Duggle

Hi rasOfir, I offered this solution as it is extremely safe, easy to build and very cheap. I shop at electronics disposal stores when I can, and there is one in my city that has various plug packs for as low as $1. It really reduces the work, the expense and weight and size demands of mounting and wiring and switching mains transformers in your box, when you only need a few mA's at different voltages. The only drawback is that the core regulator will run a little warmer (due to the higher input voltage) and will require small heatsink. cheers

The +V connection is not necessary in this application. It is for the clamping diodes which are not needed here. If it is connected, it should be to +5V (this is available from the Din power-in connection.). cheers

Also, of course one transistor is required per line so if your designing 11 lines, a transisor array in an IC package will be neater. There is a device that is fairly common that has 7 transistors per chip ULN2003 (often used to drive relays or lamps from logic pins) also called DS2003. http://www.ece.osu.edu/~passino/DS2003.pdf It also has 2.7k base resistor buit in. cheers

Hi, The transistor provides an inverting and level shifting function. The transistor connections: (pity I cant give a drawing) 1. emitter connected to common ground. 2. base connected to resistor (any value from about 1k up to 47k ), other end of resistor connected to o/p port of VE. 3. collector connected to DIN port. There are many signal NPN's that can do this job BC547 is a commonly available one. cheers

hi, I had a quick look. It seems that both the VE module and Toolkit are required (thats 50 + 120 euro). As far as interfacing, the VE module uses logic 0..3V while MBHP Din needs 0..5V input. This can be achieved by driving the base of an NPN transistor so that a logic 1 from the VE is like pushing a button on the Din. I would be more concerned with the software required by VE. Do they have lots of demo programs that you can hack to suit your needs or do you have to go through a big learning curve just to get it running? cheers

hi, As you will read, the primary bootstrap loader is once only (per core) operation that needs to be performed using a programming device such as JDM. Thereafter all MIOS and application uploads are done via midi without the need for the JDM. This is why some people buy their PICs off SmashTV with the bootloader and MIOS preprogrammed. (I've had very good experience ordering from Smash's webshop). The JDM is easy, cheap to build. Some people have problems with it, but I never have. There is a new, more reliable programming device available called MBHP Burner comming out. I recommend the "Sysexbox" applet for managing the MIOS and application uploads. MidiOX works but is harder to set up. Unfortunately, there is no other documentation apart from what's at uCApps.de (that I know of) If you patiently read everything and follow the instructions you can be confident of success as many hundereds of midiboxers have been this way before you! If you have the knack for documentation your offerings will be greatfully received at the midibox.org WIKI. cheers

hi, alternatively, learn about the standard applications that do not require you to program in assembler. They are extremely flexible in both the hardware configuration and the event mapping. If you can adapt your aims to achieve what you want using an existing framework such MB64 or MB64E, you will be up and running so much earlier. It is likely that some of your ideas about what you need for midi/music workflow may change after using your control surface for a while. Sysex is used to upload and burn applications to your midibox. It is also used to upload configurations that may be edited in a text editor or vmidibox. (youll discover these as you keep reading...) cheers

hi, Midibox LC (and indeed other MBHP applications) are made by an interconnection of standard MBHP modules. This is why the design docs on the uCApps LC page (*.pdf's) consist of connection diagrams of standard modules. A constructor needs to refer to these diagrams as well as the schematics of the individual standard MBHP modules to form a complete picture. Sometimes people ask "why is it not all together in one place with a complete parts list for the project?". I think the answer is that there are endless variations and customisations that are often desired. Another answer is that it is just plain hard work assembing and maintaining documentation (any volunteers?). As a builder of various midibox related projects, I've found it necessary to become really familiar with all the modules, to print out the various schematics and keep them in folders for each project. Everyone has their own prefered way to keep organised. Anyhow, good luck with your project. cheers

hi, I think the lack of bypass caps has been noted before. it is standard engineering practice and should be followed. However, I don't know of a case where this has caused problems with MBHP projects. Many of the basic modules (Din, Dout,Core ) have been in service for some years now, so we know they work reliably. Having said that, when I built my MB64 over 2yrs ago I included at least a few 100nF monlithic ceramic caps, on each board mounted directly across the power rails. It is easy to retrofit ceramic caps (either SMD, or leaded variety), so I think a redesign is really uncalled for. Perhaps something to think about including in future designs, or if a board is being redrafted, though. cheers

