bit.
Dang it -- this makes me wish I had sent you my Famicom AV (and AC adapter) as well. I thought about doing that just as a "worst-case scenario".
It really sounds like there's something physical going on with the combination of your FDS RAM adapter and your Famicom. It's probably something pin/edge-connector-related like what LocalH mentioned.
Also, did you try your FDS + my FDS AC adapter + your FDS RAM adapter? Part of me ponders voltage, since you're normally driving your FDS off of C batteries.
Another test would involve $$, and that'd be getting yourself another Famicom to see if the problem goes away.
This is where I wish all the hardware that's malfunctioning could be shipped to someone like, say, lidnariq or blargg to do all sorts of EE-related tests on it (dudes with oscilloscopes tend to be able to track strange things down, but I don't have the knowledge -- just a simple caveman... :-) )
So, if you exchange your RAM adapter with Koitsu's, you'll both end up with fully working FDSes, no ? Although it doesn't answer to why a particular combination doesn't work, honestly this is really crazy.
Getting a new Famicom or FDS is expensive, but plain RAM Adapters are very common and cheap on EBay (or at least were last time I checked), so it shouldn't be too hard to have a solid stock of a few adapters if there is a need to.
@koitsu : I honestly doubt this kind of problem would be "easy" to trace with an oscilloscope. You'd rather need a logic analyser and look at the timings on the CHR bus, but that equipment is much more expensive and complex to set up than a simple 2-trace analog oscilloscope.
@Bregalad -- ahh okay. (See, EE isn't my thing... heh :-) )
Yeah, I suppose a worthwhile thing to do would be to buy an FDS RAM adapter off eBay or somewhere and try it out (using Zycrow's Famicom + FDS).
Part of me thinks it's got something to do with the edge connector contacts between the FDS RAM adapter and the Famicom, which could explain why Zycrow's FDS RAM adapter worked just fine on my Famicom AV. I should have tried wiggling it or applying more pressure to it while it was inserted. Zycrow, you might try that -- just try shifting around or applying a bit of pressure to your FDS RAM adapter while attached to your system (doesn't matter whose FDS from the sound of it).
If a replacement FDS RAM adapter fixes the problem, then it might be possible to repair the questionable one by replacing the cartridge connector in it.
For an unrelated yet equally bizarre-as-hell issue with someone's FDS RAM adapter,
check out this thread at Famicomworld. It starts getting interesting (and into specific models of chips in the FDS RAM adapter that are glitchy/buggy in some way?) at the 2nd page. Ignore the crap on page 3 onward -- just adolescents fighting over nonsense -- although there is an interesting tidbit on page 5 (talking about FDS RAM adapters with two different kinds of CHR RAM chips I think).
P.S. -- Zycrow, there was something I noticed about your FDS RAM adapter -- one of the screws in one of the corners looked like it wasn't screwed in all the way. I didn't touch this/change it at all, but if you didn't open it then someone in the past may have, indicating someone may have tried to repair or fix something previously. Totally speculative, but thought I'd mention it.
LocalH wrote:
My suggestion is, if you have any of the games on
this list (which contains all Famicom carts with CHR RAM) is to test one of those games and see if you also have corruption. The only thing to throw a wrench into that line of reasoning is that your FC worked with koitsu's RAM adapter but not your own. That should rule out the FC itself, except for the fact that your RAM adapter worked on koitsu's FCs but not your own.
Maybe it's an issue of dodgy pins on both your FC and your RAM adapter, but the pins aren't dodgy enough to fail when used with other hardware, but when used together you have a pin not making a connection at all. Sort of like how the Guitar Hero 3 guitars with removable necks inevitably get dodgy connector pins, but swapping necks can cause different behavior from the pins being worn differently.
I took a cursory look over that list and I have at least one game on it (Rainbow Islands) and haven't experienced corruption with it. At the moment I don't think I have any untested games for my FC, and I haven't experienced anything remotely like what I have on the FDS.
koitsu wrote:
Also, did you try your FDS + my FDS AC adapter + your FDS RAM adapter? Part of me ponders voltage, since you're normally driving your FDS off of C batteries.
