/tilted/

Cimo, don't you dare listen to these crazed Americans. You're cake is awesome, and it does NOT need any of that high-fructose-corn-syrup garbage!! ;)

Another thing that you want to find out is how much MIDI the program responds to. Are the funtions you plan to use accessible via MIDI? Is there a MIDI learn type function? Is the program compatible with an existing interface, such as Mackie Control Universal (ie Logic Control)? Is there a comprehensive midi routing capability?

Here 'tis. http://www2.yamaha.co.jp/manual/pdf/emi/english/synth/FB01E.PDF A MB front end would need to speak using sysex, and therefore might not be able to add much in terms of real-time control, but you'd certainly be able to increase the number of patches by using banksticks.

Tell me you also have +5v on the OPL3 board...

It could be possible to do what you are saying. This project would require an extensive re-write of the LC application. If you can automate it so that you bank up once or twice (ch 9-16, 17-24), then arm the appropriate channel, then banks back down to channels 1-8 again. This is conceptually viable, but would require some memory to store flags for armed/disarmed status. Also, you would need to do tests to determine how quickly you could change banks back and forth, while still reliably getting all the information you need. Here's a little more info and some questions: * The SMPTE display, Save Session, Jogwheel, and 5 of your 6 transport buttons are all native in the LC app. * The Return To Zero button may or may not be possible, depending on your host. * What does the Pre/Post button do? * What do the "Play" buttons do? Is it just the opposite of "Record", or does it also serve a mute/solo function? Perhaps you could achieve this using a combination of buttons... By far, your easier option here would be to simplify it to an 8 channel version, or to use 3 midi ports (via a 3x3 or 4x4 midi device) Check out the wiki for more info.

I'm confused... exactly how is it that making nice squelchy filter sounds is "not really essential"? ;D

Here you can see, with Axel's LC, he's used an entire MIDIMAN USB 4x4 PCB to interface with the host system. By doing this, you only need to connect one USB cable. This is still 1 MIDI port per LC.

The cores cannot be linked in Midibox LC application. Each core takes one MIDI port, and uses 8 channels on that port. You cannot use channels 9-16, as the Mackie interface does not support this, and in fact, Channel 9 does Master fader control. I hope this helps, as I'm not strictly sure I understand the question...

One per midi port. Any other number depends on your host application, and operating system. Win 98SE and 2000 can do 10 ports max, so therefore 80 faders. Win XP has no limit on MIDI ports that I know of, so you can have 8 times infinity faders... No idea about Macs...

I was thinking about this one too, and in typical /tilted/ fashion, it was all think and no do... Maybe you could just use FETs attached to AOUTs? (more or less)

Let's just hope it's not to make blue perspex and 2x4 MB 6582s...

I'm also working on an idea using an inverting buffer to do a similar trick. I haven't tested it yet, but of course have already started the design for the PCB... :P so my theory goes: [tt] Vpot matrixed cathode 595s +---> Inverting Buffers ---> Vpot cathodes (blue) L---> Vpot cathodes (red) Bi-Colour LEDs Vpot/meter anodes <------------- Common anodes [/tt] I just hope the anode 595s can sink enough current... else I'll need to use transistors. This would mean that the V-Pots have a red pointer, with the remaining LEDring glowing blue.

Look for this inside LC_LEDS: [tt]LC_LED_Set_Loop_LEDOn movf LC_LED_LOOP_CTR, W call MIOS_DOUT_PinSet1 rgoto LC_LED_Set_Loop_Next LC_LED_Set_Loop_LEDOff movf LC_LED_LOOP_CTR, W call MIOS_DOUT_PinSet0 ;; rgoto LC_LED_Set_Loop_Next [/tt] And change to: [tt]LC_LED_Set_Loop_LEDOn movf LC_LED_LOOP_CTR, W call MIOS_DOUT_PinSet0 rgoto LC_LED_Set_Loop_Next LC_LED_Set_Loop_LEDOff movf LC_LED_LOOP_CTR, W call MIOS_DOUT_PinSet1 ;; rgoto LC_LED_Set_Loop_Next [/tt] This would work assuming all your leds are the same. If not, (ie a mix of bicolour and normal), then reverse the polarity of the normal LEDs, and tie their anodes to VCC instead of GND. I think this would work.

