TK.

Dann hast Du Glueck gehabt ;-) Nun schliesse mal den BankStick an. Und zwar aus dem Grund: nach dem Formatieren (dauert > 1 Minute, wie bereits geschrieben), wird kurzer ein Ton gespielt. Und zwar direkt ohne dass der MIDI Kanal oder der Preset-Sound irgendeine Rolle spielt. Ist der BankStick erstmal formatiert, wird dieser Ton ca. 5 Sekunden nach jedem Einschalten gespielt. Ergebnis? Gruss, Thorsten.

Argl! Mach' die Bruecke weg, die erzeugt einen Kurzschluss und koennte den 74HC595 zerstoeren! :-/ Den Widerstand kannst Du weglassen, aber daran wird sich nicht viel aendern. Im Schematic steht ja "it isn't required anymore...." und "could be left out". Mit der Bruecke erzwingst Du nun, dass der SID keinen Reset mehr erhaelt, und das wird so ebenfalls nicht funktionieren. Was hat sich bei der direkten Verbindung der beiden Pins ergeben? Die Pins duerfen dabei keinen Kontakt zur Leiterplatte haben! /edit: es macht wirklich Sinn, das mal auf diese Weise zu testen, glaube mir einfach... ;-) Gruss, Thorsten.

Hallo, es ist normal, das manche SIDs (vor allem die ersten Revisionen) ziemlich heiss werden. Oder hast Du den Eindruck, dass der Chip wesentlich heisser als im C64 Gehaeuse wird? Nun, das wuerde alles erklaeren. Der Reset ist Low-Aktiv, das heisst: solange an diesem Pin keine 5V anliegen, befindet sich der SID im Reset und wird garantiert keinen Mucks von sich geben. Ich halte es fuer unwahrscheinlich, dass es am 10k Widerstand liegt (zumal es sich hier ja um einen Pull-Up handelt). Vielleicht liegt es an der Bruecke zwischen dem SID und dem 74HC595? Sie hat hoffentlich keinen Kontakt zum Oszillator-Gehaeuse? Ist es Dir gelungen, den Reset Pin mit dem SID Interconnection Test "wackeln" zu lassen? Du koenntest mal Pin #5 des SIDs und Pin #2 des 74HC595 zur Seite biegen und direkt miteinander verdrahten. Gruss, Thorsten.

A 2 HE solution would be really interesting! But an inportant tip: try to arrange the "real" components before ordering the panel! When you compare my first panel idea with the final one, you know why... ;-) It's always a good idea to prepare all boards before ordering a panel! To the button functions: EDIT: changes to the common edit view, steps are muted/unmuted with the 16 general purpose buttons. 1/2/3/4 to select the track, A/B/C to select the layer. MUTE: changes to the appr. menu page which is used to mute the tracks with the general purpose buttons. PATTERN: changes to the pattern menu. 1/2/3/4 to select the pattern, GP buttons to enter the pattern number SONG: changes to the song menu. 1/2/3/4 to select the pattern, GP buttons to enter the pattern number, A selects the song number, B: song position, C: End/Repeat/Loop points (in fact the cursor within the menu page is controlled by 1/2/3/4/A/B/C) SOLO: only the currently selected "edit" tracks will be played. With the 1/2/3/4 button you are able to select multiple tracks at once (like known from the MIDIbox SID) ALL: allows you to set a value for all selected tracks, steps and layers at the same time with a single encoder FAST: faster encoder movements - works for all encoders at the same time. If you are using encoders with inbuilt push-button, you can activate the function also from there Best Regards, Thorsten.

Some countries or language areas have their dedicated boards for exchanging informations about local electronic shops, for starting collective orders, for requesting help (i.e. PIC burning), etc. I just noticed that something similar is missing for english speaking nations. So my question: would it make sense to start at least three additional boards for Australia, UK and Canada/USA? Maybe also for english speaking asian nations? Or would it make sense to start two new international boards "collective orders" and "PIC burning" (suggestions for alternative names are welcome) Best Regards, Thorsten.

