TK.

Thanks for your understanding :) The big difference of MB64E compared to MB64 is, that 128 encoder events + 64 button events are stored in RAM, and the MIDI receiver can parse for these events at the same time. The MB64 can only handle 64 pots and 64 button events, therefore there is enough RAM free for features like the snap mode. If - for example - the MIDI events wouldn't be free configurable, but hard programmed (e.g. only CCs at a specific channel), the RAM consumption would be much less, and it would be a piece of cake to include an additional soft-takeover function for the pots. But MB64E was designed into another direction (high flexibility, high number of parameters). And programming an application with a different concept would basically mean twice the effort. However, maybe you should have a look into the upcoming C based MIOS application, where changes can be made much easier. E.g., there isn't an "app_defines.inc" file anymore, you can declare variables and arrays directly within a .c file, and the compiler/linker will take care for the memory allocation. This makes it easier to copy features from one into another application. Best Regards, Thorsten.

Hallo Frank, erst neulich hatte jemand ein aehnliches Problem - Kurzzusammenfassung (die Details stehen in diesem Posting: http://69.56.171.55/~midibox/forum/index.php?topic=4400.0): die Chip Select Leitungen dieses Displays sind invertiert, und dies muss MIOS mitgeteilt werden. Du muesstest also folgenden Zeilen in die USER_Init-Funktion der MB64 Applikation einbauen: ;; select LCD type #7 (custom LCD driver, hooks in app_lcd.inc will be used) movlw 0x01 ; if 0: non-inverted CS, if 1: inverted CS# movwf MIOS_PARAMETER1 movlw 0x00 ; here you could forward an additional parameter movwf MIOS_PARAMETER2 movlw 0x01 ; selecting the KS0107/KS0108 driver call MIOS_LCD_TypeSet [/code] Ja, die MB64 unterstuetzt keine graphische Anzeige, Du wirst lediglich ein paar Textzeilen sehen. In main.asm koenntest Du zumindest die Displaygroesse "4x20" einstellen, so erscheinen mehr Informationen auf dem Screen. Wenn Du mal ein wenig mit Graphik herumspielen moechtest, empfehle ich den SDCC Wrapper :) Hier muesstest Du die Initialisierung in main.c, Funktion Init() einbauen: MIOS_LCD_TypeSet(0x01, 0x01, 0x00); Gruss, Thorsten.

Hallo Martin, im Februar letzten Jahres gab es eine kleine Hardware-Aenderung in MIOS und der MIDIbox SID Applikation: die Taktleitung der Schieberegister muss nun an J10:MD angeschlossen werden. Kann es sein, dass Du das SID Modul noch nach dem alten Schaltplan verdrahtet hast? Das wuerde den von Dir beschriebenen Effekt erklaeren (ansonsten scheint ja schon eine Menge zu laufen! :)) Die Details findest Du hier: http://www.ucapps.de/mios_v1_7_hardware_mods.html Gruss, Thorsten.

The MIOS C Wrapper release is now available at the http://www.ucapps.de/mios_download.html page (sdcc_skeleton package) Note that MPASMWIN is not supported anymore, since it cannot always handle properly with the generated code of SDCC Although the SDCC team fixed these imperfections in the meantime, I cannot suggest to use the "binary snapshots" instead of the official SDCC V2.5.0 release, since they could be incompatible with the MIOS wrapper (we had this several times in the last months) So, please only use following tool versions: SDCC V2.5.0 - http://sourceforge.net/project/showfiles.php?group_id=599: sdcc-2.5.0-setup.exe GPUTILS v0.13.2 http://sourceforge.net/project/showfiles.php?group_id=41924: gputils-0.13.2-1.exe Best Regards, Thorsten.

These applications got a new patch mode feature some months ago, but I never found the time to write a documentation about the new possibilities. However, they also include some bugfixes which had been reported by users, therefore I release them "as they are": o a patch mode has been implemented which can be selected as an alternative option to the common "bank" mode in main.asm In patch mode, each BankStick contains only one bank configuration (means: pot/button/label setup), but 128 snapshots! This allows you to prepare one configuration for your MIDI gear, and to store/recall/morph up to 128 scenes. o there are two new SysEx commands for storing/restoring patches o patches and banks can be changed from external with a "Program Change" command. The channels are specified in main.asm. Default: Channel 16 - Banks; Channel 15: Patches o some minor bugfixes and improvements [/code] Best Regards, Thorsten.

