Alain6870

Thank you for the advice!! By actually comparing both source code of the MB64 and MBCV application I could see that the Aout.inc was implemented differently in the MBCV program. With some code adjustment I managed to include the Aout "module" that can handle the 3 types of Aout Hardware in the MB64 program and I don't use the original Aout.inc module anymore. the new MB64 program compiles with no error but I haven't checked the hex file in my PIC. I'll report later on that. What I have to look at also is if the Amux.inc file uses the same variable than those sent by the generic Aout file of the new module. (I saw some different variables in the App-define of both codes that makes me believe it won't work as quickly...) Does someboy know if this statement is right?? Anyway, a lot to do to finish my interface: - building an Aout_NG board, - building a new SHx8 board interface with CMOS switches and analog S&H cells, Kind Regards,

I would like to use a MIDIBOX64 to drive 50 parameters of an analog synth (VCF Fre, reso, VCA, LFO, etc...) and I only have an AOUT_NG Board to generate the analog voltages. As I checked the source file of the MB64 program, I saw that it only supports an AOUT board with 2 MAX525 daisychained to the core. What can I change in the prog of the MB64 to be able to adress an AOUT_NG board? PS: I want to use the MB64 facilities such as patch storing, handling AMUX to have up to 64 different CVs and this is the reason why I want to use MB64 instead of MBCV program. Has someone already done this? or is there an other way to get what I need? Thank you for your kind help,

Hello evreybody at ucapps, Thank you to those who took the time to read my previous message. I found out that my problem lied on the fact that I had my table between 2 labels of my JUMPTABLE and there were too many lines between the rgoto instruction and the labels and this generated this adressing error! I just wrote my table at the end of my ".inc" file and everything is now ok. ; branch depending on PARAMETER NUMBER (0 Ã 127) movf TMP1, W andlw 0x7f JUMPTABLE_2BYTES_UNSECURE rgoto Para_typ_1 rgoto Para_typ_2 ... rgoto Para_typ_128 Para_typ_1 ..x lines of code... Para_typ_2 >>>> error [126] Argument out of range (1133 not between -1024 and 1023) because GPASM doesn't find the flag... Best Regards, Alain

Hello evreybody at ucapps, I'm trying to program a software based on "midimon" for helping me to read parameters of an Evolver synth on a display. For this I read the sysex message from the midi output that the instrument sends when an encoder is turned. When I get the parameter value, I look in an indexed table to print a corresponding string. The problem is that I use large tables and it seems that I get overflow message when compiling the prog. One of my tables is made of 75 entries of 16 bytes strings: mullw 16 TABLE_ADDR DEST_NAME_TABLE movf PRODL, W addwf TBLPTRL, F movf PRODH, W addwfc TBLPTRH, F movlw 16 ; (16 chars) goto MIOS_LCD_PrintPreconfString DEST_NAME_TABLE db "No destination " db "OSC 1 Frequency" db "OSC 2 Frequency" db "OSC 3 Frequency" db "OSC 4 Frequency"... and so on 75x and when I compile I get this message: ./include/asm/macros.h:172:Error [126] Argument out of range (1133 not between -1024 and 1023) Thinking this table was too long I splitted it into two tables of 38 entries each, but as I compile the program with only one table I have not errors, and when I add the second one I get the same errors, it seems to me that I reached a memory limitation of the PIC. Am I right? I'm really a newbee in programming PIC in assembler, could someone give me a hint for overcoming this problem? Is there a specifical allocation possible for this tables? PS I use a PIC 18F452 and I don't use the Tracememory block of the original midimon. thank you for your help, Best Regards, Alain

Hello every MB builder and user, I was looking for a midi controller that could send SYSEX messages like this one: F0 (sysex start) 01 (factory brand ID) 20 (synth type or ID) 01 (file version) 01 (program parameter) xx (parameter number) varies fom 0 to 127 (in decimal) but is dedicated to one control ie one pot for instance yy (parameter value = LSB nibble) zz (parameter value = MSB nibble) F7 (sysex end) Is MB64 soft only handling CC events? or has someone developped a special soft or routine to help me? By the way is someone knowing if it would be possible to handle checksum too in that kind of sysex message? Thank you very much for your gentle help

