TK.

I like the idea, but I would vote for small prices for each category, instead of giving me money. The fact, that this will result into an interesting contest with innovative and fancy ideas, improved documentation, happy community gives me enough. :) As one of the prices I would contribute 4 CORE and 4 SID modules as can be seen in this picture: http://www.midibox.org/midibox_sid_blm_photos/05.jpg They are second-hand, but well tested over a long "burn in" period of 4 years. They were housed in a non-dusty C64 case during this time. Soldered with leaded tin, best quality components, prepared for optimized PSU circuit and MBSID V2. Mounted on a pizza box (doesn't smell anymore). And the best: they are blotted with original fingerprints of TK! ;-) SIDs and PICs, DIN/DOUT, IIC modules are not included! As it doesn't make much sense to give this stuff to somebody who already owns a MBSID, I would like to see it as a price for the best user application. The nominations could be done in a seperate forum section. Best Regards, Thorsten.

Maybe it makes sense to use a special low-drop regulator for such a project. There should be good schematics available in the web, just google for "low-drop 5v schematic" Best Regards, Thorsten.

Ok, I've added the two last features - now it's design freeze, next feature request won't be taken into account for v3.0 Exactly - your old patterns should still run like before, but it also makes sense to tweak on the parameters during live playing and to check, if the sequencer behaves like expected. All the event modes are not tested good enough, especially drum mode. I mostly use Note/Vel/Len, Chord2/Vel/Len and CC/CC/CC by myself. All event modes have to be tested with different track modes (Normal, Transpose, Arpeggiator), Scales, Directions, Progression Parameters, Trigger assignments, MIDI output ports and channels, Track Lengths, Loop points, Track chains, Groove modes, Random (Humanize) modes, Morphings, Clock divider settings, and all what I forgot to mention here (which is normaly the stuff which won't work, maybe I overlooked a certain dependency ;-)) I'm also not sure if the Song mode changes are really working reliable, and all the new stuff which has been quickly built into the firmware on John's requests... Best Regards, Thorsten.

A common design technique is to use 100 nF and elcaps ( electrolytic capacitor, directional caps) in parallel. It doesn't matter if these are 10uF, 15uF, 22uF, ... Yes, the requirements for RGB LEDs is too high, this would require additional source and sink drivers, and a software variation which inverts the outputs. You need the additional DIN/DOUT modules which are listed under http://www.ucapps.de/midibox_seq_v2.html (the hardware requirements for MBSEQ V2 are also valid for MBSEQ V3) Best Regards, Thorsten.

Ok, build #71 contains two last features for v3.0: o there is now a BankStick format menu accessible from the main menu. It allows you to re-initialise a BankStick, and to check the selected Type/Size and availability (n/a print when BankStick not connected) There is no quick-access option for this menu, so that this menu isn't easily accessible during a live session. Just change to the main menu and search for "Format BankSt." [/code] [code] o The Copy/Paste/Clr function now also works on the track configuration Just only the MIDI channel and Port will never be overwritten for your own comfort! o The Copy/Paste/Clr function now also works on the track configuration Just only the MIDI channel and Port will never be overwritten for your own comfort!

Yes, MB64E is the "better" MIDIbox controller. Especially because it can handle 128 pot/encoder events and 64 buttons within one bank. :) My website doesn't make this very clear, I would like to see a discussion about the MIDIbox variations in the Wiki. Best Regards, Thorsten.

Hi JD, where did you read the info about J6? I would like to remove it from my website, with the MIOS variant this solution is obsolete. Best Regards, Thorsten.

Note that there is also a "global channel", which overrides all MIDI Channel definitions for the pots and buttons. It can be enabled in the appr. menu page Best Regards, Thorsten.

Great work! (moved to the MIDIbox of the Week section :)) Best Regards, Thorsten.

The 303 mode is the "bassline" engine, but much better than in MBSID V1, especially the sequence editing should be improved :) Best Regards, Thorsten.

