latigid on

Eh, I should read better... Looks like there are already MOSFET drivers on the lamps. As they look biased to 25V, it seems that the 5V signal from a DOUT will not be high enough (did not check the transistor specs). But it could be tried with 5V supplied to the gate of TR7/TR8. If you know what IC4 is, then the simplest way will be to match that voltage as driven from a DOUT. Please post the whole service manual if possible.

Looks like the lamps are already driven by some sort of line driver/level shifter. This is good, as one can lift the LEDs out and wire to DOUT that way. Considering the way they are connected, it seems like the 2981 chip will work better. If "IC4" still works, another option would be to connect to its inputs. The driver chip looks quite strange; what is the part number on the IC? Is the idea to use both the normal function of the mixer and something MIDIfied? Or is it a carcass just to build a controller?

In principle that sounds about right, but the current seems very high. Of course you have no resistor there, so the bulb will take what it can from the PSU. I think the 2981 can source only -500mA. If your requirements are bright lamps, then you should use relays or MOSFET transistors instead. What sort of 12V lamps are we talking? If they're the small panel indicators, the 2981 will be fine. The other way around also works: ULN2803 provides eight sink drivers. You then power the bulbs at 12V and connect to the DOUT via suitable resistors. One drawback of the current boards is that there is no space for resistors onboard, so any current has to flow through the pinheader and typically thin ribbon cable wires.

I think it should fit in place of the resistor array, but please check the pinout. There's no header nor PCB trace for power; you have to solder something to the DIP socket. I recommend limiting the current after the 2981 with suitable resistors. It might be better to use the DOUT as-is and put together something on strip/veroboard with a 2981. What sort of current? Also check here: It depends on your particular revision of DOUT. I believe the latest ones require no trace cutting.

UDN2981 can even fit into the DOUT boards. They require a secondary voltage though (595 won"t run on 12V).

Well done! I hope the customisations worked okay. Let's see some build photos of the MBCV! Best, Andy

I suggest to unzip all of the files here http://www.ucapps.de/mios/midibox_cv_v1_3.zip into something easy like C:\CV\ you would then navigate as cd C:\CV make perhaps you just have the asm in which case there is no command available. All of the subfolders (src etc.) are required to build a hex.

in MSYS cd %PATH% is the command to change directories, where %PATH% is the location of your .asm ls lists the directories and files cd can be run for one folder at a time, using ls to see what else is there cd C: //opens C: drive or as a path cd C:\2487\trunk\apps\controllers\midibox_ng_v1 In your case, use cd until you are in the correct folder, then run make. If you have set up the dependencies and system variables correctly, your .asm will be used to build a .hex. If you haven't followed the other instructions to setup the system properly you get no file and a list of error messages.

MSYS

Come now, don't be lazy! You're so close! Otherwise the MBCV will sit around wondering what it could have done with its life.

Have you done everything written on the page? Come back when you have.

You could also try the following: not sure if it's different for MIOS8.

You have to setup the build environment on your computer. Follow the instructions given on the wiki page. Then run MSYS, point the path to the directory containing your .asm and run the "make" command. This converts the coded file into a hex ready for transmitting using MIOS Studio.

Hopefully this helps: http://www.midibox.org/dokuwiki/doku.php?id=windows_toolchain_core&s[]=msys also, what is your OS/MIDIbox app etc.?

Actually, yes, you're in luck! TK. has recently developed a user-friendly language called "MIOS"! So as long as you want to run existing software applications (like SEQ, SID etc.) instead of programming in assembly and C, all you have to do is edit definition files to suit your hardware! It gets better: with the old MIOS8 one had to setup a build environment with dependencies and then run MSYS or some other compiler! Now with MIOS32 it's as easy as editing a text file stored on an SDcard! I hope that answers your question! With a process this simple and considering your obvious high levels of enthusiasm we look forward to seeing your first MIDIbox in action soon!

Doch, yes you CAN :)

Looks more or less like the schem; should be okay. Can you post your proposed frontpanel layout? If it's in some sort of grid shape it could be easier to wire onto the hardware directly.

Not possible Nope

http://wiki.midibox.org/doku.php?id=mididocs:seq:beginners_guide:start#mixer_maps

Switching the hardware connections is not trivial. Maybe it could work with three Cores: one SEQ, one NG and one to forward the control surface/display data. To me it doesn't make sense as the hardware/labelling is specific to each. I think most people would prefer two separate units with dedicated controls.

Hi Zam, Yes, maybe I even overheated the pin whilst soldering or I just got a bad one. Best, Andy

If the inaccuracy is only 1 cent, I'd say that's pretty good. From memory my channel 1 had some linear portions but others were off by a couple hundred millivolts. It just comes down to the DAC; e.g. MAX525 is simply a better (and much more expensive) chip.

The offset trimmer is only used in "bipolar" mode. Only the scale pot is used for unipolar. If you've tried calibrating from low and high ends of the scale (e.g. C2/C4) and the result is still not linear, then this is how your DAC behaves; there is nothing else to do. My channel 1 CV (AOUT_NG) is very non-linear and as such I can only use the gate.

If you don't want to add any more DIN/DOUT shift registers then you have to use the matrix routing. Depending on your skill/patience level and the particular layout of the panel, it's certainly possible, but likely a lot of frustrating work. Probably an 8x8 matrix will work without sink transistors as long as the LEDs aren't driven too brightly. It won't hurt to add them.

Yes, or connect the switch to JD8 and the diode to JD5 Just follow these: http://www.midibox.org/dokuwiki/lib/exe/fetch.php?media=mb-6582:mb-6582_cs_dout_wiring.pdf http://www.midibox.org/dokuwiki/lib/exe/fetch.php?media=mb-6582:mb-6582_cs_din_wiring.pdf You will soon see why we normally use PCBs for matrix routing :) It is possible to do this and I recommend it in the absence of a front panel PCB. You will need one DIN pin for each switch, with the other pin wired to 0V (ground). No diode is required. For LEDs do the same: you need one DOUT pin + resistor for every LED. If you are completing a comparable control surface to the MB-6582, you'll have to expand the SRIO chain, labelled as J8 and J9 just up from the bottom-left of the Base PCB. and below the J8/9_Core4 header. You can either use DINX4/DOUTX4 modules or wire the chips/resistors/headers etc. onto your stripboard.