TK.

Update: I was able to reproduce the MIDI Overrun problem with a direct feedback (MIDI Out->MIDI In). Normaly this shouldn't hurt. I will try to find the root cause tomorrow. It isn't a performance problem, but a bug somewhere in the firmware (I hope that only MIDIbox MF is affected ;-)). Best Regards, Thorsten.

Hi Damien, I don't plan to use other LCDs for my own MIDIboxes, but the "Display Plugins" will allow to customize the MIDIbox NG for other display devices: bigger dotmatrix LCDs, graphic LCDs, LED displays --- programming skills are required for such adaptions. Another thing is the display sharing issue. I hope that with the MIDIbox Link (which will also be available for PIC16F) only one LCD will be necessary. Best Regards, Thorsten.

Find the changes here: http://www.ucapps.de/midibox16e_changelog.html And the new tutorial here: http://www.ucapps.de/midibox16e_tutorial_en.html Best Regards, Thorsten.

Initially I planned to nominate only one well-made MIDIbox design per week, but this time it's really difficult.. in the last days I got photos from three outstanding projects which are the dream of everything from us :) So here they are - just choose your private favourite: MIDIbox64 with integrated TB303 synthesizer by Tom303: MIDIbox64 as "typographical synthesizer" by Rob: Multi-MIDIbox environment with a lot of motorfaders/encoders/buttons by Axel: Best Regards, Thorsten.

Hi Andre, normaly there isn't a buzzing sound - just download the schematic from the MBHP SID page and follow the tracks. Check especially the polarity of the rectifier and the electrolytic capacitors! I've no experiences with the MultiPic burner - are you able to verify the flash content? On this way you can ensure that the programming was successfull. Therafter you can do the following test to check if the program is running: start MIDI-Ox and send a program change to the MIDIbox SID: it should response with a lot of controller events (the sound parameters) Best Regards, Thorsten.

Hi Christoffer, the second AINX4 has to be connected to exactly the same pins like the first AINX4, expect for the analog outputs of the AINX4 module - they have to be connected to J5:A4-J5:A7 a connection diagram can be found here: http://www.ucapps.de/mbhp/mbhp_ainx4_64pots.pdf Best Regards, Thorsten.

first check the diodes, if they aren't connected in the right direction, the fader will not move. I think that this pictures explains more than 1000 words ;-) And check the voltage behind the LM317 during the motor is running. It should be stable and shoudln't go down under load. Best Regards, Thorsten.

I cannot recomment it, since the current drain of a LED (ca. 20 mA) is too high for the JDM circuit. It would only work with an external power supply for the LED + a transistor as electronic switch, but do you really want to spent so much effort just for a LED? Best Regards, Thorsten.

Hi Ian, since the current MIDIbox MF design is a realtime system with a defined update cycle, the faders won't be moved slower or faster... but possibly some features will be provided in the next years which are really usefull for the daily work and cannot be implemented with a PIC16F anymore (especially RAM consuming features) Best Regards, Thorsten.

Does the rest of the circuit work correctly? Does your computer receive any MIDI data? Did you adjust the contrast pot so that you see anything on the display after poweron? Best Regards, thorsten.

yes - and with an additional shift register for the DIN, you can control all LEDs with 32 buttons. I will add a more detailed description about the button/LED assignments in the next days. Best Regards, Thorsten.

Hallo Malte, ueber Pin 2 und 3 werden normalerweise keine Pulse gesendet - das sind analoge Eingaenge. Wenn Du schon ein Oszi zur Verfuegung hast (ich hoffe, dass hier nicht der Eindruck entsteht, man koenne die MIDIbox nur mit teuren Geraeten debuggen ;-)) dann schaue Dir mal den XTAL1 an - an diesem Pin muss ein 20 MHz Sinus zu sehen sein. Interessant ist auch die Enable-Leitung zum LCD: die muss waehrend des Einschaltens wackeln. Mit einem PIC-Programmer koenntest Du auch den Flash-Inhalt ueberpruefen. Dazu musst Du lediglich die Firmware, die sich im PIC befindet/n (sollte) in IC-Prog (o.ae.) einladen und den Verify-Button anklicken. Gruss, Thorsten.

Es geht noch einfacher: Du kannst den Tx-Pin des PICs auch direkt an den Rx-Eingang des GamePorts anschliessen --- ohne Optokoppler dazwischen. Zusaetzlich muessen dann auch noch die Masseleitungen miteinander verbunden werden. Im MIDIbox Plus Hardware HowTo findest Du dazu einen Schaltplan (demnaechst wird es auch eine eigene Seite geben, auf der alle Moeglichkeiten aufgelistet sind) Gruss, Thorsten.

