SeverityOne

As far as I can tell, they are not connected directly to one another, but both to the 74HC14A. That's what the measurements showed. I'l have to leave it at this for a little under two weeks, because of a holiday. - Peter

Now that's a coincidence: just as I was about to post, you had posted your message about the SoundBlaster and the multi-meter. If we take the following pin out, according to the print on the DB50XG: 1 2 3 4 5 6 ... then both pins 4 and 8 are connected. If you weren't confused before, this would be a good time. The DB50XG has all components - without exception surface mounted - on one side of the PCB. Only the 26-pin WaveBlaster connector is soldered through-hole and sits on the other side of the board. It's a clever design, so they can use the huge grounded areas on the non-component side as a shield. (At least, that's what I presume, I could be talking total nonsense here.) If you look closely, there are are two small leads on the non-component side coming from the 26-pin connector. They go on for about 2.5 cm/one inch and end up in through-holes (which makes them excellent to use with a multi-meter). These come from pins 4 and 8. On the other side of the through-holes is a 74HC14A, a Schmitt-trigger inverter. From the Motorola datasheet: Well, that would make sense. Another lead from the 74HC14A goes to pin 26, 'Reset', so that also appears to be correct. The only conclusion that I feel is safe to draw at this moment is that you can't trust anything you find on the internet... - Peter

Hmmm, let's see how easy it is to follow the leads on the SoundBlaster card. :) - Peter

Indeed, that was the idea. For the DB50XG it'll be useless, because it doesn't have a MIDI out, but since that pin on the I2C chip would remain unused anyway, there's no harm in connecting it. Unless it would lead to unwanted behaviour and 'ghost' signals being received, but one would expect Creative to have thought of that. And otherwise, you can always disregard the stream altogether. For the next two weeks I'll be abroad and likely not be working on this... - Peter

OK... as a architect/designer, the first thing you do is create a functional overview of the desired end result: This doesn't actually add anything, but since I had so much fun using JFDraw (a freeware Java-based vector drawing programme), I included it here. Although I may have gone a bit overboard with the XG logo. :) The MIDI out from the Waveblaster connector won't do anything in the case of the DB50XG, but the WaveBlaster standard supports it, and it won't cost anything significant in extra components. Next up is a drawing of what MIDIbox modules would be used. - Peter

Something like that. I'd have to look into MIOS and the MIDIbox 64 first, of course. But since these days my programming limits itself to distributed interactive web sites in Java, getting my hands dirty on some plain old C could be fun. Assembly, the last bit I did was a super-optimised loop of 68000 code to calculate a Mandelbrot set on an Amiga, so I might not be ready for that just yet. - Peter

I did! Thanks! ...a day after I paid €1,30 for the same article downloaded from the Elektor site. :) Then again, it's in Dutch, which happens to be one of my two native languages. (Something that my wife says can't happen, having two native languages.) But since the article was originally written in English, it's good to have the original next to it. Thanks again, - Peter

Whilst going through the documentation, it became clear to me that there's just one MIDI path throughout a core module. So yes, I'd need a separate MIDI Out (and In, for those Waveblaster boards that support it). My plan is to design a PCB with more or less the components of the MIDI I2C module, minus the opto-coupler, plus the components needed for the PCB that el-bee mentioned, again minus the opto-coupler. The problem I have at the moment, though, is that I can't for the life of me figure out how to add a shape to a schematic design with Eagle Light. But give me time. :) Another idea that came to me is an adapter for Yamaha PLG boards... but let's start with this one first. I'll have a look at that Elektor magazine. Thanks again everyone, - Peter

Looking at the schematic of the core module, the MO pin of J11 is physically connected to the output at J12. So sending different signals to those two ports is not going to happen. Obviously, I could connect the MI pin to the MIDI in of the DB50XG, but then I wouldn't be able to send messages from knob movements and button presses to it. So, back to the drawing board. - Peter

Just for the record, so I will need an extra module to extra the second MIDI out at TTL level from the core? Is there no way to extract it one way or another from J11, like the LTC board gets it from there? - Peter

So... I had a look. My question about the two MIDI Outs may not have been entirely clear. What I meant is that there should be the usual MIDI In and Out, to communicate with the box. The In, well, that should be pretty clear, and the Out would do things like patch dumps. However, apart from reacting to the messages coming from the MIDI In, the DB50XG would have to react to the signals coming from the Core module (turning knobs, pushing buttons, etc.), and these would be different from those going to the MIDI Out. Unless it's too much of a hassle, and not worth it, in which case the MIDI Out could be sacrificed. But I think the second MIDI Out, at TTL level, might be connected to J11. However, this is the point where my understanding of the MIDIbox platform hopelessly falls short. Any hints in this regard would help me a great deal, because it looks like a Core (which I have one spare), a DIN and a DOUT (both of which still have to built into a SEQ box), an AIN (for the knobs or sliders) plus a small custom PCB would do the trick. Thanks again, - Peter

Well, my EX5S is a rompler too, among other things, but you can control quite a few parameters. It's been a long time since I last looked at the manual of the DB50XG (which comes with the complete MIDI specification) but you can control quite a few parameters, from what I remember. It's not unlike the CS1x keyboard, and identical to the MU-10 module in regards to sound capabilities. Admittedly, that's not a lot of controls. :) So on the one hand it's a simple General MIDI device, but this being Yamaha, you have three effects processors as well, plus set of resonant filters. More info can be found at Sonic State. (Karaoke chip? Oh dear.) This is the reason that I had this idea, because it's such a lovely little board. - Peter

Thnaks for the replies so far. Well, finding the DB50XG is not the problem; I bought it at a musical instruments shop way back, which has since long gone. The problem is that I'm not very familiar with the electrical properties of electronics: logically I understand everything; it's when it gets to complex figures, capacitors and resistors that my mind draws a blank. And, of course, I wouldn't want to risk damaging the DB50XG. But it seems such an excellent project to convert it into a MIDIbox XG or something like that. - Peter

Hi, The other day, whilst going through old PC hardware, I found my old ISA SoundBlaster 16, with the Yamaha DB50XG daughter board stuck to it. Now the SoundBlaster 16 is useless in this day and age - you can't even rip an FM chip off it - but the DB50XG is an interesting little sound generator. Which made me wonder, how difficult would it be to make a small box with a small PCB and some MIDIbox components? There are several resources on the internet about either the Waveblaster or the DB50XG (like this one, this one and this schematic), but these deal with building a board that converts the TTL signals of the board to MIDI, via an opto-coupler and buffers for the MIDI Thru (the interface does not have a MIDI Out; at least, the DB50XG doesn't). Since the MIDIbox hardware also works on TTL level (right?), what would be needed, apart from a Core module and something like an AIN (for knobs), DIN (for buttons and data wheels) and/or DOUT (for LEDs) modules? The MIDIbox boards would need to be able to handle MIDI In and Out (for proper MIDI ports), and an extra MIDI Out (on TTL level) to control the daughter board. Software... that's another issue, but it should be feasible enough. Any help would be greatly appreciated. - Peter

This is incredible - an FM synth as well! Now, true to the FM tradition, the user interface would probably consist of four buttons and a two-segment LED display :D but do you have a general idea of the amount of buttons involved? I'm going to order them for the SEQ and the SID, see. :) Great job! - Peter