TK.

Yes, I'm using my V2 control surface as well. I've added a second 19" box for a button/LED matrix and 4 group select buttons, but this is only a freaky "nice-to-have" extension and not required to access all available functions. I used my sequencer 6 months without it, before I built it just for fun, and not because it's a "must-have". No, they will be formatted during the first startup with v3 - the data structure has been changed for 32 step tracks and more control informations. A v2 pattern compatibility is not provided by the firmware itself, but in theory patterns could be converted with a special program So, in fact you only need to replace the PIC18F452 by a PIC18F4620, and you will already be able to use the new firmware. Hint: some PIC18F4620 have been recently offered for cheap money in the Fleamarket section by people, who bought it for their MBSID V2 (but which will need a PIC18F4685 in future) In addition to the new PIC, at least one MBHP_IIC_MIDI module is useful. Details are listed here: http://www.midibox.org/dokuwiki/doku.php?id=mios_pic18f4620&s=pic18f4620 Just consider, that MBSEQ V2 has been programmed 4 years ago... since this time I collected ideas of users and myself, which couldn't be integrated into the small PIC18F452 anymore, but which would be nice to have once bigger PICs are available. I also worked on solutions how to overcome known problems and limitations. To say it short: the MBSEQ V3 firmware is a work of 10 months. It's not just a small upgrade, you will experience a completely different machine! :) Best Regards, Thorsten.

It is a known problem with MBSEQ V2, and in fact there are much more inconsistencies. The whole engine has been overworked in MBSEQ V3 based on the experiences I made with V2, so it's really worth a try! Best Regards, Thorsten.

Ueberpruefe mal die Versorgungsspannungen an den 74HC14 (siehe Schaltplan), evtl. ist Vss oder Vdd irgendwo nicht angeschlossen (Loetstellen ueberpruefen), und die Chips/LEDs werden ueber den Druckerport versorgt - das wuerde fuer ein schwaches aufleuchten der LEDs ausreichen, nicht aber zum Brennen eines PICs Gruss, Thorsten.

MBSEQ V3 provides four different approaches. 1) static randomisation within the utility menu: Just press the button below "Rand" and a new random pattern with random notes/velocity/length/gates will be generated Press and hold the random button and turn one of the rotary encoders to increase/descrease the randomness (+/- x) 2) "humanisation": Adds non-static randomness to Note/Velocity/Length, the intensity (+/- x) can be selected as well 3 and 4): random Gate and Value for specific steps: Part of the trigger layers - allows you to trigger the gate and/or to play the Note/Velocity/Length value randomly on steps where this function is activated Min/Max values are nothing else than defining the "mid" value within the layer, and adding the selected "intensity" (+/- x) to this value Different Note Scales are provided for "harmonic randomness" When I'm speaking about Notes, I mean MIDI events - it could also be a CC or it could be a chord of up to 4 notes. And don't forget: there is a loopback port which allows you to control any sequencer parameter randomly - so, not only the step events, but every parameter which is listed in the CC implementation table Just think about random morphs between two patterns, random progression parameters (jump back and forth), random clock dividers... So, of course, there is always something which somebody has seen in another sequencer, and what he would like to see in MBSEQ V3 as well, as requesting features - and especially asking "is it possible..." is so easy. But from my oppinion, nobody should do this before he has tested the already existing features - there are a lot of things to explore :) Best Regards, Thorsten.

I read your thread again, just to be sure, but I haven't found anything, which isn't already possible with MBSEQ V3, or did I miss something? ;) Best Regards, Thorsten.

Hi Rio, it took some time to find out, how to reproduce this, because you missed to mention, that this only happens when the paste target position is not the first step (or in other words: when a part of the pattern is copied to another position) Next bug, please! Gruss, Thorsten.

There is a new build #72 which got a bugfix in the paste function (triggers were overwritten when paste target was not the first step) Best Regards, Thorsten.

No - for me the existing randomization possibilities are sufficient Best Regards, Thorsten.

Ja, sieht nach einem Designfehler aus. 100 nF werden fuer 3.3V chips hergenommen, ich haette das Datasheet sorgfaeltiger lesen sollen :-/ Danke fuer den Hinweis! Werde den Schaltplan demnaechst updaten Gruss, Thorsten.

Miss mal die Spannungen direkt am Druckerport. Im Schaltplan habe ich die Signalleitungen mit "Clock", "Data In", "Data Out", "Vdd enable" und "Vpp enable" bezeichnet. Diese Signale muessten sich mit P18 zwischen 5V und 0V hin- und herschalten lassen. Kannst Du das bestaetigen? Gruss, Thorsten.

