jimhenry

I have never seen documentation of how to rework a PCB, sometimes called "cut and jumper." The basic idea is to make two cuts across the PCB trace, perhaps an 1/8" apart and remove the trace between. Sometimes a soldering iron will release the cut trace. Leave enough trace that you can lay a wire flat against the trace and solder it to the trace. If done well the solder should "hug" the shape of the wire rather than being a big blob. Mechanically support the wire somehow, perhaps a dab of hot glue. Find an old PCB and practice before attacking useful boards.

Look at these two example projects: http://www.ucapps.de/mios/sm_example1_v2.zip http://www.ucapps.de/mios/sm_example2_v1.zip The scan matrix approach has not been used many people so you'll be somewhat on your own in pursuing this. For 16 switches, I'd be seriously tempted to cut traces and just wire each switch individually. Not elegant but less work on the Midibox end.

I haven't done enough with DIN boards to give an answer tempered by experience but based on analysis I don't think the length of the connection to the key switch inputs will be a problem. The cabling between DINs and to the Core carries all the 128 inputs on a single wire and the clock signal to make that work. You are dealing with fairly high frequency signals on that cable. The inputs are essentially DC signals.

Just trying to answer the "maybe you want to drive the pins active high" part. If the answer is no, you can ignore the rest. A DIN reads a digital input meaning an input that has only two states, ON or OFF. The two states need to be represented by two voltages, +5 volts (the power voltage for the Midibox) and 0 volts (the Midibox ground). The usual way to read a simple switch is to connect one side of the switch to ground and the other side to the DIN input pin. When the switch is closed, the input will be 0 volts. To get the +5 volts when the switch is open, the input pin is also connected to a resistor that is connected to +5 volts at the other end. This resistor is on the DIN board. The resistor has enough resistance that it won't supply any appreciable amount of current when the switch is closed. The result is an input for the DIN pin that is +5 volts when the switch is open and 0 volts when the switch is closed. If for some reason you wanted to connect your switch to +5 volts instead of ground, then you need to reverse the voltage on the other end of the resistor. This would be the "drive the pins active high" configuration.

It probably will be easiest to use two power supplies. The lamps that need 20 volts (not 24 volts?) probably don't need a very regulated voltage level. The 5 volt logic supply should be more carefully regulated and that regulation is provided by the Core module. So an appropriate wall wart supply to Core will give your 5 volts and whatever DC supply works for the lamps can be used for that. Just be sure the negative sides of both supplies are tied together to be the ground and you have about +9v to the Core and about +20v to the lamps.

Yes, not only can the two device share the same ground but it could create problems if they do not share the same ground. The idea of ground is very much like providing the foundation for everything. The voltages are all measured with reference to ground. So it is necessary to use the same ground for everything so the relative voltages remain the same throughout.

Might I suggest a contest for the best "Introduction to MidiBox" document? This is to be judged at least in part for its ability to make the Midibox accessible to more people.

You'll be using the MIDIIO128 project. You only need the DOUT half of the project. This will give a digital on/off signal in response to MIDI messages. You'll have to figure out how you interface the DOUT signals to the Yamaha.

Do you want the Yamaha to send MIDI, respond to MIDI, or both? Do you want to add MIDI to what is there or replace the analog electronics with MIDI?

I agree this is another great idea from SmashTV! It will simplify getting to the point of having a basic functioning MidiBox for beginners. In my experience getting people to that point is the biggest hurdle to overcome with the MidiBox. Once they have a MidiBox that does something, then they start to see how the whole MidiBox concept works and they start becoming self-sufficient. I think preloading MidiMon might be another step in this direction. I don't know if MidiMon enables MIDI Thru, but I think it should so that it can be installed in-line for its intended purpose. As a start-up aid having MIDI Thru and also putting out some sort of MIDI Time Code so that MIDI In and Out functionality can be confirmed even without an LCD would be a further help to the beginner. In the same vein, perhaps the CORE Board might eventually be revised to include LEDs to monitor MIDI In and Out. That probably would be useful in all applications of the MidiBox.

This sounds like a MIDI Loop to me. See if disconnecting one of the MIDI cables between the MidiBox and the computer stops the freeze. If so, then we can try to figure out why you have a loop.

All possible 128 inputs and outputs are accounted for by the software. You can connect as much or as little of the hardware to use this support as you want. Switch matrix is not fully developed. Don't bother with it. It is of no advantage for a project like this. Yes, you will do all your uploads of additional software via MIDI with free programs available on the MIDIbox site. Yes, you can do this easily. Yes, you write exactly the MIDI message you want sent for each switch on/off.