Another test would involve $$, and that'd be getting yourself another Famicom to see if the problem goes away.
I will do some power-related tests tomorrow and see what I get.
I really don't want to have to buy another Famicom, especially since I'd need an AV or an AV modded one to reliably get it to work on the TV I use.
koitsu wrote:
Part of me thinks it's got something to do with the edge connector contacts between the FDS RAM adapter and the Famicom, which could explain why Zycrow's FDS RAM adapter worked just fine on my Famicom AV. I should have tried wiggling it or applying more pressure to it while it was inserted. Zycrow, you might try that -- just try shifting around or applying a bit of pressure to your FDS RAM adapter while attached to your system (doesn't matter whose FDS from the sound of it).
Since the issue reminded me of some of the graphical garbling I used to get on my NES when I was playing with a poor pin connection, I did try this a bit before I got on the forums, and it didn't seem to make a difference. Will try it again, though.
koitsu wrote:
P.S. -- Zycrow, there was something I noticed about your FDS RAM adapter -- one of the screws in one of the corners looked like it wasn't screwed in all the way. I didn't touch this/change it at all, but if you didn't open it then someone in the past may have, indicating someone may have tried to repair or fix something previously. Totally speculative, but thought I'd mention it.
This was probably just from when I opened it up earlier to take a look at the chips inside. The garbling was the same before and after I did this (and I probably just failed to screw that last screw in all the way, yeah.)
Bregalad wrote:
So, if you exchange your RAM adapter with Koitsu's, you'll both end up with fully working FDSes, no ? Although it doesn't answer to why a particular combination doesn't work, honestly this is really crazy.
Getting a new Famicom or FDS is expensive, but plain RAM Adapters are very common and cheap on EBay (or at least were last time I checked), so it shouldn't be too hard to have a solid stock of a few adapters if there is a need to.
Exchanging our RAM adapters is my preferred course of action. I'd be willing to trade back if mine starting malfunctioning with Koitsu's system, even. But that's entirely up to Koitsu. I'm just really happy that you guys have been willing to try to help me out on this at all.
Please try the power-related tests (AC adapter vs. C batteries).
There's something that just doesn't sit well with me about this situation in general -- it's probably the OCD part of me wanting an actual explanation for what is going on, rather than just swapping FDS RAM adapters and saying all is well. I worry that if we swap, fast forward a few weeks/months (or maybe even a year), and suddenly "something odd starts happening", I'm going to be kinda ticked off. I only say this because I think I've spent something like US$500+ on Famicom and FDS stuff to get a working/reliable setup and I'm a bit reluctant to change what I know works when we don't have a completely definitive technical explanation for what's going on. I'd really have expected your FDS RAM adapter to malfunction on my stuff... :/
I have an alternate/different solution that will probably work (if not we're in no different shape than now), but it will take me some time to get it to you (roughly 2-3 weeks). You can use my stuff in the meantime; no rush to ship anything back yet.
tepples wrote:
But does it really take extra time in the PPU to flip a sprite horizontally? I thought it'd be just a mux between the flipped version of the sliver and the not-flipped version, selected by the H bit of the sprite's attribute 2. Vertical flip is just an XOR gate between the line-within-sprite and the V bit.
Sorry for the wild tangent, but I felt this overwhelming need to find where this was in Visual2C02.
Nodes
1860 and
1842 are the gating signals that select horizontal mirroring bits or not respectively, and they allow conduction from e.g.
_db0 or
_db7 to later processing logic (and ultimately
spr0_l0 and
spr0_h0). They're the two halves of an SR latch, gated on
/spr_loadFlag NOR
pclk1, and ultimately fed from
spr_d6.
So, yeah, that random hypothesis neither made sense (in hindsight) nor has anything to do with the actual instantiation.
koitsu wrote:
Please try the power-related tests (AC adapter vs. C batteries).