hi, SMD devices can be easily terminated with 0.25 mm diameter enamel coated wire, the type used for winding coils and transformers. -Strip the wire 1mm from the end by dipping it in molten solder on the soldering iron tip. -Avoid breathing the fumes and wipe the excess solder off the soldering iron tip. -Tin the leads of the IC using fresh solder. -Wipe the excess solder off the soldering iron tip again. -Re-tin the tip of the iron, then bring the wire to the IC lead, and apply the tip of the iron. If done right, the junction heats instantly and sufficient solder flows to form a perfect joint. It happens very quickly, and reletively little heat is transferred into the IC chip itself so it is safe. A DIP adapter can be made using this technique to terminate the IC to a low cost DIP socket. (keep the wires short in length). Various methods can be used to provide extra mechanical robustness to the finnished product e.g generous coatings of circuit laquor, epoxy resin, etc. I have recently completed my MBFM control surface using this technique to wire all SMD versions of the DINs and DOUTs shift register chips directly to the control surface matrixboard with encoders, LEDs and buttons. It makes for a very compact and concise assembley with a single 6 conductor ribbon connecting the control surface to the Core. I'll have photos for my "Midibox of the Week" entry soon! cheers

hi, If you want people to comment on possibilities such as this, then provide some background info that may be of use to others: -Where do you get this chip? -Why is it of interest? -What documentation data sheets have been located (eg datasheets). Some links. I did a google search after reading the first post in this thread and found very few hits, with nothing at all informative, so thought the device was too obscure for anyone to be interested in. (btw; I find the WIKI QA103 a bit harsh, as I dont mind having various devices brought to my attention.) cheers

hi TK, will it be possible to implement digital audio I/O streaming through this interface? (Say through spi or something). Anyhow, looks great! cheers

hi, For a single channel converter using stock parts its a good solution. However, with 16+ resistors per channel the effort is transferred to construction time and board size. As suggested, there must be low cost, integrated solution for such a generic requirement. cheers

hi, A standard Midibox64 can be assembled without any pots if desired (thus no Ain board required). You can have up to 64 configurable midi buttons and LEDs. cheers

Hi, I had a look at the keyboard I/f site of your link. Why not use Midiout from one of the Velocity sensing scanmatrix controllers wired to Midi in of a MIDIBox64 derivative (they have Merge function already). You will be up and running very quickly this way. You can assign control functions and further customize the operation of your rig by editing and downloading MB64 ini files. You could then begin learning MIOS functions (and PIC programming) and further customize your setup. For example you could write a function that simply translates incoming velocities to whatever curve you wish, create splits and crossfades and whatever other master keyboard functionality you wish. I suggest you use standard modules (ie kbd adapter like you suggest, and a MIOS core) and use them to customize to get exactly what you envisage. This way you will be operational very early and can concentrate on adding the unique features you desire. This way way you avoid spending all your time reinventing the wheel. On the subject of deriving velocity curve, if you parameterise the response curve to several CC's representing breakpoints and slopes, assign these to control surface knobs, experiment. If you discover some "ideal" response curves they could be encoded (in various ways) to MIOS application user functions that can be recalled via pushbutton for example. Of course this requires programming skill but it's scope is limited to converting CC's into a numerical function that shifts and scales velocity bytes (and of course developing some knowledge of MIOS). cheers

If the C4 was routed with a midi monitor program between it and the application you would see the interaction between it and the host. Reverse engineering would not be that hard, but you would have to have a C4 unit to run the tests. You could record a midi file of all the midi events transfered between the application and the C4 and decipher the protocol later, even. cheers