I don't know what the Maxim chip is but maybe it's a UART or maybe it is just an integrated RS-232 interface, which is more likely. Either way, what you are asking is going to require some non-trivial C64 programming. Was there some C64 code to go with that schematic? This is going to be a long way from a plug-and-play project. If you can't decipher that schematic I'd think twice and three times about getting into this. You'll be getting a lot of education if you manage to make this work. And I don't think there are many people left who can offer much help.

I could be completely off-base but I see the MIDIbox as an interface between MIDI signaling and "real world" things like switches and pots. You need a PC to MIDI interface to connect the MIDIbox to a PC. I think you are looking for a C64 to MIDI interface and I don't think a MIDIbox is going to be much help. It's been ages since I did C64 hacking but I think you might be able to do a MIDI interface by "bit banging" input and output pins directly from the C64. If I remember correctly, C64 RS-232 was done without a UART. You might just need to add the current loop interface which can be copied from the MIDIbox or a bunch of other places.

Here's a plug and play solution for you: http://www.organworks.com/Web/products/products.asp?productid=25&categoryid=17&category=MIDI%20Works

The Miditzer produces MIDI out. By directing it to the appropriate synthesizer hardware you can get anything you want for sound and electrical levels. I believe the same is true in one way or another for the other virtual organs. The FluidSynth that is bundled with the Miditzer may have confused you. It is the default configuration so that beginners will immediately hear the right sound with little effort. Even with that, it purely a function of the sound hardware on the computer and the amp input as to whether the two can be directly connected. Many sound cards have a Line Out that can be directly connected to a Line In on an organ console.

Perhaps you could work out something that would pick up the tap sound and use that to trigger the MIDI. Trying to get something on the dancer is not going to be easy.

Why do you not want to use the Atmel project? My guess is that you could do this with a PIC processor but that it is going to be a LOT of work.

The default PIC header of 0 is fine. It only matters when two or more cores are connected. I think an LCD is necessary. When you are getting started it will be very reassuring to see things happening on the LCD. When everything is working you can disconnect the LCD or just leave it inside the case in case you need to troubleshoot. The LCD is not needed for normal use. The diodes are used in switch matrix designs. The are more efficient in terms of hardware. But when you are making only one, efficiency of the hardware is not the most important thing. Efficiency of your time is. MIDIIO128 is "dumb" in that you have 128 inputs that are just looking for a high or low level provided by a simple switch. No tricky wiring. No diodes. No problems.

I think the reason the DOUT board was mentioned earlier is because you use a DOUT with the scan matrix input. With 1 DIN and 1 DOUT you can theoretically have 1,024 inputs. Even if it makes more sense to make the PCBs locally, or use point to point wiring, I would seriously consider buying the PIC preprogrammed from someone who can provide it with the bootstrap loader already programmed in. It undoubtedly will save you much time and expense no matter what it costs, which I don't think will be very much. Be sure to include an LCD in your project. It is marked optional but it is necessary unless you are experienced with MIDIbox so you can fly blind. Greenfox posted some pictures here: http://www.virtualorgan.com/forum/forum_posts.asp?TID=433&PN=1 You might need to register for the Miditzer Forum to see them: http://www.virtualorgan.com/forum/

I would suggest using the MIDIIO128 project to MIDIfy your harmonium. The scan matrix will work but it is conceptually a tiny bit more difficult. The big thing in favor of MIDIIO128 is that it is a project that has been in use for a long time and there is much more information and help available for it. You will need a CORE module and 3 DIN modules for your harmonium. As has been pointed out, you need to provide some type of electrical switches on the keys and pedals. This is primarily a mechanical issue and you'll have to figure out what is going to work for your harmonium. It could be as simple as a metal "chopper bar" that connects two spring wires. You may wish to add some additional controls. I would particularly suggest adding combination pistons under the keyboard if possible. They are of great value with the virtual organs that you will want to control.

Look at this topic in the MIDIbox WIKI: MIOS: MIDIbox Operating System

I have never done anything with drawbars.

It is possible if you are prepared to do some programming and engineering. One way to sense velocity is to have two contacts that close at different points in the key travel. You measure the time to go between the two contacts to determine velocity. As far as I know, no one has done that yet with a MIDIbox.

The scan is on a fixed rate of once every millisecond. Removing the DOUT table would not change the scan rate. I've never looked but I would expect that removing the table would create a lot of problems. I wouldn't touch it.