MIDIbox MM V2.0 is the first application which has been completely written in C. For myself it is a prototype for the MIOS Wrapper and a nice demonstration example, how HUI protocols can be handled with MIOS. Another application which will be available soon is the ported MIDIbox LC (it's nearly finished, but I want to complete the MBSEQ hardware option before the first release - means: you will be able to use a Mackie/Logic Control emulation with your MIDIbox SEQ V2) In addition I'm planning to port MIDIO128, MIDIbox TC, MIDImon and MIDIbox CV in the next months to improve the re-usablility. There will also be a new application: MIDIbox KMK (Korg microKontrol) emulation However, here the changelog for MIDIbox MM MIDIbox MM V2.0 ~~~~~~~~~~~~~~~ o the complete application has been ported to C More infos about how to compile the code can be found under http://www.ucapps.de/mios_c.html MIDIbox MM V1.0 ~~~~~~~~~~~~~~~ Initial version which emulates the Motormix nearly completely: o eight motorfaders send and receive position data with 10 bit resolution o eight touch sensors can be connected o eight virtual pots (rotary encoders) send relative events o one additional rotary encoder ("jog wheel") sends relative events o all 67 button functions are available o all 72 LED functions are available o two status digits (7 segment digits) o messages are displayed on a 2x40 LCD (only!) Additionally this application provides following functions: o a dirty hack makes it possible to add eight LEDrings to the V-Pots. However, the host application must send the absolute V-Pot position to the display - currently it only works with Sonar, Cubase doesn't get use of the so called "rotary pointer" (see Motormix specification) o a layer function which allows to map Motormix button functions and LEDs to a smaller number of buttons (additional "Shift Key") o 128 general purpose controllers can be send with the V-Pots in GPC mode o optional "dont-send-if-touchsensor-not-pressed" function (disabled by default to avoid confusion, but heavily recommented if touchsensors available) Following function of the original Motormix is not implemented: o different graphical forms of the "rotary pointers" at the lower LCD line, since the graphics are not documented. Instead, always a vertical bar + the decimal value will be displayed Additional Notes: o Meters are not supported by the Motormix protocol o MTC digits are not supported by the Motormix protocol o LEDrings normaly not supported by the Motormix protocol. The LEDring hack works only with Sonar, and not with Cubase o the two 7 segment digits have to be connected directly to the shift registers --- no multiplexing required o only one 2x40 LCD is supported by the Motormix protocol o tested with the MIDIbox NG hardware, buttons and LEDs are assigned according to the existing design. The mapping can be changed in mm_io_tables.inc o *IMPORTANT* since I'm a Logic user (and therefore prefer the Logic Control emulation), this application has only be tested with following demo programs: o Sonar 3 Producer Demo Edition o Cubase SX Demo V1.02 No guarantee that it also works with other versions o Note for Cubase users: most buttons are not assigned to a function by default - you have to do this by hand in the device configuration! It seems that these functions don't have a LED feedback, but this could be different in newer Cubase versions. Please read also the "Remote Control Devices" Documentation which can be found at the Steinberg Website [/code] Best Regards, Thorsten.