One way of interfacing such a circuit is to connect the square wave to the CCP1 or CCP2 pin of a PIC and perform a capture (period measurement). This is to provide a frequency measurement. Program logic would use changes in pitch to trigger different notes. Could integrate pitch bend. This requires good level of programming skill. Due to pin and other resource allocations MIOS may not be an option for such an application. cheers [edit] hi moog, the point being that the circuit you posted outputs an AC signal, whereas Midibox uses "changing DC" input. Tracking an AC signal is more involved. It's doable but the Midibox platform has no existing interface for such a signal (as yet). cheers

Hi, Midi-in generally requires a hardware UART. Unfortunately the Core has only one. The exception to this is the Midimerger application which uses all its CPU power to implement a second UART. The answer (not eligant but works) is to integrate a second Core board into your Midibox, configured as a dedicated merger. cheers

Hi, MBHP AIN modules convert 0-5V DC into CC's (0 to 127). This circuit produces a square wave 9Vpp. A Frequency to Voltage IC (or other circuit) could be used to convert this into a MBHP useable signal. cheers

I agree, my first thought is to beware of noise introduced by the motor drive! Remember that this project has been devised and successfully implemented for non audio application, rather DAW automation (optically isolated midi etc). For example the H bridge clamping diodes are mounted on the pcb, but where localised noise could be an issue, I would mount them directly across the motor with capacitor bypassed Power supply lines so that the back-emf current spikes are not carried through wiring close to sensitive audio circuitry (yuk!). I'm not saying what youre aiming for can't be done (it clearly can) just do lots of prototyping ( and proper testing) before you commit to a major project. good luck

Hi, I want to make some big changes to the position of buttons and LEDs in their respective Shift registers (MBHP Logic/Mackie). This is to cater for a customised modular system (2*4 input channels) that has a PCB with buttons encoders and leds but also with SR's integrated to reduce wiring. Can the LEDring/Meter column driver (common cathode) be freely reassigned to different registers in the chain? I understand that the 12 row drivers would need to be contiguous and aligned (2xSR's). My idea for this is to duplicate these SR's (row drivers) on the second module that has LEDrings/Meters for 4 channels. The idea is to "tap off" the data signal at the imput of the first (row driver) SR and take this signal to the duplicate (row driver) SR's on the second 4 channel module. I hope I expressed this clearly, its a simple concept that a schematic would express instantly. I'm wondering if there is anything I've overlooked and what source files contain the assignments that need to be changed. Thanks for any help/comments cheers

Hi all, Yes, the prototype PS2Midi converter has made significant progress. 4xPS2 Ports with basic keyboard and mouse autodect support, midi merge, have been implemented (using old 16F877 etc). Yes, some more exciting projects tend to steal time from this but I do revisit it regularly and will release the design in the next month or so. In the mean time, I ask people to keep a lookout for a source of cheap guarranteed PS2 touchpads (e.g VOTI seems to have one). Also, more detailed specifications on PS2 keyboard behaviour than is normally hit in a common google search. In particular I have noticed that the scan codes of certain keys are unexpectedly altered when the numlock LED is activated. I cant find info on this anywhere. It is easy to workaround but more info would help. cheers

Hi Smash, nice work, I would suggest that from an engineering point of view the purpose of the 330nF and 100nF caps is actually to guarantee the stability of the regulator at high frequencies (prevent it from oscillating) and mounted physically close to the regulator in the PCB layout. The 2200uF is most definitely for "smoothing", I agree. The 10uF is to reduce the output impedance of the circuit at high frequencies, (In other words, to provide a stable voltage even when a sudden spike of current is drawn.) Remenber a linear regulator is actually an amplifier with gain and negetive feedback. So to summerise, my recommendation is to depict the ceramic caps in the "regulation" box, and to rename the 10uF cap "output decoupling" (it decouples the regulator from sudden changes in load current). It probably seems pedantic, but if the effort is going to be made, its worth looking at the detail. Anyhow it's a very nice presentaion. cheers

hi, MBHP is an excellent platform to develop the user interface and controller of your project, but for the sampler functionality, it will need a dsp/digital audio subsystem which MBHP does not (yet!) have. Miss Parker looks like a very promising open source dsp unit that is midi controlled. I dont know about its RAM capacity/expandability, but the combation of Midibox control surface and DSP would be a great combo! cheers

Guillatine for cutting sheet materials works perfectly and no dust.