There's something that just doesn't sit well with me about this situation in general -- it's probably the OCD part of me wanting an actual explanation for what is going on, rather than just swapping FDS RAM adapters and saying all is well. I worry that if we swap, fast forward a few weeks/months (or maybe even a year), and suddenly "something odd starts happening", I'm going to be kinda ticked off. I only say this because I think I've spent something like US$500+ on Famicom and FDS stuff to get a working/reliable setup and I'm a bit reluctant to change what I know works when we don't have a completely definitive technical explanation for what's going on. I'd really have expected your FDS RAM adapter to malfunction on my stuff... :/
I have an alternate/different solution that will probably work (if not we're in no different shape than now), but it will take me some time to get it to you (roughly 2-3 weeks). You can use my stuff in the meantime; no rush to ship anything back yet.
Perfectly fine and understandable. Meanwhile I'll try some of those other tests.
OK! I tried a couple other things, but the results weren't terribly surprising.
First, I used my FDS and RAM adapter and powered it on with Koitsu's AC adapter. The results were garbling, as usual. I then used this same configuration switching to Koitsu's RAM adapter, and the garbling was gone.
I also moved my RAM adapter slightly in and out of the Famicom while a game was on. This would usually cause the game to freeze or the whole screen to go into a sort of plaid glitch-out, but I didn't see a visible effect on the sprites. (I performed this test with Super Mario Bros. 2, where the garbling is easiest to see from the get-go.)
If you guys would like for me to try some other tests, I'm happy to, and Koitsu, just tell me when you'd like me to ship your stuff back to you and I'll do so. Once again I'm super appreciative of all of your help.
Bummer, was hoping as a last resort that the AC adapter might rectify something. Oh well, at least we've ruled some bits/pieces out.
I'll be sending you something (it's a surprise) in a couple weeks to see if that solves your dilemma. If so (crossing my fingers!), keep it (free of charge -- let's just say knowing people who live in Japan comes in quite handy). After that, yes, please send me back my FDS + FDS RAM adapter + AC adapter + box + Kid Icarus disk. :-)
Have you tried holding START and SELECT on controller I while powering on with the RAM adapter to start the RAM test? I doubt anything will come out of it but it's worth a shot.
koitsu wrote:
@Bregalad -- ahh okay. (See, EE isn't my thing... heh
)
For an unrelated yet equally bizarre-as-hell issue with someone's FDS RAM adapter,
check out this thread at Famicomworld. It starts getting interesting (and into specific models of chips in the FDS RAM adapter that are glitchy/buggy in some way?) at the 2nd page. Ignore the crap on page 3 onward -- just adolescents fighting over nonsense -- although there is an interesting tidbit on page 5 (talking about FDS RAM adapters with two different kinds of CHR RAM chips I think).
I remember that thread. It was a heated discussion and it looks like some of the posts have been deleted since then. Anyway that incident is what made people to start suspecting CHR RAM chips to be faulty and is why I thought this was something similar. It has been researched in this thread
http://www.famicomworld.com/forum/index ... pic=8512.0 but in the end no concrete results have yet to appear.
Pokun wrote:
Have you tried holding START and SELECT on controller I while powering on with the RAM adapter to start the RAM test? I doubt anything will come out of it but it's worth a shot.
I had no idea there was such a self-test. Hidden features...! :P
There's also a hidden message on newer BIOS revisions. After starting up with the self-test, hold RIGHT and A buttons and Takao Sawano's hidden message appears. Doesn't work on mine because I have a RAM adapter with the older BIOS.
Pokun wrote:
newer BIOS revisions
Plural? I only knew of "original" and "Twin Famicom"...
I don't know how many versions of the BIOS there is, but counting the Twin Famicom there are at least 3 versions.
I have the older 7201 FDS drive and the included RAM Adapter also seems to have the older BIOS as that hidden message doesn't appear, while on my friend's RAM Adapter (he has the newer 3206 drive) it does appear. I've tried the trick on both RAM Adapters myself.
It should work on the Twin Famicom BIOS version as well, I think the hidden message says what BIOS build it is.
Considering there is no jump table for BIOS calls, the room for revising the BIOS is extremely thin. How does both versions differ (if those 2 versions are actually confirmed to exist) ?