I did modify it, but the mod itself is minimal. "all" you need to do is remove the diodes, then connect your bridge rectifier where the diodes were (looking at the picture close up, you connect to the right-most diode pins, which connect to the inputs of the regulators, not the left-most pins, which connect to each other!!) yes. The picture is very small. It seems the "free" image host re-sized it for me...

Wait. Don't panic. You can still save things here. Option 1: You can attach heatsinks to the regulators. (This might be a good idea anyway) Option 2: You can modify the circuit you have, to be full wave rectified, BUT... You may need a new transformer. If you have a 12v Centre Tap, you need to get a new one. If you have a 24v Centre Tap, all is well. Then you pull out the diodes, and use a bridge rectifier component. (there is another slightly messy option) Then connect the AC inputs of your bridge rectifier, to the outermost taps of the transformer (ie, not the centre tap) Then connect the rectified outputs (+ and -) of the bridge, to the board you have, at the approriate locations. Also connect the centre tap of your transformer to the (A2) point on the board shown. Please excuse my awful MSPaint work!!

Oh! another thing... Check the resistance between pins 15 and 16, to the EL traces for the backlight supply. (I'm looking at the data sheet, and it mentions an EL option?...) You just want to be really sure you don't connect the 100v AC from your EL inverter, to the core module, as this would make smoke happen... :-\ That said, if the lcd is not EL backlit, none of this matters... Edit It seems you already said it is a 14 pin connection. Any way you look at things, there's not going to be a pin 15 or 16...

It sounds as though you have a short or low ohm somewhere after your regulators. The supplies you specified should be adequate to your needs (although the 12v might be close), so I would suggest you re-check your voltages. Disconnect everything from your core (except the PSU), and check for the correct +5v. Then re-connect one module at a time, and re-check your voltage each time. If your voltages suddenly drop with the addition of a particular module, then check and re-check that module, then move on once you have restored the correct voltage. I'm also curious, you mention a 300mA 12v supply... Is that regulated? And if so, are you trying to regulate an already regulated supply? that would cause some heating, as the reg struggles if it doesn't have a little surplus...

Off hand, I'm not sure if you need to change anything in the app to select 4 bit mode. I would imagine that since those pins have been re-located for the CAN bus, you have no option but to use 4 bit mode. The CAN bus may also explain why your LCD is not quite functional... The LCD would have difficulty understanding the ramblings of the CAN bus, I imagine. Perhaps someone can jump in here? Someone who's built a V2?

Assuming you're talking about a MidiBox SID V2 (which, given the location of the post, you are?) - Then you should connect the LCD using 4 bit mode. This means not all 16 pins of the LCD are conected to the core. More details at uCApps.de or the wiki No. The LCD should be fine. The interface is pretty tough (and believe me, I've done some dumb things to LCDs!!)

This statement alone created a large, red, flashing neon sign that says "check your connections" I would hasten to suspect a bad solder joint, or possibly a short. Are you using the new PIC with the new MIOS and the new connection (ie 4 bit)?

What if you got a 1U rack case, and mounted the FB-01 next to the MBFM? Best of both worlds??

Are your 0v lines connected? ie, by using two seperate psus, you may have two seperate 0v lines, one for 5v, a second for +/- 12v. You'll need these to be connected, otherwise your rails will be out.

Perhaps it would help to mount the "keys" more like vibrophone keys? Ie drill two holes from one side to the other. One hole about 1/5th down from the top, the other about 1/5th up from the bottom. You then mount the keys using string passed through all the keys, usually tensioned at one end. Could cut down a lot of the crosstalk...

Not sure if I understand this bit: http://www.midibox.org/dokuwiki/mc_protocol_mappings Might be helpful...?

I wonder if you've considered the MBLC_HOLD_LAYER function? I think this would be very helpful if you don't want to compromise too many buttons. Have a good read of the LC_IO_TABLE file. you can nearly double the functionality of a given number of buttons... ;)