Hi Gert, no, MIDIO128 doesn't support BankSticks, it's a very dumb application which only allows a single configuration. For MIDIO128, MIDIbox Link can only be enabled within the initialisation file (download the mk_midio128_syx package) The error message is very strange, I never saw this before. What is "s/Perl/site/lib"? Which perl version are you using? > 4) When i'm using different banks i can assign the same button to different Midi notes. Am I correct?? for MB64/MB64E: yes Best Regards, Thorsten.

Code Generators are very common in SW development today, but the typical problem is the resulting code size and the bad performance, as the implementation has to be "more generic". Thats one of the reasons, why modern software is sometimes so slow Sometimes it's also difficult for the designers to add special features, just because the GUI which they have used to generate the code doesn't support a certain function. Without a team of specialists which permanently works together, it's better not to start with such an approach for MIOS applications. Best Regards, Thorsten.

I had a conversation with Wilba in the last days about the requirements for an alternative frontpanel, which takes the features of MBSID V2 into account. The result can be found in the Wiki: http://www.midibox.org/dokuwiki/doku.php?id=wilba_mb_6582 Hope it helps. It's really very very very difficult for me to finalize the specs for an enhanced frontpanel, when the project is still in the starting phase. :-/ Best Regards, Thorsten.

From your postings it's difficult for me to find out, which ones of the previous questions are already answered. First of all I want to highlight, that the DUO-LED solution was intended as some nice fun for myself, I never thought that anybody else really wants to spend the same soldering effort (it took some days). After I got unexpected requests from users, who wanted to build the same (unbelievable in my eyes ;-)) I created this schematic in the hope, that somebody else creates a PCB layout to simplify the construction. And this is the current state - informations are published, I've stopped my own activities in this area in the hope that somebody else continues with this extension, and writes a proper documentation I soldered the caps at the bottom side of my breadboards They are very important because of the high power consumption of the LEDs. At least 10 uF should be added in addition to the 100 nF for each 74HC595 As you can see in this picture: http://www.midibox.org/midibox_seq_blm_ext/07.jpg, the 10k resistors for the 74HC165 can be seen at the top left corner. There are 16 x 150 Ohm resistors for two 74HC595, and the third 74HC595 has no additional resistor, as this one is not required (LEDs are connected from one DOUT pin to another DOUT pin, and they are not between DOUT and ground) - an additional resistor at the third 74HC595 would only limit the voltage. I think the best solution would be, if somebody would create a PCB layout for the whole circuit. Best Regards, Thorsten. P.S.: I got my Duo-LEDs from Reichelt: LED 5 RG-3

Build #70 is now available for download. It can be considered as release candidate for v3.0 This means: no new features will be added for v3.0, only bugfixes Since I made a lot of small changes at different places in the firmware, there is a certain chance that something is not working anymore. I will see this in the next days - beta testers: please help! :) From the CHANGELOG.txt o OPT menu page overworked, so that the layout and usage matches with the other menus o the Global Scale can now be changed in the OPT menu Not only the number, but also the Scale Name will now be displayed! [/code] [code] o A sustain function has been added. If enabled, all played notes of a track stay active until the next note is played. This option overrides the gate length! It is especially useful for long chords (e.g. strings) since it saves you from modifiying the gatelength for each step. It's also a nice effect during a tune is played! The sustain option is selectable within the Track Mode menu at the position which was formerly used by the Global Scale o A new event mode has been added: Note/CC/CC This one has been forgotten during the definition of the event mode. In order to get a consistent selection, the new mode has been added between Note/CC/Len and Vel/CC/CC This leads to a slight incompatibility of old patterns. If a track has used Vel/CC/CC, CC/CC/Len or CC/CC/CC before, the event mode has to be incremented by 1 within the Event Mode menu!

