Outboard MIDI

[d0pefish]

Are any of the the Show Info - No, Yes, LCD-On (FB), LCD-Auto (non-FB) for MT32-Pi intended to provide an OSD over the HDMI output of Mister? I don't have the OLED display as indicated but at times would like to see the information on screen to validate it's working. None of the options I have tried work. LCD-On shows a dark gray box, but nothing inside of it.

Either you haven't configured mt32-pi correctly (see previous post) or you're not using the custom user port->Raspberry Pi wiring scheme. You can't do this just with plain MIDI; there needs to be an I2C connection between the Pi and the MiSTer.

[shimian]

Are any of the the Show Info - No, Yes, LCD-On (FB), LCD-Auto (non-FB) for MT32-Pi intended to provide an OSD over the HDMI output of Mister? I don't have the OLED display as indicated but at times would like to see the information on screen to validate it's working. None of the options I have tried work. LCD-On shows a dark gray box, but nothing inside of it.

Either you haven't configured mt32-pi correctly (see previous post) or you're not using the custom user port->Raspberry Pi wiring scheme. You can't do this just with plain MIDI; there needs to be an I2C connection between the Pi and the MiSTer.

I have full control over the mt32-pi via the OSD, I wired it myself to the arduino header with some jumper wires. I can change modes/etc of the mt-32pi from mister GUI, but when I toggle Show Info, I get nothing. Which is the correct Show Info mode to use with no LCD attached to the pi directly?

[d0pefish]

You can use any of the Show Info modes whether there is a real OLED attached or not - but you still need to enable type = ssd1306_i2c in the lcd section of mt32-pi.cfg.

[shimian]

You can use any of the Show Info modes whether there is a real OLED attached or not - but you still need to enable type = ssd1306_i2c in the lcd section of mt32-pi.cfg.

So this is odd. When I enable ssd1306_i2c, I lose mister control and the mt32 submenu disappears from my Mister OSD. Thought? I have everything wired to the schematic.

If I set LCD off, it works perfect.

[d0pefish]

Check your wiring, keep it as short as possible; <30cm. I2C/I2S isn't designed for long distances, so if the wires are too long the data will be corrupted.

[shimian]

Wiring is clean, they're 20cm breadboard jumpers. Both I2C wires look fine. Super odd that enabling the LCD makes it stop working. Just validating that this pinout is fine for I2C

Pin3/D+ - i2c sda
Pin7/drain - i2c scl

I have those wired to arduino pins 1 and 6 respectively using breadboard jumpers. I'll double check everything again.

[shimian]

wiring is good, matches the schematic. Sound via mt32 works perfectly. OSD lets me control modes, soundfonts, etc. Once I enable the LCD though in mt32-pi.cfg it no longer communicates with MiSTer via i2c (I lose the OSD options).

Tested a few times just now back and forth.

lcd=none - everything works perfect, sound, osd, mt32/soundfont mode, etc
lcd= ssd1306_i2c - I lose the mister control of mt32-pi.

Do i need to adjust the LCD i2c address in the config? Is there a debug flag I can set on mt32-pi that might log out why it's having trouble? I don't think it's a wiring issue as it works fine when I don't try to specify an LCD.

I am using latest mt32-pi and latest Main Mister/AO486 core. Running on a Pi3 b+

Let me know if a GitHub issue is a better place to report this too, more than happy to open one up.

[d0pefish]

Just for the benefit of those following along, this was resolved after some chat outside the forum and looks like it turned out to be an electrical signalling issue.
If you're rolling your own wiring, bear in mind that I2C is extremely sensitive and can be broken by noisy signalling - keep wires short and use shielded cabling if possible.

Enjoy

[V_Amo]

Just for the benefit of those following along, this was resolved after some chat outside the forum and looks like it turned out to be an electrical signalling issue.
If you're rolling your own wiring, bear in mind that I2C is extremely sensitive and can be broken by noisy signalling - keep wires short and use shielded cabling if possible.

Enjoy

Thanks for the update. I’d double checked my config and wiring but still couldn’t get it to work. Now I’m thinking my issue is the same as this. I’m using short breadboard wires connected to a short USB 3 cable. I’ll try disabling the display output and, if that works, shortening my cables further.

