TK.

Hi Matteo, search for "CS_MENU_BUTTON_SID3" and "CS_MENU_BUTTON_SID3" in cs_menu_buttons.inc and change it in the following way: CS_MENU_BUTTON_SID3 movlw 0x00 movwf SID_BANK goto CS_MENU_MS_NotifyBankChange CS_MENU_BUTTON_SID4 movlw 0x01 movwf SID_BANK goto CS_MENU_MS_NotifyBankChange [/code] Best Regards, Thorsten.

Cool! Danke fuer den Tip, werde ich heute noch auf der mbhp_mf Seite einfuegen. Fuehrt Conrad auch Inkrementalgeber? (Rotary Encoders) Gruss, horsten.

It's even harder for you to change the code then changing the wires ;-) Toggle state: just save the new state with the snapshot button once all LEDs are off Best Regards, Thorsten.

Hi Robin, you are making things more complicated then they are... you don't need a specific address for the table, and you can put it into the assembler code where you want - once you are using a label (BANKSTICK_DUMP_TABLE) it will get automatically a reference which can be used in the code (-> TABLE_ADDR BANKSTICK_DUMP_TABLE). It must only be ensured that the table is not located within a function. Means: you have to make sure that the PIC never executes the data of the table (so, never do a "goto BANKSTICK_DUMP_TABLE"). Normaly I'm putting tables at the end of an .inc file, the function above the able must end with a "return" or a "goto"/"rgoto" Best Regards, Thorsten.

an auto detection for the MIOS ID itself is infeasible, each core module in the MIDI chain needs an unique ID, otherwise it wouldn't be possible to address informations (code/data) to a dediated PIC. But you don't need to change any code line in order to change the IDs of the sysex blocks, you only need to re-generate the .syx file from the .hex file with the hex2syx.pl script like described in the README.txt of the MIOS package, and you have to follow the step by step instructions which can be found http://www.ucapps.de/midibox_sid_csB.html (they are still valid) Best Regards, Thorsten.

Hi Julian, 8 AOUTs and 8 gates are planned Best Regards, Thorsten, (P.S.: the user manual is finished: http://www.ucapps.de/midibox_fm_manual.html)

Hi Martin, check the "planned additional features" list under http://www.ucapps.de/midibox_seq.html - this is the stuff which will be implemented sooner or later and which should fit into a single PIC. Once this has been done (maybe in one or two years or so...) I will think about extensions ;-) Arp styles are already available. In fact the definition of an arpeggio sequence is the definition of a style (the concept is not like that what can be found in common synthesizers) Best Regards, Thorsten.

Hallo, entweder taugt es wirklich nichts, oder es benoetigt eine gewisse Grundbelastung, damit es sich nicht permanent automatisch ein- und ausschaltet. Dieses Problem hatte ich mal mit meinem C128 Netzteil, abhilfe: ein Fahrrad-Gluehbirnchen als Grundbelastung ;-) Gruss, Thorsten.

Sure, this would be a solution... Best Regards, Thorsten.

uploading attachments should be possible now Best Regards, Thorsten.

Hi Robin, such tables can be stored everywhere in the flash memory, so you could also save it in the same file like the code which accesses it, or you can create a new file which contains all your own stuff. Note that the tblrd*+ method I demonstrated here works only when the table is located in program flash. Storing the table in EEPROM or BankStick requires the appr. MIOS_*_*Read* Funktions To check the allocated memory locations I'm normaly looking at the end of main.lst where the "memory usage map" can be found. Serge's Tool: which data do you mean exactly? Best Regards, Thorsten.

This: http://69.56.171.55/~midibox/forum/index.php?topic=3852.0 is possibly the best solution for high resolutions Best Regards, Thorsten.

Hi Luke, yes, you can cut and/or chain them You can save some money by purchasing some1x40 instead of smaller ones Best Regards, Thorsten.

Hi Michael, you don't need to modify MIOS, you could write a dedicated driver which hooks on USER_SR_Service_Finish and is part of an application Best Regards, Thorsten.

Hi Tom, yes, all seperate encoders of StepC are useful, I don't want to miss any of them :) Best Regards, Thorsten.

The PIC doesn't provide an integrated DAC. Sometime people are using the PWM outputs to generate different voltage levels, but there are only two sources, the voltages are not adequate enough and a cheap(er) DAC would do a better job Best Regards, Thorsten.

Both assumptions are correct! Best Regards, Thorsten.

Hi, the pictures are not accessible. You could use the Joystick ports as external BankStick slots. Best Regards, Thorsten.

Hi, It's even possible to build a least-cost version without an AINX4 module. See the comments in main.asm: ; Number of connected pots ; Allowed values: 1-64 #define DEFAULT_NUMBER_POTS 64 ; number of connected pots ; ; Multiplexers (AINX4 modules) connected? -> set this value to 1 ; No multiplexer connected= -> set this value to 0, allowed number of pots in this case: 1-8 #define DEFAULT_MUX_ENABLED 1 ; enable multiplexers [/code] In "unmuxed" mode the pots have to be connected directly to the J5 Best Regards, Thorsten.

Hi Luke, the technical problem which prevents me from implementing a tap function is a lack of resources within the PIC. An additional timer would be required which provides a reference clock which runs independent from the "BPM timer". Without this timer it isn't possible to measure differences between the tapped and the current tempo. Best Regards, Thorsten.

Can you please tell me why you don't configure the merger mode in your midibox64e.ini file (-> see mk_syx package)? Best Regards, Thorsten.

Hi, can you please write your questions in a more readable form, because I'm not able to find out which buttons should be useless with pots, and if you already know that only two displays can be connected to the core module or not... (resp. which displays are supported by the firmware) To the part list: maybe somebody else can check this? Best Regards, Thorsten.

Hi, yes, the wiring of the LEDs is very simple when you are *not* using a stripboard. I took 4 long blank cables for the cathodes (in a row), and 6 shorter blank cables for the anodes (in a column). These cables are directly connected to the DOUT module. Since the LEDs are clued into the panel and sitting very stable, the danger for shorts is very low. The clueing and wiring was a job of about 1..2 hours Best Regards, Thorsten.

Hi Wise, I tried your code and think that it works in principle, but needs some finetuning. One error: "MIOS_TIMER_Init(4,39062)" is wrong, I guess that you want to select a 1:4 prescaler, the appr. number is 2 4 is an invalid value, but I can tell you that it selects mode 0, which means: prescaler 1:1 The program works better when the timer is running faster. I checked it with a detented ALPS STEC11 (24 pulses per revolution) and MIOS_TIMER_Init(0,15000) which is ok, but not the optimum. From my impression the physical force which is required to move the encoder from one to another detent makes a fine resolution of 6bit impossible. I also tried it with the (so called) "chinese low-cost encoders" which were organiced by Ian Hurlock some time ago, they are working much better than ALPS (in your case!!!) with MIOS_TIMER_Init(0,10000) Yes, maybe it makes sense to reduce the resolution to allow a better (and not so random-like) handling. It could also make sense to use a table which maps the counter values to the output values Best Regards, Thorsten. P.S.: debug hint: you can write: void Tick(void) { MIOS_LCD_CursorSet(0x00); MIOS_LCD_PrintHex2(counter); } [/code] to debug the counter. Tick() is the best function for such debug messages, because it is called when nothing else is to do. It's only important that this function never runs longer than ca. 300 uS to reduce the risk for a MIDI In buffer overrun Best Regards, Thorsten.

Fine! :) Touchsensor: see the schematic at the DINX4 page Best Regards, Thorsten.