Help out development immensely

[Nach]

Currently, the amount of people doing hardware tests to verify what we need to emulate is basically limited solely to byuu.

Unfortunately, the only copier byuu has is a UFO 8.3j, which can't do all the tests he needs to do.

byuu really could use a Wildcat DX 2 or a Game Doctor SF 3, or possibly both of them.

Is anyone able to help get these for byuu?

Finding a place where they could be bought, which isn't a scam would be enough. If someone owns one but only uses it for playing games and would like to donate it to byuu, that'd work too.

Once byuu gets one, it should help out both bsnes and ZSNES.

Thanks in advance.

[byuu]

To clarify, I would need this for special chip testing, most notably the SPC7110, S-DD1 and S-RTC. The rest I can do on my UFO now.

I am willing to pay for the hardware, but I will only pay someone I trust to actually deliver it in fully working condition, and I don't want anyone sending money to someone they don't trust. I also do not want a copier from anyone working in SNES emulation -- I don't wish to inhibit someone else's ability to do good things for the scene.

It really just needs to be a copier that can pass through read requests from $[00-3f|80-bf]:[2000-5fff] to the cartridge. My UFO just returns 0x00 every time. It also has to turn on when you insert special chip games like Star Ocean, FEoEZ, etc. My UFO just displays a black screen with those games connected to it. I know the GDSF7 does the same for FEoEZ.

Right now, I do have a SNES unit with a disabled CIC chip, so in theory I could create a special program that uploads to RAM and waits for a keypress before activating the test. From here, I could add a copier header, transfer this to a floppy, run it on the UFO, then rip the UFO out with the power on, jam a special chip cart into the running system, and then run the test. And I would be forced to read any results on-screen, rather than storing large pools of data into save RAM for PC analysis.

Both apathy to the much more involved process, as well as fear of damage to my only copier prevent me from trying this.

[powerspike]

I have a friend living in Hong Kong. If someone could tell me the name of a store that sells them used he could get one. I'd need the address in English though since he's illiterate in Chinese. A high resolution picture of the copier would help a lot too. As far as money goes it probably wont cost that much so you wouldn't need to pay anything byuu.

Edit: And I'm pretty sure it'll end up being a GD SF3.

[zidanax]

I know the GDSF7 does the same for FEoEZ.

So I suppose the SF7 is probably no good, then? Oh well... that's the only copier tototek.com has left in stock. Robert Webb used to have Super Wildcard DX2's on his site, but I don't see them anymore. Although he did charge an arm and a leg for them anyway, if I recall correctly.

EDIT: Apparently, this person had a Game Doctor SF3 for sale a few months ago. Maybe you could PM him and ask if he still has it?

EDIT: this thread(registration required) mentions a ton of copiers for sale, including an SF3. The seller said "I need to sell everything
by this Thursday 7/17/08", but the seller bumped the thread just today, which indicates to me that the stuff is probably still up for grabs.

EDIT: didn't notice until just now that those two posts are from the same username.

[neviksti]

I know the GDSF7 does the same for FEoEZ.

So I suppose the SF7 is probably no good, then?

For romhacking and testing I strongly recommend the SF3 or SF7 (it even lets you hack the BIOS of things like the XBand or SuperGameBoy due to its great pass-through capability). The only downfall of the SF7 is the slightly extra capabilities also mean slightly more delay of signals. For the cartridges that barely make it (DSP or SuperGameboy for example) this slight extra delay is a bit too much.

For that reason, I suggest the SF3 (not as good with memory mapping, but better for testing). The only potential problem I know of the SF3 is that it inverts the clock line. This doesn't seem to bother the games I tested. As usually, the clock line is kind of finicky anyway... the whole reason it is inverted is the copier is trying to boost it to help out games. Caitsith cleaned his machine very well and doesn't appear to have any trouble, so any connection finicky-ness I have is probably my fault.


Byuu,
The FEoEZ no copier start up symptom is not something "wrong" per se with your copier or the game doctors (yes, it does it to the SF3 as well). The problem is the SPC7110 cartridges don't respect the /CART signal... in fact they don't even use it. So the copiers raise /CART to tell the cartridge not to respond to /RD signals and it does anyway. I had this going on for quite awhile before I figured out that I was probably hurting the copier and the cartridge.

All successful tests I ran with un-modified cartridges were on the SF3. I would run something that loaded to ram. I would insert cartridge and run test, and remove cartridge when done. Repeat cycle.