I should definitely try to dump mine with TapeDump.
The only difference I know of is that the hidden message doesn't exist on the older version(s). Maybe I should try dumping my RAM Adapter since the newer BIOS is already dumped. I have no experience in dumping though.
OK! It's time for a new chapter in Zycrow's Weird FDS Problems.
Today I received in the mail another package from the endlessly generous Koitsu, and this time it was a third RAM adapter (we will call it subject "C").
The bad news is that subject C did not solve our problems. In fact, to make things even stranger, subject C is actually less reliable than my original adapter (we'll call it subject A).
So far, I've tried all of my games with subject C and my own FDS. I may try some other combinations before I ship all of Koitsu's stuff back to him. The results:
Super Mario Bros. 2: Game freezes on the title screen; sometimes the screen is blank or completely glitched out. Mario's sprite is garbled like with subject A, but not quite as much.
Moero Twin Bee: Garbling on the sprites during the intro sequence; game freezes afterward.
Doki Doki Panic: Game freezes immediately.
Metroid: Game freezes after title screen.
Palutena no Kagami (my copy): This was the only game that actually loaded all the way. Pit's sprite was garbled, but not as badly as on subject A, and there was some other graphical glitchiness here and there.
I have not tried out Koitsu's copy of Palutena no Kagami yet.
So, wow! What a weird situation. I think what I'm gonna have to do is just hunt down RAM adapters until I get one that works.
EDIT: I should note that before and during my testing, I did several rounds of cleaning on subject C's connectors, as well as my Famicom's connectors, just for good measure. This appeared to have a small effect on SMB2, though I was never able to get past the title screen there.
The fact there's now two suspect FDS RAM adapters involved (one of which worked fine on my own setup here) seems to imply the issue may in fact be with the Famicom, possibly the cartridge/edge connector not making full/good contact with all FDS RAM adapters.
I can't confirm or deny that the FDS RAM adapter we call "C" works, because my FDS setup is with you. :-)
So if there was something to try replacing, I would suggest starting with a Famicom. I know your Famicom is modded or something to work on your setup (video-related? Maybe RGB mod? Or component mod?), but it's starting to sound to me like you just have a crappy cartridge slot on your Famicom that works "mostly well" with carts but possibly used FDS RAM adapters have a slightly different edge connector thickness or something and it's just enough to make your Famicom wig out. I've seen this myself on my old (now-given-away) original Famicom whose cartridge connector was quite worn -- slightly touching a cart would result in graphical garbling. It's the best guess I have at this point, and it's all speculative on my part.
This is partially why I wish I had sent you my Famicom AV as well, but to ship all that in one box would have been quite a lot (more than I already paid). Otherwise I do not have a spare Famicom or Famicom AV to send you. :(
I also don't mod anything I own (nor do I buy modded stuff) solely because I don't trust the modifications. For all I know one of those bazillion mods causes some kind of weird voltage or electrical problem that manifests itself in rare edge cases. I'm not an EE guy so again all this is speculative. This is one of those things where someone familiar with all that needs to be present physically in person to actually troubleshoot what is going on, otherwise all we're gonna end up doing is just spending money on products + shipping until "we find a combo that works" (when aside from clone incompatibility, all this stuff should "just work").
EDIT: I should note *prior* to sending you FDS RAM adapter "C", I cleaned the edge connectors using the same methodology I've cleaned all my NES and Famicom cartridges with (glass top cleaner followed by post-clean-up using ~80% isopropyl alcohol).
Yeah, my Famicom is modified to use component cables - it's one of the mods that appears professional (there are cable ports on the side of the system rather than sticking out of the controller wiring, for example). I've historically had a very difficult time getting Famicoms to work with an RF switch, and I vastly prefer the original Famicom design, which was why I went for the modified one rather than the official AV model.
Possibly at some point I'll buy a second Famicom, but given that I've bought a Famicom, a disk system, and a lot of games over the past several months, I may wait a bit before deciding my next move.
Meanwhile, let me know if you guys want me to try anything more within the next couple days before I send Koitsu's hardware back to him.