Hello Thorsten, First of all I want to apologize for the misspelling of your name, in my last mail I forgot the h after the t. This simple ommission is due to the fact that I know some Torsten too, as you know in Germany both are usual with and without the h. But let's go to the subject... I tested the new version of the soft and it rocks. V1.1 is ok with the added menu for step length in song mode. it's exactly what i wanted. Thank you for your very quick update. I noticed a minor "bug" in the display of the timebase menu. Whe you display the string "Sync to bar:" and then go back to an other menu that displays the string "Selection: " you have as a result the message "Selection:r". So I changed the length of the string "Selection" in the source Cs_m_trkdiv.inc and now everything is ok. I presume you could change this with the next update. Grüsse, Best Regards,

Hello everybody, As promised I send 2 pics of my MB808: The case is out of wood. Next step will be painting of the MDF surfaces. it will be maybe white then. IMG_3232.JPG IMG_3233.JPG IMG_3232.JPG IMG_3233.JPG

Hello Torsten, Yes your suggestion would fullfill my expectations in most cases. For instance when I have a song written in 3/4 I can programm the patterns or the tracks in every timebase I want and finally indicate the sequencer that for this song chaning has to occur every 12 of 16th steps. Even further ,it would be cool to have a dedicated timebase for the pattern chaning in song mode, it could be then more accurate. IE: Song Timebase 8 and chaning_steps 7 for instance. For my other problem with the fill_in pattern I could let some rest to adapt the pattern to the overall duration. kind regards,

Hello, Sorry, I'm confused I wasn't clear on my first post... :-[ I have no regulator on the board, and only one regulator after the tori and the rectifier. Now I had my MB808 working all the day without a problem. It seems that the "reset default" was due to a bad calibration of the luminence of the LCD display, as I explained on my previous post ,causing, I presume, a breakdown on the pic. Thanks to everybody for your involvement. Explenations are welcome as I have only presumptions no evidences.

Hello Torsten and the other MB gurus, Thank you for your piece of Software. It is so cool... I've finished my MB808 last night and could play with it for some hours. Except a small Hardware problem with my leds everything is ok and it was only pleasure: tchak tchak boom! I noticed when playing in song mode with a 32' scale and a 12 steps length (for all the pattern) that this pattern was then repeated untill 1 measure was reach before it jumped to the next pattern. Is this intend or did I miss something? Maybe the steps are always given in 16' scale (quarter note), I don't know! What I explained is a little bit embarrassing when you want to play a fill in for a song and it doesnt exactly match one measure or for special rhythmic scale ie (7/8 beat etc...). Sorry for being not so clear. Let me know if what I wrote is really not understandable, and thank you for your patience.

Hello guys, Thank you for your support! :D In the mean time I did some measures and could find out that I am drawing a current of 330 mA with only some leds on and it goes up to 430 mA with all the 16 leds on. This is normally no problem for the 78MCT05 I have. (And it stays really cool so what?). I measured the voltage in both cases an could see that with one led on I had a voltage of 5.03 V and with all the leds on I could measure 4.97 V. (Is this the problem I don't know?) I decided to reduce the luminance of my Display (I forgot to tell you I had one in my previous message) to get less current drawn, because most of the power is used for the LCD (about 200mA at full luminance). And now my MB808 ist in working order for more than 3 hours and I don't experience the problem any more. Question: is the LCD capable of creating a current cunsumption that generates a micro breakdown in the voltage supply of the PIC? Is the BC337 ok for driving the display or is there a more adequate transistor? Promise: I take some pictures when my TR... sorry MB808 is finished.

Hello everybody, It's my first post in this forum, I hope I did it rigth ? ??? I have recently build a MB808 with my own PCB for the voice board and PCBs from smach TV for the Midibox. (1x core, 2x DOUT, 2x AIN : Very good quality). Everything is working perfectly, except for the loop menu or the spare page. I think the problem is when all may 16 leds from the step button lit together, it makes me from time to time reset the PIC. I use only low power leds and a external 5V pover supply made around a 7805, but the current seems not to be a problem. Could somebody give me a hint or explain me were my problem could come from? I am a little bit dry! Thank you for your help.