Did you try the ENC example 1? Then you should continue with ENC example 2 which handles more than 1 encoder... (it always makes sense to read the comments in main.asm ;-)) Best Regards, Thorsten.

Hi, I didn't wrote that this is a Steinberg bug, I wrote that they violated the protocol. Or in other words: either they haven't read the specification and implemented the meters in the wrong way, or they thought that they could "enhance" the protocol, but did this in a very unusual and incompatible way. The biggest problem with the Steinberg implementation: it's not fail-save. Means: there is no flag or something else which indicates the first and the second byte. This means that the protocol handler could get out of sync (e.g. during the startup phase or if the MIDI cables are temporary not connected). Something like a time-out mechanism could help here as a dirty workaround. Second problem: they don't use the meter events in the same way like Logic. The emagic guys were clever, Logic sends only the peaks, and Logic Control decreases the meter level automatically: MIDIbox LC decreases the meter levels also automatically (of course). But from your logfiles it can be assumed that Cubase sends every single LED event. This costs a lot of bandwidth (means: motorfaders won't be moved in realtime, very slow LCD messages, etc...). Third problem: I don't know why it would make sense to send LED meters with 7+4 = 11 bit resolution. Who really wants to have 2048 LEDs for a single meter? ;-) Result: the implementation is very different from the original LC. Either you've misconfigured Cubase (a user wrote some time ago that Cubase provides some kind of generic metering via MIDI - hope that you haven't enabled it, because it wouldn't be compatible with MC/LC), or a special function for Cubase has to be programmed (without the automatic decrease and with a parser for this broken protocol). Of course, not a big problem with MIOS, but who wants to spent the effort? Best Regards, Thorsten.

It's a EMC problem. Add the 100nF caps (maybe also to the LCD), and check your power supply (5V@core should still be stable when *all* motorfaders are moved) Best Regards, Thorsten.

Hi Frank, I will try to reproduce this. I never tested this code... since I don't own so much ledrings ;-) Best Regards, Thorsten.

Oh, by the way: I'm also planning to release a MIDIbox LC setup file for MIDIbox SEQ (however, without Motorfaders the handling isn't that much perfect, but it's a cheap solution for people who want to master their tracks after the sequences have been recorded :)) efficient scheduling ;-) (I'm working on the MIOS based MIDIbox SEQ application since some months, mostly not more than 2-3 hours a week) Yes, definitely. My effort for the MIDIbox projects is about 60% to answer postings and emails, 30% for implementing applications and writing documentations, and the remaining 10% to make music with all the stuff... however, this is a totally different topic, so I will stop here. ;-) In my opinion this would be perfect, also for beginners who don't know where to order all this stuff. Btw.: Arumblack - it's a 3U panel Best Regards, Thorsten.

Did you really follow the instructions? See http://www.ucapps.de/mios_bootstrap.html MIOS has to be uploaded with the primary loader, this means: within 2 seconds after power-on (before MIOS boots) Best Regards, Thorsten.

Ok, scheint wohl nicht das Problem zu sein. Bleiben noch die oben erwaehnten Tests uebrig. Gruss, Thorsten.

Fuer das Multiplexen koenntest Du Dir vielleicht mal das ENC Example 3 anschauen, aus dem Header: ; ; ENC Example, Part 3 ; ; Purpose: ; -> Sends a MIDI event when one of 64 encoders (#0-#63) is moved ; -> Prints the value on LCD ; -> updates the encoder value if CC#0 - CC#63 has been received via MIDI ; -> handles 64 LED rings ; ; For interest: ; -> variable definitions in app_defines.h ; -> ENC initialization in main.asm, USER_Init ; -> ENC pin definitions in mios_tables.inc, MIOS_ENC_PIN_TABLE ; -> single encoder handler in main.asm, USER_ENC_NotifyChange ; -> simple display handler in main.asm, USER_DISPLAY_Tick ; ; -> MIDI events which have been defined in mios_tables.inc, MIOS_MPROC_EVENT_TABLE ; -> MIDI event handler in main.asm, USER_MPROC_NotifyFoundEvent ; ; -> LED ring handler in ledrings_handler.inc (called by main.asm, USER_SR_Service_Prepare) Der Code, der Dich interessieren duerfte, steht in ledrings_handler.inc Du musst natuerlich nicht alle 64 Ringe (=768 LEDs) anschliessen, man muss sie auch nicht ringfoermig anordnen... und das Beispiel funktioniert auch mit weniger LEDs ;-) Zu den Haendlern, die den PIC18F452 fuehren (nein, der PIC18F442 ist nicht kompatibel): siehe die CORE Bestell-Liste: http://www.ucapps.de/mbhp/mbhp_core_orderlist.txt Gruss, Thorsten.