I've had absolutely nothing but success with multiple Famicoms working on US TVs using RF adapters, specifically the ones that were included with the NES (top or front loader) or the SNES (they're gray boxes with no "game/TV" switch on them --
here is a picture of one). I forget which channel they worked on, but it was either 3 or 4.
I talk about that
in this thread. In there I go over what television set I use, and I also went to the extreme of trying the same setup out on my neighbour's LCD TV (which is an actual Japanese TV but using a NES RF adapter -- it was found on either channel 10 or C31, but that's because the TV is actually Japanese).
There's speculation in that thread that the Famicoms I had "were modded" but they were not to my knowledge (the one I tested anyway, but I've done this before on other Famicoms) and I present evidence of it on page 3 last post. Once I got a Famicom AV I gave the non-AV models away, but like I said I had absolutely no issues. But apparently other people have had issues (see thread), so I really don't know what to say. It's always "just worked" for me.
I don't think I have a spare RF adapter I could give you, but I could send you one if you want to try it (I have one for my top-loader NES and one for my SNES, but I forget if I have a 3rd for my front-loader NES or not -- either way they're all interchangeable).
Pokun comments took everyone by surprise here!
I didn't publish what I know before just because for some reason I expected that such info was already known by most users here. I was obviously wrong.
Pokun wrote:
Maybe I should try dumping my RAM Adapter since the newer BIOS is already dumped.
Correct. The version tagged as 'original' here is actually the newest known revision. The Twin Famicom one is the same but with changed graphics
*, everything else (including the hidden message!) remains untouched. BTW, I think we should dump our RAM adapters anyway
Pokun wrote:
After starting up with the self-test, hold RIGHT and A buttons and Takao Sawano's hidden message appears. Doesn't work on mine because I have a RAM adapter with the older BIOS.
This must be done before any text appears, otherwise it's too late.
Spoiler, yay!I still hold a dump of the older BIOS, but it was (very) badly copied so it's useless...* Edit.
Ah thank you for the correction! It was a long time ago I tried this on my friend's RAM Adapter. And thank you for the clarification of the different BIOS versions. I guess Sharp didn't know about the secret message when they rebranded the BIOS and thus the copyright notice still says "Nintendo". I wonder if it can be dumped with Kazzo or TapeDump.
koitsu wrote:
I've had absolutely nothing but success with multiple Famicoms working on US TVs using RF adapters,
I talk about that
in this thread. In there I go over what television set I use, and I also went to the extreme of trying the same setup out on my neighbour's LCD TV (which is an actual Japanese TV but using a NES RF adapter -- it was found on either channel 10 or C31, but that's because the TV is actually Japanese).
There's speculation in that thread that the Famicoms I had "were modded" but they were not to my knowledge (the one I tested anyway, but I've done this before on other Famicoms) and I present evidence of it on page 3 last post. Once I got a Famicom AV I gave the non-AV models away, but like I said I had absolutely no issues. But apparently other people have had issues (see thread), so I really don't know what to say. It's always "just worked" for me.
It's known that using RF with Japanese consoles (including the red & white Famicom) doesn't work on some American TV sets and most European TV sets.
It might be easiest to get a picture if you set an American TV to "analog cable" and look in channels 95-99.
Here's an update!
So I bought a standard white/red Famicom with no mods over on Famicom World. I initially couldn't get it to work, but I had a second CRT lying around and was able to boot it up using channel 95. This Famicom worked perfectly with my original FDS RAM adapter. I haven't tried out the other one that Koitsu sent me, but after playing 30 minutes or so of SMB2 with no graphical errors whatsoever, I think I feel comfortable in ruling this as an issue with my modded FC.
...which is a shame, because the picture quality from that machine is much better and I greatly prefer the TV setup I have in that room. But that's entirely my own problem.
There may be one more chapter to this story, as I was able to pick up a junker FDS with a working RAM adapter, so once that comes in I'll give it another test with my modded FC and see what happens.
UPDATE: Third time's the charm - that RAM adapter works with my modded FC, so I didn't have to make any drastic changes to my setup.