Zusammenfassung: o to-COM spielt mit Deiner Windows XP installation nicht zusammen o MIDI-Out Deines PCs/MIDI-In der MIDIbox funktionieren o MIDI-Out der MIDIbox/MIDI-In Deines PCs funktionieren nicht Einfachster Test: schalte den MIDI-Merger ein, die to-COM Option aus, schicke mit MIDI-Ox eine Note zur MIDIbox: Die Note sollte genauso zurueckkommen, im MIDI-Ox-Fenster sollte alles, was zur MIDIbox gesendet wird, doppelt angezeigt werden (einmal als internes, einmal als externes Event). Falls MIDI-Ox stattdessen ein Pitch Bender Event anzeigt, kann der Fehler eigentlich nur noch am MIDI-In Deines PCs liegen. Der CNY17 ist ein wenig problematisch, mit einem 6N138 oder PC900 wuerde der MIDI-Eingang sicherlich zuverlaessiger funktionieren. Probiere mal verschiedene Pull-Up und Vorwiderstaende aus, so wie unter http://www.ucapps.de/howtodebug.html beschrieben aus (was fuer die MIDIbox Plus gilt, ist natuerlich fuer jede andere Schaltung mit dem CNY17 gueltig) Gruss, Thorsten.

Hi Frank, I never have tested the MIDIbox MF with less than 8 faders, maybe the timer interrupt consumes more cycles in that case and therefore causes a MIDI overrun. I will check this when I'm home again and evtl. release a fixed version. To 1): this error occurs when the MIDIbox receives a MIDI byte before the last byte has been processed. In my interrupt routines I have to ensure that this case never happens - I will check it. To 2): The feedback elimination routines are working on the following way: when a fader is moved manually, all incoming events which are assigned to that fader will be ignored for 250 mS - this ensures that the motor doesn't start to run during the manual move. For the other direction: when a fader has been moved by the motor, no MIDI event will be sent for 250 mS - on this way a feedback (MIDI program -> MIDIbox MF -> MIDI program -> MIDIbox MF -> ...) will be prevented. Best Regards, Thorsten.

Hi Frank, yes :D you must know that the LM317 has a different pinout compared to the LM78xx. Take a look into the PCB layout, where you can see, which component has to be connected to which pin. Random values: you should take care for the connections between the analog inputs of the PIC and the faders. The cables should be very short (less than 30 cm) and shielded to ground. Otherwise the analog signals could be affected by the motor driver signals. Best Regards, Thorsten.

Yes, the core will work (it's 1:1 compatible) --- but the 20 MHz crystal has to be replaced by a 10 MHz type, since the PIC18F452 contains an internal PLL which quadruples the clock rate. Best Regards, Thorsten.

Hallo Stefan, dieses Fehlerbild ist mir voellig neu. Wenn die Baudrate falsch waere, muesste die MIDIbox z.B zufaellige Events senden. Pitch Bender Events sollten nur dann versendet werden, wenn sie auch einem Poti/Button zugewiesen wurden.... Der Fehler muss irgendwo ganz anders liegen... Kannst Du mir mal mehr Input geben? Welche Events siehst Du auf dem Display (einfach mal durch die verschiedenen Display Modes blaettern). Welche MIDI-Schnittstelle haengt bei Dir am PC, wie verhaelt es sich mit der to-COM Option und dem Roland Driver? Gruss, Thorsten.

Tja, so ist das mit Windows... :-/ Verzeihe die dumme Frage, aber hast Du im Merger Menue auch die richtige Interface-Option eingestellt? Wenn Du die normale MIDI-Schnittstelle verwendest, muss to-COM deaktiviert sein, ansonsten passt die Baudrate nicht (passt zum Fehlerbild). Gruss, Thorsten.

Jep, das passt schon. Du muesstest dann halt noch im Source code in led.inc die LED-Register invertieren ('movf LED_x, F' durch 'comf LED_x, F' ersetzen). Am besten gleich mit v0.34, fuer aeltere Versionen uebernehme ich keine Gewaehr (ich weiss ja, wie ungerne Du updatest ;-) Hast Du mittlerweile eigentlich schonmal die neue MIDIbox MF beta ausprobiert? Gruss, Thorsten.

Ok, this explains everything. Before trying various resistors, you should check if the optocoupler circuit is really correctly connected to the Rx pin. You could also try the second optocoupler which is currently connected with RA4. Best Regards, Thorsten.

Great to hear that it finally works :) Best Regards, Thorsten.

It's available at Farnell, see the http://www.ucapps.de/mbhp_bankstick.html page for the order number. Best Regards, Thorsten.

If you have to switch between the modes very often, I could compile a special version which allows to select the mode with the spare pin (J14 of the core module). In this case, you have to connect a 10k pull-up resistor between J14 and +5V and can select the to-COM option by using a jumper to ground. I have to test this version, because I don't know if it already works correctly. You will get it next week (I'm not at home and currently have no access to my source files). Best Regards, Thorsten.

Yes, I forgot to add this to the list (fixed) It would be great if somebody could evaluate the quality of this programer, just to save my effort. I already own the PICstart kit by Microchip and therefore have no problems to program the newer PICs... Best Regards, Thorsten.