Hi, do you see a gray/black line when the LCD is powered and not initialized? This would indicate, that the contrast voltage is ok. If not, this Wiki page helps: http://www.midibox.org/dokuwiki/doku.php?id=troubleshooting_lcd_displays Best Regards, Thorsten.

The difference between 120 BPM and 119 BPM is 136 uS (@24ppqn) It's a systematic problem, that such a jitter cannot be compensated with MIDI, as sending one byte takes 320 uS In addition, your MIDI interface (USB?) will add some additional delay and jitter, and this is normaly much higher than the accuracy of a MIDIbox. See also the MIDI benchmark: http://www.midibox.org/forum/index.php?topic=2342.0 So, the only solution is to use Logic as MIDI clock master Best Regards, Thorsten.

Hi RM, This is already possible in the Length menu. :) You can even control the loop and end point via CC, and it can be controlled from another track (internal loopback) for very organic sequences. Yes, this was already possible with MBSID V2, a step can trigger an event up to 4 times with a delay of 1..31 microticks between each trigger. This can be defined for each step seperately in the length layer Best Regards, Thorsten.

Thank you very much for the research! I've updated the OPL3 PSU schematic, so that it only uses a single bridge rectifier. I think that I never found this issue by myself, since I'm using transformers with seperate secondaries - and I'm not sure, if I connected the two middle AC lines together. I will check this once I open the case of my MBFM again. The PSU of JD Sleep looks like a good alternative solution for people, who have a single secondary transformer lying around which they want to re-use for the AOUT or OPL3 module. A wiki page about this topic would be great! I would add a link to this page from ucapps.de, so that everybody can read more about different variations. Freeware schematic drawing tool: I'm using xcircuit since more than 10 years now. I don't know if there are better tools in the meantime, but this one works under different operating systems, which is a big advantage. Best Regards, Thorsten.

Yes, it was about 300 bytes for one pattern. A complete pattern backup is not provided by MBSEQ itself, this would require a SysEx tool like TL's patch manager, but nobody has implemented it for MBSEQ. Songs cannot be dumped. If you just want to create a backup of the BankStick, then you could use the MIOS functions for this - see http://www.ucapps.de/mios/mios_backup.txt It's the fastest way... Best Regards, Thorsten.

Please note that the assign layer of the OSC section is a *must* for MBSID V2, as it contains the free assignable "knob" functions - each patch will have its own assignments which allow interactive sound control of specific parameters which are relevant for "cool sweeps" Without this layer, you can only control these "knobs" from an external MIDI controller, or with your PC Best Regards, Thorsten.

Hi, in MBSEQ V3, the size is 651 bytes. Just press the GP buttons below ">>SEND<<" within the SysEx menu Best Regards, Thorsten.

A hold function is not available for MBSID V1 (as it requires some extra memory), but will be available in MBSID V2 Best Regards, Thorsten.

Episode #1: the reincarnation of a DIY synth :) -> mbsidv2_demo1.mp3 Best Regards, Thorsten.

Hi Reiner, unfortunately the repeat rate is hardcoded into MIOS (it's 32 mS). I will make it configurable with the next release Best Regards, Thorsten.

Rio: I will check this and release a new version if required Stryd: ok, I'm a young guy, I can wait ;) Best Regards, Thorsten.

Schoen, dass ihr die Sache unter Euch ausgemacht habt! dj3nk: ich werde nun erstmal von einem Bann aus dem Forum absehen. Es ist verstaendlich, dass Du momentan im Stress bist, aber die Umgangsform gegenueber anderen sollte darunter nicht leiden, egal ob Du sie persoenlich kennst, oder ob sie am anderen Ende der Welt sitzen. Gerade im anonymen Internet kann das sehr schnell in eine ungewollte Richtung eskalieren, wie man in diesem Thread nachlesen kann, und davon hat niemand was. Gruss, Thorsten.

20mS - this sounds like the ADSR bug, see also: http://www.ucapps.de/howto_sid_wavetables_1.html at the end of the page Best Regards, Thorsten.

So, if the RC and SC signal is available, the only possibility I could imagine is that the 74HC165s are not powered (bad soldering at pin 8, 15 or 16). Could you please check this with removed chips? Because otherwise it could happen that you measure an "output voltage" caused by internal nets instead of the "input voltage" from the power supply. Best Regards, Thorsten.

And here the next monster MIDIbox :) Dietmar has added MIDIbox LC based LCDs, LED digits and LED meters to his Tascam US2400 controllers in order to get a visualisation of parameters and values. There are some buttons in addition which allow to trigger functions, which are not accessible from the Tascam Controller.