For what it is worth, I consider Tototek very trustworthy. Just be very patient if you need to contact a human (well, tomy, as he does everything) as they get swamped. I've ordered from them 4 or 5 times, and talked to the owner plenty of times.

As mentioned in the SPC7110 thread when discussing all this copier stuff, I have some SF3's (they were like $30 bucks each, so why not grab a couple). I already sent one to Caitsith2 during this so he could help out. I only have one spare one now, and I already promised it to someone else once I change out the broken floppy drive (old hardware doesn't last forever unfortunately). If there doesn't seem to be an easy way, I can tell my friend that I'm sending it to someone else instead... I'd really prefer not to do that.

Another option would be Caitsith2 finished laying out a PCB which is an upgrade of the PC parallel port <--> SNES cartridge interface we used to use. It lets you control everything, and right from your PC.

[byuu]

The FEoEZ no copier start up symptom is not something "wrong" per se with your copier or the game doctors (yes, it does it to the SF3 as well). The problem is the SPC7110 cartridges don't respect the /CART signal... in fact they don't even use it. So the copiers raise /CART to tell the cartridge not to respond to /RD signals and it does anyway. I had this going on for quite awhile before I figured out that I was probably hurting the copier and the cartridge.

All successful tests I ran with un-modified cartridges were on the SF3. I would run something that loaded to ram. I would insert cartridge and run test, and remove cartridge when done. Repeat cycle.

Oh, crap ... if it's still required to rip out the copier anyway ... :(
I was under the impression someone on NESdev got it working without doing that. Sigh ...

If there doesn't seem to be an easy way, I can tell my friend that I'm sending it to someone else instead... I'd really prefer not to do that.

No, absolutely not. Please don't do that. I'll find one some other way. But thank you for the offer.

Another option would be Caitsith2 finished laying out a PCB which is an upgrade of the PC parallel port <--> SNES cartridge interface we used to use. It lets you control everything, and right from your PC.

Yes! That would be awesome, too! I've always thought it would be neat to have a cart <> PC interface, so you could dump games and write PC programs to poke at the special chips. Make it think it's connected to a real SNES. Use a cheapo "SFC<>SNES" ripoff adaptor so you only have to solder that instead of individual carts.

But SNES<>PC would be great for running hardware tests. Get something in RAM somehow running, and then upload new tests from the PC so that you don't have to keep ripping out the copier with the system on. That would not be a total solution right there, but a great step toward improving the lifespan of the copier.

I know blargg wrote a SNES->PC serial linkup through his controller port, but I don't know if that was bi-directional, nor do I know if that's necessarily fast enough, but something like that might be feasible, too; if blargg wouldn't mind posting schematics. Heck, I can always give it a try myself, if I can get all the parts I need at Radio Shack. But somehow I think I'll fail :P

[powerspike]

Ah that's a shame that you don't need one after all. I've been meaning to ask my friend to pick me up a bunch of cool stuff from over there.

[byuu]

Well, I could still definitely use one ;)

I'm sure it will be helpful for games that respect /CART, and having a backup to my current copier would be nice.

[neviksti]

All successful tests I ran with un-modified cartridges were on the SF3. I would run something that loaded to ram. I would insert cartridge and run test, and remove cartridge when done. Repeat cycle.

Oh, crap ... if it's still required to rip out the copier anyway ...
I was under the impression someone on NESdev got it working without doing that. Sigh ...

I think you misread what I wrote.
You don't need to rip out the copier. The copier stays seated in the snes the whole time. All you do is insert or remove the cartridge in the external cartridge slot of the copier. Only have the cartridge inserted while code is running in ram.

When I made the mod cart, I disabled the part of the SPC7110 cart that was clashing so I never had to remove it (I was destroying the cart in the mod anyway, so I figured I might as well make it easy to use). This disabling is very easy... just cut one trace and solder a pin to the one next to it. If that sounds acceptable I can explain with details (and maybe pictures I guess if I borrow a camera).

Yes! That would be awesome, too! I've always thought it would be neat to have a cart <> PC interface, so you could dump games and write PC programs to poke at the special chips. Make it think it's connected to a real SNES. Use a cheapo "SFC<>SNES" ripoff adaptor so you only have to solder that instead of individual carts.

At one point you could buy SNES cartridge connectors from electronic stores (I got some from MCM electronics many years ago). I think you can still get them online from ebay or specialty stores. That is what the PCB currently uses (the first one I made used a gameboy connector, but that doesn't pass some pins through).