The public release of MIOS V1.8 is finally available at the http://www.ucapps.de/mios_download.html page Changes: the interrupt handler now also saves FSR0L and FSR0H. This allows the use of C programs compiled with sdcc a wrapper for the SDCC compiler is now available in a seperate package. More informations can be found under http://www.ucapps.de/mios_c.html the motorfader driver has been overworked. The new concept works with a "dynamic deadband" and a definable PWM which will be used once the fader reaches the target position. The details are described at http://www.ucapps.de/mbhp_mf.html on a MIOS up- or download, the BankStick number is now selected with an address extension, which is part of the SysEx command the "Brown Out Reset" (BOR) settings in the configuration sector are changed automatically when MIOS detects unproper settings - the appr. cell in this sector will be overwritten and the device will be reset. New Setting: Brown Out Reset at 4.5V This ensures that the PIC won't access devices during power-off which cannot operate at voltages < 4.5V (e.g. the BankStick, which could get an unintended write command) support for PIC18F4620 (64k flash device) --- in the source code only (no precompiled binary available) Unfortunately this derivative is currently not qualified for MIOS due to a silicon bug in the EUSART peripheral (rev3) Some words to the PIC18F4620: Synapsys informed me about this new derivative some days ago (thanks Bob!) - it is a great replacement for the PIC18F452, because it is pin compatible and provides 64k flash (two times more than the 18F452), 3966 bytes RAM (> two times more than the 18F452), 4k EEPROM (four times more than the 18F452) There is already an updated Bootstrap Loader in the CVS which is located to 0xfc00. Also the mk_syx script has been updated in order to support the upper memory area (thats the reason for the new "address extension" in the SysEx upload/download command). So: this could be the future chip for the MBHP platform, because the price isn't much higher, but the benefit is -- let's say --- unbelievable! :-) It opens a lot of new possibilities for applications which ran out of memory, especially for MIDIbox SID, FM and especially SEQ BUT: the current silicon release contains the same EUSART bug which I found on the PIC18F4550 (I checked this with a rev3 sample) - It makes the chip unusable for high-performance MIDI applications! Therefore please DON'T BUY THIS CHIP yet, until Microchip has regognized that this is a serious problem, and released a new silicon version. This could take some weeks, some months, ... nobody knows. There is a workaround, but I must say that this won't be supported from my side due to development effort and performance drawbacks: by using an external USART, which is connected via SPI or IIC it should be possible to realize reliable MIDI transfers. But this can only be a "tinkering solution", it requires changes in MIOS and increases the CPU load I prefer to wait for a full functional silicon. Once available, just replace the PIC18F452 and be happy about a lot of new possibilities :) ----- Following applications have been updated: mf_calibration: due to new MF driver mf_direct_control: due to new MF driver midibox_hn: due to new MF driver midibox_lc: due to new MF driver, got also some minor bugfixes midibox_mm: due to new MF driver, has been completely rewritten in C skeleton_v1_8: due to new MIOS release sdcc_skeleton_v1_0: first C wrapper release Have fun! Best Regards, Thorsten.

Kann es sein, dass Deine MIDIbox noch mit einer alten MIOS Release laeuft? Gruss, Thorsten.

Hi Heitor, thanks for your interest. :) It is very important that you are working exactly on the current flow, otherwise you will produce a lot of additional effort at my side (please don't! :-/) Moxi: could you explain how you edit the pages in the meantime? Best Regards, Thorsten.

Why did you open a new thread to the same topic? As I already wrote, there is no memory free for snap mode - not only flash, but also RAM (snap mode requires a copy of the last pot values) Please believe me that an enhancement into this direction requires a lot of conceptional changes in the MB64E (e.g. forget the possibility to handle 128 positions in RAM) There is no 1-2-3 solution, otherwise you would find it already in the application (isn't this clear?) Best Regards, Thorsten.

How are these rubber switches normaly adapted to a keyboard controller? I've no idea how the hardware looks like, do you have a schematic? Best Regards, Thorsten.

Thanks Marcel - a lot of stuff to read! Best Regards, Thorsten.

Hi Jeff, Do you know more details why MIOS doesn't run on this chip? normaly not, you have to downstrip MIOS (e.g. remove the routines for graphical LCDs), and you have to put the "application" directly into the MIOS source in order to save memory. You can download it from the CVS - here a link to the web interface: http://cvs.sourceforge.net/viewcvs.py/mios/mios/ It is possible to use MIOS without bootloader, but this means that MIOS itself cannot be updated via MIDI, and that a PIC programmer has to be used instead. This is much less comfortable, and especially slower than updating the code via MIDI. Yes, just search for "PIC_DERIVATIVE_TYPE" in the bootstrap loader and MIOS source code These are the places which have to be adapted Best Regards, Thorsten.

Hi Adam, the output voltage of the 74HC595 is 0V..2.5V, it depends on the following analog circuit if a biased voltage is required or not. If I should ever make a PCB for the AOUT_LC module, I will prefer a biased voltage, because all filters I own are working with biased voltages, too... Best Regards, Thorsten.