Just to clarify if this is really true (sounds like a poor implementation from Steinberg if they violate the protocol and if they really use the same events for two different informations) - could you post a logfile with the metering information of a single channel? Sweep the volume from low to high to make the coding clear to me. Dan, if you read this: did you notice the same effect with your MIDIbox LC? Best Regards, Thorsten.

I just found the part numbers: the encoder knobs are from http://www.albs.de DK13-164, black-soft DK13-164, gray-soft Each knob costs 0.65 EUR Datawheel also from http://www.albs.de DK38-144 I got this as free sample These knobs cannot be found in the (poor) webshop, I contacted Albs directly via EMail. If you have luck, you will get an immediate response. If not, it can take some days or weeks ;-) However, if more guys are interested, we could start a collective order (Albs normaly delivers only to big customers in quantities of 100 pieces and more) The buttons are from http://www.reichelt.de DT 6 <colour> <colour>: BL, GN, GR, RT, SW Each button costs 0.43 EUR Best Regards, Thorsten.

It always makes sense to use the latest MIOS version... ;-) Best Regards, Thorsten.

Mir ist noch eine weitere Fehlermoeglichkeit eingefallen: mit welchen Spannungen werden die beiden Module versorgt? Siehe auch: http://www.midibox.org/cgi-bin/yabb/YaBB.cgi?board=troubleshooting;action=display;num=1072108209 Zum Debuggen wuerde ich empfehlen, die Module mit zwei getrennten Netzteilen zu versorgen. Gruss, Thorsten.

Ja, macht trotzdem Sinn. Ueberpruefe auch mal die Spannung am Reset Pin (#5): 5V Und am RW Pin #7: 0V Vielleicht solltest Du auch nochmal den mbsid_interconnection_test durchgehen. Und wenn das nicht weiterhilft, wuerde ich vorschlagen, dass Du mir das SID und CORE Modul einfach mal per Post zuschickst ;-) Gruss, Thorsten.

Nein, wie ich schonmal erwaehnt hatte, gibt es einen default Sound. Welche Spannung misst Du an Pin #27? Sie sollte bei ca. 4-6V liegen. Hast Du die anderen Punkte ebenfalls ausprobiert? Gruss, Thorsten.

Prima, nun hast Du endlich wieder einen definierten Stand, auf den man aufbauen kann. :) Probiere nun folgendes aus: [*]nochmals ein Audio Signal an den Eingang anschliessen. Falls am Ausgang nichts rauskommt, muss sich irgendwo zwischen Pin #27 und dem Audio Out ein Kurzschluss befinden. Wie findet man das am einfachsten heraus? Einfach SID aus dem Sockel nehmen, Pin #27 leicht zur Seite biegen, SID wieder in den Sockel stecken, und zwar so, dass Pin #27 keinen Kontakt mehr zur restlichen Leiterplatte hat. Danach ueber einen 100 nF Kondensator direkt an den Verstaerker gehen. Sicherheitshalber koenntest Du auch nochmal einen 1k Widerstand in Reihe schalten (wegen der Kurzschlussgefahr usw...) [*]in sid_tick.inc, Funktion SID_Tick den Debug-Code aktivieren (ab alpha3 standardmaessig in der Release) [*]einen BankStick anschliessen, und schauen, was sich dabei aendert. Zur Erinnerung: beim ersten Einschalten dauert es ca. eine Minute, bis der BankStick formatiert ist. Im Hintergrund hoert man seltsame Piepstoene. Also: Geduld und nicht gleich wieder ausschalten. Gruss, Thorsten.