But SNES<>PC would be great for running hardware tests. Get something in RAM somehow running, and then upload new tests from the PC so that you don't have to keep ripping out the copier with the system on. That would not be a total solution right there, but a great step toward improving the lifespan of the copier.

As an aside, one advantage of the SF7 is a parallel port. You can communicate with a PC.

I know blargg wrote a SNES->PC serial linkup through his controller port, but I don't know if that was bi-directional, nor do I know if that's necessarily fast enough, but something like that might be feasible, too; if blargg wouldn't mind posting schematics. Heck, I can always give it a try myself, if I can get all the parts I need at Radio Shack. But somehow I think I'll fail

I'm starting to get a bit lost.
Are you saying something like this:
- load program into ram using copier
- remove copier
- insert cartridge or whatever
- ram program can communicate via serial link and no more hot swapping is necessary

Is that the idea?
That sounds very neat, and would be fun to make. But the amount of effort required, for something that sounds to be solely a SPC7110 issue, seems immense.

(And why do you always think you'll somehow fail if it involves hardware work? )

EDIT: By the way the controller port is bidirectional. So bidirectional serial communication is possible.

[powerspike]

Well, I could still definitely use one

I'm sure it will be helpful for games that respect /CART, and having a backup to my current copier would be nice.

Haha alrighty. I'll kick his ass then to get one for me. Maybe pick up some NES bootleg stuff since I get a good laugh out of those. :]

Edit: found some place called Ho King Shopping Centre. I'll have him look around there.

[neviksti]

Since shopping sites came up,
Where's a good place to search out rare-"ish" older system equipment?
There are a few things I've always wanted to try out (XBand SNES keyboard for example), but don't know where to find.

[Snark]

I have a spare SF3 I could donate.

[powerspike]

Well it would probably be faster seeing as my friend works a very demanding job over there. I mean I could get it, but it might be like a month from now or longer for said copier.

[badinsults]

Since shopping sites came up,
Where's a good place to search out rare-"ish" older system equipment?
There are a few things I've always wanted to try out (XBand SNES keyboard for example), but don't know where to find.

Here is something that might interest you, neviksti:

http://www.gamesniped.com/2008/07/23/sn ... est-carts/

As for the Xband keyboard, good luck. If one ever popped up on ebay, I bet it would go for several hundred dollars, possibly over 1000. Just look at the extremely rare MACS rifles, which are probably more common.

[byuu]

Just an update: I received a GDSF7 from Tomy (yes, I paid for it.)
Just need to get a 9v adapter, don't want to overheat the voltage regulator with my 10v adaptors. Heh, I'm going to miss the eye candy UFO8 UI.

Using neviksti's trick, I should be able to run codein RAM, then connect the FEoEZ cart afterward to probe the registers. But I want to verify that won't damage anything before I try it, of course.

Also, blargg made me a PC<>SNES serial "joypad" interface and a dev RAM cart. With that, we've been able to verify power-on timings and default register states, and more stuff should come in the future. It should also provide yet another way to run custom code without the need for floppy disks. Best of all, it should allow for stop'n'swop with either a copier or the dev cart.

I'm planning to add a core randomization config option that will initialize stuff like memory to "random" patterns, simulate the power-on and long-dot drifting that you see on real hardware, etc.

Really cool stuff all around, that should help out immensely, if I can just stop being so lazy lately and run more tests!

I feel pretty confident that all of this can keep me hacking for quite a while, the joypad thing will even bring new life to the UFO with the dead SRAM.

Of course, I won't turn down any offers for more cool stuff :D

[mozz]

Once again, I'd just like to say thank you, to byuu and others, for continuing to push for more accurate and complete info about the hardware of the SNES and (in this case) special chips. I think in the long run your efforts will be very helpful in having quality accurate emulators. (Already there are great strides in this direction, in just the last couple of years. Its very encouraging.)

I'm glad you guys are motivated enough to design and perform the experiments to collect this trickle of obscure knowledge. These facts are not going to come from any other source. The people (original engineers) who knew the answers you seek have either forgotten them by now, or would not be allowed to share them with you anyway. The facts exist, manifested in these physical pieces of hardware, but buried and inaccessible to most of us. Through poking and prodding and reverse-engineering, you coax these devices into giving up their secrets, and then share them with the community. Other people might not care, but I appreciate the service that you're doing, in seeking out these precious scraps of knowledge.