[rhester72]

For just _standard_ outboard MIDI to a MT-32 (read: USB-to-MIDI M-audio Midisport Uno), what settings are required in ao486?

In the OSD for UART Mode, I have:

Connection: MIDI
MidiLink: Local
Type: MUNT
Baud: 31250

MidiLink.INI relevant options:

Code: Select all

MIDILINK_PRIORITY   = -20
MUNT_OPTIONS        =
MP3_VOLUME          = -1
MUNT_VOLUME         = -1
MODEM_VOLUME        = -1
FSYNTH_VOLUME       = -1
MIXER_CONTROL       = Master
MUNT_ROM_PATH       = /media/fat/linux/mt32-rom-data
FSYNTH_SOUNDFONT    = /media/fat/linux/soundfonts/SC-55.sf2
UDP_SERVER          = 192.168.1.130
UDP_SERVER_PORT     = 1999
UDP_SERVER_FILTER   = FALSE
UDP_FLOW            = 0
TCP_SERVER_PORT     = 23
TCP_TERM_UPLOAD     = /media/fat/UPLOAD
TCP_TERM_DOWNLOAD   = /media/fat
TCP_TERM_MP3        = /media/fat/MP3
TCP_TERM_MIDI       = /media/fat/MIDI
TCP_TERM_SYNTH      = FluidSynth
TCP_TERM_ROWS       = 23
TCP_FLOW            = 3
TCP_SOUND           = TRUE
DELAYSYSEX          = FALSE
#USB_SERIAL_MODULE   = /media/fat/pl2303.ko

[AO486]
MIDI_BAUD           = 31250
UDP_BAUD            = 38400
#TCP_SOUND_CONNECT   = /media/fat/SOUNDS/connect_sav.wav

I'm running SOFTMPU with the options /SB:220 /IRQ:5 /MPU:330 /OUTPUT:COM1

I get the string "MiSTer MidiLink! BB7" on the LCD, but so far, I've tried X-Wing for DOS and Kyrandia 1 for DOS, and get no MIDI sound (or activity light on the USB cable) at all.

[rhester72]

Huh - it appears SOFTMPU was the culprit. I stopped loading it, and bam - MIDI!

Was one of the options I was passing wrong, or is it just not needed?

[onaryc]

Hi,

I'm making a custom USB to gpio pi cable. The pb is that i have a 5.5 XL IO board : i do not have the 10k pullup resistors added in the 5.6 revision (https://www.atari-forum.com/viewtopic.p ... 1b#p368980) and i think in the latter revision a supplementary i/o user connection was added (for genesis/megadrive controller with snac board?).

From the schematics, i found that i should add 10k resistors between usb pin1/vbus (or rpi gpio pin2) and :
* usb pin 2/d- (or rpi gpio pin 10)
* usb pin 3/d+ (or rpi gpio pin 3)
* usb pin 5/tx- (or rpi gpio pin 35)
* usb pin 7/GND_DRAIN (or rpi gpio pin 5)
* usb pin 8/RX- (or rpi gpio pin 40)
* usb pin 9/RX+ (or rpi gpio pin 12)

I do not have the extra pin, but it does not seem to be connected to the rpi (a pull up resistor shall be added for it if my board was modded).

The thing that is bothering me is that i though (at least on the 5.5 revision) that usb pin 1/vbus was delivering a 3.3v voltage but is connected to the 5v line on the rpi gpio pin (to power it i suppose). Am i missing something? is the last io board revision capable of delivering 5v on the vbus line? Also there is a 3.3v on the rpi gpio pin 1, can it be used to power the rpi instead of the 5v?

[onaryc]

ok, after @blue1 answer some of my questions, it appears that :
* usb pin1/vbus IS 5v and shall not be used for the pull up vcc
* usb pin 6/tx+ can be used for the vcc (as in the 5.6 board version).

So pull up resistors shall be added between usb pin 6 and 2, 3, 5, 7, 8 and 9.

The next pb is that if i add the extra io pin, where should i take the vcc for the pull up in order to have a cable compliant with all existing board revision (or is the io6 selector shall be used in this case?)

[syntax_x]

I was having issues getting sound back into my MiSTer using a cut USB cable like a few others here.

Turns out id used the wrong end of the USB cable and the colors were swapped on the twisted pairs.

Working fine now