Hi, yes with 2 OPL chips you have the choice of 4 audio outputs The number of external filters is limited by the number of audio outputs, the CVs for the filters can be mapped to any voice no, this has to be managed by yourself - in real life this is no limitation, up to 8 BankSticks can be connected -> makes 1024 patches. It's possible to play chords in poly mode, but in this case all 3 oscillators are routed through a single filter. The "super poly mode" which allows you to play 4 SIDs over a single MIDI channel in polyphony is still a planned feature, but with some sequencer tricks it's already possible - see http://www.ucapps.de/midibox_seq_tutorial1.html Best Regards, Thorsten.

Yes, it should. You could use MIDI-Ox or another MIDI monitor in order to check if any message is sent by the core. If this is the case, you can search for the reason in your MIOS Studio Setup Just follow each point of the MIDI Troubleshooting Guide until you find the error Best Regards, Thorsten.

Hi, I guess that you need at least a keyboard matrix with double pole switches at the hardware side, and n delay counters (n = number of keys) at the software side. This should allow to measure the time which passed when you started to press a key (upper contact gets deactivated), and reached the bottom of the key (lower contact gets activated) The main reason why you cannot find an application which supports velocity is just, that I don't own such a keyboard which I could use for experiments Best Regards, Thorsten.

these are high-frequency signals which cannot be measured with a multimeter obviously not, otherwise it wouldn't send a request message It's clear, MIOS hasn't been uploaded completely. I'm not sure why, maybe you haven't selected the right MIDI In Port in MIOS Studio, so that the upload tool doesn't receive the return message from the core. This would explain why only one block has been uploaded. So: check the MIDI configuration in MIOS Studio (you should see the request message in the MIDI IN monitor of MIOS Studio) Alternatively you could try to upload MIOS with the "Don't use feedback from core" option Keep in mind that MIOS itself has to be uploaded within 2 seconds after power-on. Without the feedback option, you could enable the "Wait for upload request" option instead. power-on the core and wait until all blocks have been uploaded --- but this feature really requires a proper MIDI connection Best Regards, Thorsten.

Evtl. die Release Zeit der VCA Huellkurve zu hoch gedreht? Gruss, Thorsten.

Es ist knapp an der Grenze, doch mit normalen LEDs (Rot/Gelb/Gruen) treten damit keine Probleme auf. Blaue LEDs sind jedoch eine Ausnahme, manche benoetigen 30 mA und mehr (Herstellerspezifikation lesen!)... hierfuer ist ein 74HC595 (ohne zusaetzlichen Treiber) sicherlich ungeeignet Gruss, Thorsten.

Hallo Markus, fuer den MIDIbox Link musst Du den Source Code nicht modifizieren, bei den meisten Applikationen laesst sich das auch vom User Interface (bzw. vom .syx Konfigurationsfile) aus einstellen. Der MIDI Merger des letzten Cores sollte auf "MBLink EP" (Endpoint) eingestellt werden, der Merger der uebrigen Module auf "MBLink FP" (Forwarding Point) Entferne Deine Modifikationen am besten wieder, die bringen den Datenfluss nur durcheinander. Gruss, Thorsten.

ja, das koenntest Du bspw. machen. Der Tristate Mode ist vor allem dann interessant, wenn die Ausgaenge mehrerer Shift Register zusammengeschaltet sind (Bus-Treiber) Ausgang QA bis QH aktiviert/deaktiviert Gruss, Thorsten.

Hallo, zu welchem Deiner Postings gehoert diese Frage? Ich fuehle mich verwirrt ;-) (Bitte die Thread Struktur einhalten) Gruss, Thorsten.

Hallo Moritz, ja, beides ist richtig. ja, das war gemeint. Bei der MIDIO128 kann man bspw. auch eine 1-zu-x Steuerung realisieren, bei der ein einziges Event (bspw. ein Program Change), mehrere ausgaenge gleichzeitig schaltet. Ja, ihr benoetigt 4 DOUTX4 Module und ein CORE Modul. Zusaetzlich ein PIC Programmiergeraet (JDM oder besser PIC_BURNER), wobei hier jedoch optional jemand aus dem Forum den Bootstrap Loader flaschen koennte (MIOS und MIDIO128 werden ueber MIDI aufgespielt) Gruss, Thorsten.

yes, no free memory for all this stuff Best Regards, Thorsten.