nesdev.org

If anyone has a console that reports the SDRAM Error, this mapper set will hopefully fix that problem. No boot ROM flashing or hardware changes needed. It works here on the 8 various consoles I have but won't be official until more peeps try it out. Should also mean that the SNES mini will now be supported. I still need to get the new portable SNES to test

Also adds an autosave option for battery RAM at the saving question screen. This will save to the same file that was loaded when you started the game.

Download at: http://www.retrousb.com/downloads/POWERPAK132103b.zip

Has the sram issue with rom expanded ExHiRom games been fixed? I tried out dq3e and it still doesn't save.

I'll try it when I get home. Does this also fix the DSP-1 detection error?

shadowkn55 wrote:Has the sram issue with rom expanded ExHiRom games been fixed? I tried out dq3e and it still doesn't save.

Nope, just those changes noted for now.

mic_ wrote:I'll try it when I get home. Does this also fix the DSP-1 detection error?

Yes both were the same problem, where some timing signals ended significantly later on the mini and some EU 2/1/3 systems.

bunnyboy wrote: Also adds an autosave option for battery RAM at the saving question screen. This will save to the same file that was loaded when you started the game.

how exactly will this work ? you must still press the resetbutton for 5 seconds or so ? (i'm not sure how it works because mine is not arrived, till now)

are you already in conversation with d4s for the ultra 16 save-compatibility ?

Zoigl wrote:how exactly will this work ? you must still press the resetbutton for 5 seconds or so ?

You still hold reset for 5 secs and the screen asking to save appears. But now there is an option to use the ram file that was originally loaded so you do not need to navigate and choose it again. There are still the options to choose a different file or not save at all.

Zoigl wrote:are you already in conversation with d4s for the ultra 16 save-compatibility ?

Don't know what that is...

Great to see continued support for the PowerPak, will report back any problems encountered by the end of the week.

Zoigl wrote:
are you already in conversation with d4s for the ultra 16 save-compatibility ?

Don't know what that is...

When the PowerPak has switched to SRAM saving mode after holding reset for a couple of seconds, it will output what appears to be garbage data around range $00-3f:21c0-$00-3f:21ff.
Intended or not, this breaks compatability with devices using that memory region, among these are the LifeFitness bike interface and Ultra16, which is a BIOS for the SNES of sorts I developed a couple of years ago. Ultra16 menu on youtube

Unless there's no way around it, it would be greatly appreciated if you could either disable that data output or move it to some other region.

Mini-derail, please feel free to respond via PM.

d4s, how did you get info on the LifeFitness interface? Did you disassemble the game, or did you actually buy one of those crazy ass things? Heh.

Fascinating to hear that it connects to the SNES' B-bus like that. I figured it was just a fancy controller.

$21FF is used for the SDRAM DMA port, so that won't be moving. Might work to use the DMA=0 config option so it never touches that port. Sounds like cool hardware tho!

I just updated to the latest mappers and it is working on my SNES jr now. I will keep an eye out for any new bugs.

Mitch

bunnyboy wrote:$21FF is used for the SDRAM DMA port, so that won't be moving. Might work to use the DMA=0 config option so it never touches that port. Sounds like cool hardware tho!

Thanks, that's good to know!
Using $21c0-$21c3 myself here.
I'm pretty sure the PowerPak was outputting data on the whole upper $21xx range, but only in SRAM save mode. Will look into that and the new mappers later.

d4s, how did you get info on the LifeFitness interface? Did you disassemble the game, or did you actually buy one of those crazy ass things? Heh.

Fascinating to hear that it connects to the SNES' B-bus like that. I figured it was just a fancy controller.

Yeah, I finally bought one of these interface boxes after several years of hunting on ebay. Guess I'm just a sucker for crazy hardware.

To be perfectly honest, I haven't researched it as much as I'd have liked to yet.
It is an interesting piece of hardware, though. The base unit has battery-backed RAM and an RS232 interface IIRC.
A PCB-shot should be in there somewhere: http://dforce3000.de/?p=filedump&srch=exertainment

Will send it over if you ever feel like emulating it.

I'm glad you've got hopefully all of the compatibility issues fixed now. I figured like the early NES PowerPAK issues we would see them resolved shortly. One question I wanted to ask you was about ROM loading and how you determine loading and mapping. The reason being certain strange roms like the BS-X roms have been reported not to load properly. They work just find in copiers like the GDSF7 though. So if you can share how the loader determines the cartridge setup that would be great as it would be nice to get games like the BS Shockman Zero and BS Zelda working, assuming the reports I've heard are true and they don't currently work.

It is an interesting piece of hardware, though. The base unit has battery-backed RAM and an RS232 interface IIRC.
A PCB-shot should be in there somewhere: http://dforce3000.de/?p=filedump&srch=exertainment

Will send it over if you ever feel like emulating it.

Really appreciate the offer, but I really don't think I could pull it off. Wouldn't want to deprive you of something so rare. And even if I could (I can't), the hotel overlay thing looks like a lot more fun, bleh. Besides, still have Matthew's BS-X over here collecting dust >.>

I do notice there's an EPROM on that board, with the UV window covered. So it looks like we need to emulate the TL16C550AN (which has serial interface stuff in it, looks like it was meant for a modem, maybe an online firmware update mechanism or service mode?) and connect it to that EPROM's data.

http://www.alldatasheet.co.kr/datasheet ... 550AN.html

God, all sorts of controller insanity, too. SNES to RJ45 ... I am aware you can use RJ45 as a ghetto serial cable, telxons usually do that. So all in all, wow. This is an insane amount of work for something that can't even be emulated via keyboard / mouse / gamepad. Amazing.

I posted some minor software disassembly notes here:
http://board.byuu.org/viewtopic.php?f=3&t=477

If you want to continue work on this sometime, let me know and I'll try and help out, but I don't think I want to do all of that work alone :/

First impressions:
-Autosave works great! (How about auto-loading *.srm files that have a filename equal to the ROM's that's been selected for loading?)
-2Mbit hirom games don't work yet.(example: Hungry Dinosaurs.
Not valid by official Nintendo specs, but how about having a minimum bootable filesize of 32kbyte(lorom) and 64kbyte(hirom)? Would be great for homebrew stuff.)
-no issues with SDRAM/DSP-1 errors so far on any unit.

Regarding that $21xx-region issue:
-after holding reset to enter SRAM saving mode, the PowerPak responds to read accesses from $21c0-$21ff. Here's what it looks like, although that picture itself doesn't prove anything in particular, of course: http://dforce3000.de/index.php?p=fdsingle&uid=361
-DMA=0/1 doesn't affect this.
-Data read appears not to be random, patterns are visible.

Here's a guess:
The PowerPak DMA port is accessed at $21ff by software, but you don't actually decode the full b-bus adress in hardware, just the upper two adress lines, resulting in the DMA port being accessible from $21c0-$21ff.
The port just stays on after ROM loading.
I assume the adjustment neccessary would be rather small, unless you don't have the full adress bus b connected to your FPGA, of course.

My general experience with the b-bus is that quite a few games accidentaly access the whole bus range, most prominently during memory clearing operations. I had to put a hardware lock on the Ultra16 registers to prevent some games from messing up (Lion King comes to mind in particular).
Disabling most PowerPak registers after ROM load might increase compatability. Just an idea, though.