BTW, just another idea of what might be wrong. Did you program the microcontroller fuses, and not just the hex file? Without the fuses configured, the crystal won't be set oscillate, and the FTDI chip can't communicate. (The crystal is shared between the chips, and the microcontroller is driving it.)

*FT232BL

The chip labeling is usually made using laser engraving. If it less readable on the chip that is on the board, this may be due to corrosion, but also just to a layer of flux residue sitting on top of the chip. Try cleaning the top of the chip using a q-tip and some rubbing alcohol if you care, and see if that makes it more readable. And anyway, I'd be more worried about the pins than about the top of the chip case.

Yup. smile

In the interest of promoting sample playback glitchiness, there's something else you can do. There are four 256 byte lookup tables that are for two channel wave mixing. This is the dist setting, which you can normally set to clip, shap, shap2 or wrap. If you hold A and pressing left or right 4 times beyond the leftmost or rightmost setting, you can use data from anywhere in the memory map for this table. Most of it just sounds harsh, but worth checking out. This only works when there's a sample on BOTH the software channels, otherwise this table is not used, and you will hear no difference.

564

(1,485 replies, posted in Trading Post)

I guess with those you can press left + right easily? (If you tried.)

Could you tell me the LSDj version you're using and share the save file with me?

The exported wave file should always be sample perfect, even if the audio playback is not. As noted in the thread I linked to, there is a bug in versions prior to 1.4, and you should upgrade. Other than that, the timing should be as consistent as it can be on a Gameboy with respect to CPU load. (In other words heavy use of tables and vibrato and such.) Using GBC a the emulated CPU will give you more GB side CPU power.

And as for Gambatte and the save size, I haven't experienced that problem. Again, could it be that you need to upgrade to a later version?

567

(5 replies, posted in Nintendo Handhelds)

Only the newer adapter, which has a red circuit board, has support for MIDI. The older one can only be used for file transfers. (Save data/software update.) You also need to mack sure the adapter's mode is set correctly, which you can do by running SYNC.bat included with the NL software for Windows. (Not sure if/how you can do this in OSX.)

Are you trying to use this for recording or live performance? You can do two things, sync two instances of BGB using a virtual link cable, and export audio. If you want to export audio from two instances of LSDj running synched, you snould combine these methods. If you just want the two instances to stay in synch between each other, just using a virtual link cable would be enough. It's all described here. Study it carefully.

http://chipmusic.org/forums/topic/12730 … aq-thread/

If you really, really care about getting this board diagnosed/fixed you could send it to me. Qualifications: Working on a new flasher project on my own, nowadays quite comfortable soldering even the smaller SMD variant used for the FTDI chip on the newer programmers.

First I need to ask, has it ever worked, to any capacity? Exactly how does it not work? When you plug it in, do you see a USB device n the device manager? Exactly which FTDI chip is that? Have you programmed the AVR chip? Using which method?

My first reaction is that it's indeed the soldering around the SMD chip which is the main problem. The first problem is that some of the pins may lack enough solder to make reliable contact. The second more serious problem is the flux residue. The flux, depending on type - needs to be cleaned off. What appears to be happening is corrosion from the flux residue is starting to eat the tracks. In particular the ground connection to the FTDI chip (the middle connection on the left side of the image, going out to the big copper lane.) I'm not sure from the image whether that's just dirt or if it's corrosion that has eaten all the way through.

Limitbreak wrote:

I always thought having a "GB-303" would be cool by somehow adding a filter knob, res knob and perhaps a decay knob onto a DMG. I don't know how one would go about it...it's probably overly ambitious but a designated GB acid box would be fun.

I have two different ideas for doing that, one using hardware and one using software, and will probably try implementing the software version at some point.

Different problems have different causes. Things like crashes and a corrupted logo can be caused by bad contact. Eg, dirt, misaligned cartridge connector and perhaps easy to overlook, that the cartridge/cartridge connector is slightly bent from internal pressure from a backlight panel or other mod, which might also cause bad contact.

The "dark screen" problem on the other hand is never caused by bad cartridge slot contact.That is purely an electrical problem, caused by varying power consumption when the CPU is power save mode. When the CPU's power consumption goes varies, so does the LCD voltage. There's not really a way to completely fix that without redesigning the power supply circuit. The easiest way to fix the dark screen problem is probably to edit the LSDj ROM, like I said above.

I think the Nanoloop problem is a bad contact problem, and you may have luck not inserting the card-tridge (haha) fully and see if that makes a difference. You might also want to try to burn the version of the ROM that doesn't have the custom "hello" logo. Click the following link and scroll own to "Scrolling Logo" You need a Nanoloop MIDI adapter to update the ROM.

http://www.nanoloop.de/midi/one.html

572

(84 replies, posted in Nintendo Handhelds)

bitjacker wrote:

what version of nanoloop is this?

1.2 or 1.3 iirc. Don't remember which now. One of these days I'm going to give it a good article. (Yeah right, as if I'm actually going to DO stuff.)

That the screen going black is related to LSDj's power saving feature, in conjunction with other parameters like GB type, possibly the condition of the unit (dried capacitors) as well as battery type/quality/health.

The brute force way of solving the problem would be disable LSDj's power saving feature. Older versions used to have a setting for this in the menu screen, but newer versions don't, and instead turn off power saving when playing the song, and keeping it on otherwise. You can however still disable it by editing the ROM. Open the ROM in a hex editor and look for 76 00 C9. There should probably only be one occurrence of this string in the file. Change 76 to 00. In the latest version of LSDj as of now, 4.7.3, this instruction is at address 4B9.

Keep in mind that this *will* drain your batteries slightly faster.

If you have a blue (non-USB) EMS cartridge that doesn't hold a save, or possibly won't boot, please do the following: Look in the top left corner and tell me what the text says. This text denotes which board revision the cartridge is. You may need to open the cartridge if you have a sticker in the way. It always starts with GB32, but please tell me what follows. I'm speculating that many cartridges of the type GB32I-1 has a specific problem that I think I've found a solution to, but I want confirmation from other people.

575

(30 replies, posted in Nintendo Handhelds)

Yeah, it's part of my one hour chiptune rescue guarantee. loljk.

576

(30 replies, posted in Nintendo Handhelds)

Today is your lucky day. I looked into this and produced the following IPS patch which both allows the game to be run on a GBA, and fixed the sample playback problems on GBA, using the "antispike" method that I invented which is now also used in LSDj.

http://gbdev.gg8.se/files/musictools/Po … sic%20fix/

Please report any problems with the patch here.