Raffael aka. supcha has built this MIDIbox SID: The impressing thing: just have a look on supcha's posting counter - he was able to build the whole thing without any help from our side. :) A lot of additional pictures can be found at his website: http://raffi.kicks-ass.net/gallery2/main.php?g2_itemId=1696

MIOS V1.9d is now available for download; this version got updates for PICs with code space > 64k, a requirement for PIC18F4682 and PIC18F4685 From the ChangeLog: o support for PIC18F4685 and PIC18F4682 o initialisation of 4bit CLCD overworked o 4bit CLCD selected by default for PIC18F4685 and PIC18F4682 to free the ECAN pins at RB2 and RB3 Note that the D0-D3 pins of the CLCD should not be connected to the PIC in this case, but clamped to ground instead! o new Bootloader version V1.2b: support for the new PICs Output drivers of Port B are not enabled during the startup phase anymore. [/code] An update is not required for PIC18F452 users, because the behaviour should be almost identical. Only exception: if your Character LCD is connected with the 4bit interface only, it could make sense to try the new release (it would be interesting for me, if your CLCD can still be initialised) [b]Info for programmers:[/b] under migration/ you will find the new include files for PIC18F4685 and PIC18F4682 mios.h now contains generic parameters for memory sizes, special registers and workarounds - they are mainly used within MIOS itself, but could also be useful for normal applications. macros.h contains variations for table preload macros when the source code is compiled for PICs with code space > 64k. For such derivatives, TBLPTRU is changed as well. Please note, that the PIC18F4685 and PIC18F4682 are not 100% binary compatible to PIC18F452. RAM from 0x60-0x7f is not directly accessible. BANKED accesses have to be used instead. Some bigger applications which allocate this area could crash. Since this PIC is currently only relevant for future MIDIbox SID versions, an adapted release of MBSID V1.7303 will be available soon, which will in addition support communication via the integrated ECAN peripheral (see http://www.ucapps.de/midibox_network.html) Best Regards, Thorsten.

Yes, it's 60 kilobytes per second :) Best Regards, Thorsten.

The full featured PIC18F4685 version of the bootloader, MIOS and MBSID V1.7 is already distributed to experienced users. I'm waiting for feedback to ensure, that everything works flawless before new users are starting with the overworked firmware - this is to avoid an immense support effort at my side for this new technology. Patience please! Best Regards, Thorsten.

Yes, the function parameters are correct. Great to see a new sequencer concept comming! Maybe the distance between the pots is a little bit too short for adjustments during a live session, I would reduce it to 2x2x4 pots Best Regards, Thorsten.

for me it's a nice new toy which opens a lot of new possibilities. In fact, the control surface handling is realized in a much more elegant way now :) The transfer rate heavily depends on the frame content, I can only give some approximated values, which I've measured with the scope. Without the request/acknowledge protocol the transfer rate of a RAM->RAM transfer is about 120kb/s With the HLP I'm using to ensure that a slave doesn't miss a message it's ca. half of this: 60kb/s (assumed is a payload of 8 bytes for each frame + 2 bytes of the extended ID. The remaining bits of the ID, the CRC, the stuffed bits are not counted) Compared to MIDI (3kb/s) this means an improvement of factor 20 But the speed alone doesn't make the performance, it's also the functionlity given by the ECAN peripheral. E.g., due to the hardware based message buffers and reception filtering it doesn't require an interrupt handler which collects the incoming data, this saves some CPU time, which is better invested in stimulating the SID registers. :) Thanks a lot!!! Best Regards, Thorsten.

After a lot of experiments I've frozen the HLP. A summary can be found here: http://www.ucapps.de/midibox_network.html Hope that it is flexible enough for the next 1..2 years ;-) Best Regards, Thorsten.

See MBSID V2 wishlist, search for "Modulation matrix enhancements" In MBFM it's already possible to control the frequency of a LFO with another LFO, but the MBSID V2 implementation will be even more flexible. Best Regards, Thorsten.

No, it's just only the MIOS_MIDI_MergerSet function which needs to be called from the Init() function, see also http://www.ucapps.de/cmios_fun.html#MIOS_MIDI_MergerSet Best Regards, Thorsten.

I don't think, that trying an update is a good idea when no proper upload request has been sent to the core. It really looks strange, we never had such a case before. It looks like the baudrate is not correct, but as Karlchen wrote, that he has already checked the crystal, it must be something else. The case that the MIDIbox-to-COM mode is selected via ID can be excluded (38200 baud, in this case, you would see Pitch Bender Events). Since your Edirol interface receives more 1's than 0's (FF, F8, FC), could it be that there is a short at the MIDI Out line? Do you have the possibility to check the MIDI In port of your MIDI interface with another MIDI device? Best Regards, Thorsten.