TK.

Just to ensure: when you are trying an upload of pic18f452\midi\update_with_old_mios.hex now, which messages are sent by the core? I would propose to post all messages of the MIDI In monitor here (use copy&paste) - this helps to understand, in which state your PIC is now. Best Regards, Thorsten.

Dann gehe mal schrittweise vor: schliesse ein DOUTX4 Modul nach dem anderen alleine an das Core Module an, und ueberpruefe, ob die Buttons noch funktionieren. Beim fehlerhaften Modul dann die 74HC595 entfernen, und schauen, was passiert Wenn die RCLK (oder SCLK) Leitung dann immer noch kurzschliest, mach doch mal ein Foto von den Platinen und poste sie hier (oder schicke sie mir zu, ich kann sie dann ins Netz stellen) Gruss, Thorsten.

Sorry, I hope that my answer was not too late - do you still get the upload request during power-on? Best Regards, Thorsten.

You still get an upload request during power on, this is your luck! Be very carefully now, so that not the same disaster happens like here: http://www.midibox.org/forum/index.php?topic=6560.0 First you need to revive the old MIOS version by uploading the mios_v1_8.hex file (it's available here: http://www.ucapps.de/mios/mios_v1_8.zip Thereafter upload the pic18f452\midi\update_with_old_mios.hex - please don't try any other file of the update package! Best Regards, Thorsten.

With the core module of course :) See also http://www.ucapps.de/midibox_sid_walkthrough.html Best Regards, Thorsten.

Ok, ...40 00... bedeutet MIOS Device ID 0x00 Wenn selbst der io_loopback funktioniert, kann es eigentlich nur noch am Rx Pin liegen (R.I.P.) Gruss, Thorsten.

Sieht eigentlich alles sehr gut aus. Es gibt jetzt nur noch zwei Moeglichkeiten: entweder ist die Device ID != 0x00 (dies erkennt man am Upload Request, Du koenntest die SysEx Meldung mal hier posten), oder der Rx Pin ist wirklich defekt. In diesem Fall muesstest Du Dir wirklich einen neuen PIC kaufen (EUR 6.70 bei Reichelt) - falls Du kein Programmiergeraet hast, koenntest Du mir den neuen und den alten PIC (zur Analyse) zuschicken... Gruss, Thorsten.

Verfolge mal die SCLK und RCLK Leiterbahnen auf der Platinenunterseite, vielleicht gibt es hier einen Kurzschluss? Gruss, Thorsten.

When you are typing "make", you will see the allocated blocks after the linking phase (they are displayed be the hex2syx script). 0x7fff is the maximum on a PIC18F452. The linker will exit with an error, if this boundary is exceeded I guess that you program currently doesn't allocate more than 2k, and by doing clever programming (e.g. use subfunctions for frequently repeated code, use data tables for a better oversight), you won't need more than 4k, so there will still be about 16k free for much more stuff Best Regards, Thorsten.

alright - this means that there is a problem with reading back the data line. You can follow back this line from PIC pin RB7 to LPT pin 10. You could use your multimeter in order to check the voltage levels when the data line is set to 0 (=0V) and 1 (=5V) Best Regards, Thorsten.

No, F stays for Channel 16 (0xf = 15, we are counting from zero). Alternatively you could replace the "iorlw 0xb0" by "movlw 0xb0" in order to force MIDI channel #1 on the SID, but since it will still receive MIDI data over channel 16, changing the MB64E setup is the better solution. Note that MB64E provides a "global MIDI channel" where you can overload the channel settings of your setup. It's really simple, and can be changed from the menu. Best Regards, Thorsten.

Mit main.asm meine ich das Source File vom srio_interconnection_test - hast Du ihn wirklich ausprobiert? Das ist absolut erklaerbar. Button 7 ist der erste Button in der Shiftregister-Kette. Wenn die Taktleitung nicht richtig angeschlossen ist, oder einen Kurzschluss hat, sieht das Core Modul nur diesen Button Die LEDs sind an der gleichen Taktleitung angeschlossen... Gruss, Thorsten.

Just add bsf SID_STAT, SID_STAT_CC_DUMP_REQ anywhere to cs_menu_buttons.inc Best Regards, Thorsten.

Hi *, I've finally released v1.7303 to the public. In distance to the beta versions made before, the 303 mode now sounds more punchier than ever before. :) Due to all the bugfixes which have been made during 13 months after v1.7a, an update to the most recent version definitely makes sense, otherwise you will miss a lot of cool stuff! Please note that an update to MIOS V1.9 is required as well! Download http://www.ucapps.de/mios_download.html ChangeLog http://www.ucapps.de/midibox_sid_changelog.html Bassline Tutorial http://www.ucapps.de/howto_sid_bassline.html Have fun! :) Best Regards, Thorsten.