Thanks Thanks Thanks! :) More pictures are available here: http://www.midibox.org/midibox_seq_photos Btw.: picture #11 shows that this 19" case (available for 30 EUR @ Pollin) isn't so perfect for my panel. However, such fixes are well hidden ;-) and I didn't found another way for the leftmost buttons The case is almost empty, but it will be filled with the AOUT module and perhaps also an analogue synth (-> the so called "simplesizer") in the future. Best Regards, Thorsten.

Hi Robin, This is nothing for which you have to be worried about. MIOS inserts these delays between the SysEx blocks to avoid a buffer overrun at the PC side (some MIDI interfaces cannot handle that much data). Another advantage is, that you are able to record the SysEx blocks with a sequencer, you can store them in a midi file and playback to the MIOS core w/o taking care for the required delays between the write cycles. But this has nothing to do with your application, since you have to write your own SysEx sending routine anyhow (which isn't so difficult). A BankStick read is always much faster than the transfer of a single MIDI byte. If you are searching for some numbers: reading a single byte from BankStick takes about 100 uS in worst case (mostly ca. 50 uS). The transfer of a MIDI byte takes exactly 320 uS. Everything is in your hands, so it doesn't alter the SysEx implementation. Here a small example: ;; sending a block of 11020 bytes which is stored at ;; address 0xa123 of the first BankStick ;; select first BankStick movlw 0x00 call MIOS_BANKSTICK_CtrlSet ;; set start address to 0xa123 movlw 0xa123 & 0xff ; (low-byte) movwf MIOS_PARAMETER1 movlw 0xa123 >> 8 ; (high-byte) movwf MIOS_PARAMETER2 ;; set loop counter, using TMP[12] movlw 11020 & 0xff ; (low-byte) movwf TMP1 movlw 11020 >> 8 ; (high-byte) movwf TMP2 ;; call the generic SYSEX_SendBlock routine which ;; can be found below call SYSEX_SendBlock ;; and exit subroutine return ;; generic SysEx send routine ;; Expects: ;; o BankStick address in MIOS_PARAMETER[12] ;; o number of bytes which should be sent in TMP[12] ;; o BankStick should be selected before with MIOS_BANKSTICK_CtrlSet SYSEX_SendBlock SYSEX_SendBlockLoop ;; since this routine runs longer than 1 second, the watchdog should ;; be serviced to avoid a reset --- use the clrwdt instruction only ;; on extreme long loops! clrwdt ;; read from BankStick call MIOS_BANKSTICK_Read ;; send value call MIOS_MIDI_TxBufferPut ;; decrement 16-bit loop counter, loop until counter reaches zero decf TMP1, F skpc decf TMP2, F bc SYSEX_SendBlockLoop ;; thats all return This routine will take 320 uS * 11020 = ca. 3.5 seconds, limited by the MIDI transfer rate. You could call this routine from a button handler... Best Regards, Thorsten.

I will move this posting to the MIOS board since this is not the right place for such questions

Das CS ist immer noch enabled. Man merkt das schon an der Anzahl der Acknowledge Meldungen. Die Firmware ohne CS ist wesentlich kleiner. Wie schauts mit dem Filedatum aus? Gruss, Thorsten.

Ok, das CS_ENABLED flag scheint jedoch noch auf 1 zu stehen. Die Program Change Befehle werden nur vom Control Surface gesendet. Checke nochmal das Datum des .hex und .syx Files Ist Dein MIDI-Interface Multiclient-faehig? Falls nicht, alle Programme schliessen, die auf den gleichen Port zugreifen. Am besten aktivierst Du einfach das virtuelle Keyboard von MIDI-Ox und drueckst auf die "q" taste (spielt ein C-3) Gruss, Thorsten.