Great! So, now where you've tried again how to upload an application, it's time to do the next step: update your MIDIbox to MIOS V1.9 and MBSID V1.7303 :) Best Regards, Thorsten.

Yes, the Karma applauds/smites are logged... Best Regards, Thorsten.

Wenn der MIDI Port Noten versenden kann, heisst das (unter Linux) leider nicht unbedingt, dass er auch SysEx Daten fehlerfrei versendetet. Gerade letztes WE habe ich diese Erfahrung nach der Neuinstallation von FC5 auf meinem Notebook gemacht - manchmal werden SysEx streams verstuemmelt - debugging ongoing... :-/ Du koenntest mal folgenden Test machen: den Hardware Loopback durch den Optokoppler realisieren (also: PIC raus, Rx mit Tx Pin verbinden), und mit MIOS Studio erstmal eine Note versenden. Wenn sie im MIDI In Monitor erscheint, mal ein .hex File versenden. Siehst Du die SysEx Daten? Sind sie immer gleichlang? Gruss, Thorsten.

Wenn ein einzelner Button mehrere Events ausloest, dann liegt es meistens an einem Kurzschluss an oder zwischen der SCLK/RCLK Leitung. Mit dem srio_interconnection_test kann man das sehr einfach debuggen (die Anleitung steht in main.asm) Gruss, Thorsten.

It will work with MIOS, and your code is a good starting point, because it will send something - so you will be able to debug this! However, checking for evnt2==0 is propably not correct, I'm missing the check for "evnt0==0xb0" (CC CHannel 1), etc... but I'm very sure that you will be able to complete the program by yourself once you've worked with the core module. Just try it, get some first experiences, and once you face a problem and you are not able to solve it by yourself, ask in the forum. Just two additional hints (beside of the missing evnt0==0xb0 check): use MIOS_MIDI_BeginStream() and MIOS_MIDI_EndStream() only once, not several times within a SysEx messages. For the beginning you could left it out, these functions are only required for the MIDIbox Link function (when several core modules are chained) Binary->BCD conversion: if this is really required, use MIOS_HLP_Dec2BCD, it's the simplest way Best Regards, Thorsten.

Just a prewarning that this wasn't the last tweak in the 303 sound engine option. Now where I'm working on patches without an external filter, I see some room for improvements. So, beta17 will be released soon ;-) Best Regards, Thorsten.

Hi Steve, I will propably accept this, but only under following circumstances: rename the project to MIDIbox SID <your abbreviation like -LC for low-cost) in order to make the origin of the design clear add the URL www.midibox.org (not only your own - this is like hijacking the IP of somebody else and not a good style) replace the oscillator by a direct clock connection to the core due to this issue http://www.midibox.org/forum/index.php?topic=5748.0 add R10 to the PIC - even it might currently not be required, it could be used for any purpose in future allow to add more banksticks for MBSID V2 (the subsystems require different BankSticks, see wishlist) - optionally just add a pin header for extensions consider to change the layout, so that the pin headers for DIN/DOUT and LCD could be mounted test your board intensively before offering it to other people to avoid additional support effort in the forum I can only make a final descition once you were able to get the board successfully running. Best Regards, Thorsten.

Hi Carsten, what do you mean with different numbers - different types (6510 vs. 8580) or different production dates? However, in both cases you can use the same PSU so long you ensure, that the 6510 gets 12V, and the 8580 gets 9V Best Regards, Thorsten.

Hi Olivier, this is not possible with the multiplex method - just imagine about how it works, maybe it makes sense to draw a flowchart, a diagram or whatever. So, the only possibility is to split the 16 x multiplexers to two 8 x multiplexers (or less) in order to decrease the update period The best resistor values depend on the LEDs you are using - you have to do some experiments. In normal case (when cheap LEDs with low power consumption are used), the LEDs are bright enough, so that no special measure is required. the AND opteration which masks out the first 8 bit is not really required, but it's a good programming style. The >> operator leftshifts the 16bit word - the purpose of these operations is to select the lower and upper byte of a 16bit word Best Regards, Thorsten.

Hi Olivier, yes, this is normal, because each bar is only service fourth the time now. You could increase the brightness by using smaller resistor values, or by using more dout pins (4*4 select lines instead of 4), so that the counter is reset at lower values Best Regards, Thorsten.

Have you already tried the mbsid_interconnection_test and the sid_testtone application? Best Regards, Thorsten.