Wilba

Before you all go off and redesign your layouts, just keep in mind that just because TK has a nice panel that he's not going to throw out, there's no reason why people can't design whatever layout they like. I just threw in some ideas, they're not mandatory.

This thread is turning into Biggest MIDIbox Bloopers. Don't feel bad. I fried four 6581 SIDs with 14v on the 12v supply line after converting to the "C64 optimized PSU" circuit. Ooops, I forgot that 7812. Ooops, 9+5=14, not 12. I even heard them fading away... :'( I have redeemed myself since by acquiring massive amounts of SIDs and sharing them around ;D

Don't say that! He might wire a ground line from the PSU to each encoder, switch and LED!!! ;) Essentially it doesn't matter, you can connect a button/encoder/LED pin to ground anywhere that's convenient... i.e. you can run a single ground wire through all your components and join it to any ground, or a more "starlike" arrangement with multiple wires connecting a group of components. My point was to share with you that there's nothing special about the ground for LEDs or encoders or buttons.... it's only when you're trying hard to separate analog and digital grounds (or audio and digital grounds) that it matters.

yeah I got some of those encoders a while ago... good price, nice feel... I think they're standard pinout, but haven't tested them yet. Be warned though, if you're pushing on a knob like the ones from ALBS and it's too tight, you might not be able to pull it off without pulling out the whole encoder shaft! It's not that big a deal as it appears to push back in without a problem... the encoder shaft appears to be a separate part with a pair of catches at the end to snap into a ring... this is unlike the other encoders I have.

If you follow the instructions for the LTC module, you'll get a new MIDI port available in MIOS Studio. It should just work like any other MIDI port. However, it might be cheaper/easier/more useful to just buy a USB to MIDI adapter.

If other displays work, then it's the display or cable, a short between the pins on the display, or a bad connection on the IDC plug maybe... but then it could also be a break on a track of the display itself, which is probably unfixable... your instincts are right tho... try a test app that prints ASCII 65 to whatever... look for a pattern.

Yeah the answers are searchable elsewhere, but off the top of my head, Link becomes a "Shift" button, CC & Edit by default are PageUp/PageDown, using shift key to get CC & Edit functions. If you're building "step C" for V2, consider: * a Volume "row" on the mod matrix under Filter. * an extra LED for LFO Waveform, called "Positive" (you can have postive-only waveforms). * MAYBE rename "Arpeggiator" to "Phase Offset" *** UNCONFIRMED!

... and look what I came across in another thread! Illuminated ones! http://www.tme.hu/microswitch-pcb-mounting-monostable-led-white/arts/en/a25/pb6141l-1__2.html http://www.tme.hu/katalog/index.phtml?f_szukaj=KS01-BL-3

Could have been a soldering iron ;)

Ich kann Deutch nicht, sorry... aber I can tell you that I did buy some of these knobs, months ago, the last of the "transparent blue" kind... "Drehknopf DK16-190V3 A.6/4,5 AT=14,5" for 0,90 Euro each. They are not as blue as I expected, more like smoky bluish-grey. These knobs are fantastic, they are perfect for encoders like the ones from Voti, aka. STEC16B as they cover the bushing (the threaded metal around the shaft). I actually saw a Waldorf synth and pulled the knobs off and it looks like the same rotary encoders I have ;D

Those are tactile switches, they're sold with the cap attached, they have a very clicky feel and sound, I personally have tried them in the store and didn't like the clicky feel and sound, but visually they look nice, a slightly rough finish on the top, and have good dimensions. Here's a link where I saw them: http://www.altronics.com.au/index.asp?area=item&id=S1094 But I also had a look at a tactile switch with separate button cap, these switches are practically identical to the ALPS part SKHCAEA010/SKHCAAA010/SKHCACA010 and have a much nicer feel, and the cap is waxy smooth plastic... and switch+cap is cheaper than the other ones... I'm leaning towards these for my MB-SEQ... http://www.altronics.com.au/index.asp?area=item&id=S1135 http://www.altronics.com.au/index.asp?area=item&id=S1138

1) No and no. If you look at the DIN and DOUT schematics, each shift register is connected to the previous one with the same connections as the Core->DIN and Core->DOUT connections. Also, the DIN and DOUT modules share the same SCLK and RCLK lines, so your detatchable control surface only needs six lines (+5v, GND, SCLK, RCLK, SI, SO). Now, for example, if you put a DINx1 and DOUTx1 in your "first" part, and connect the other DIN/DOUT modules at the end of the chain, the whole lot will work whether connected or not. Pay attention to the 10k pullups R33,R34,R35,R36 on the DIN module... these keep the input high if the next shift register isn't connected... there's a similar one on the Core. The DOUT doesn't need anything like that, as if nothing else is connected the bits get shifted into the air ;D Now I just thought of something, should you want to detect whether the "extra CS" is connected or not, reserve one DIN pin for this, picture it as a switch that's always pressed down... when the extra CS is not connected, that switch will be read as open, when connected, it will be read as closed. You might want to make use of that, so if you disconnect the extra CS, you aren't stuck in some menu, it will jump back to the main menu where your up/down buttons work ;D 2) Yes, you can probably use pins on J5, although it would be better if you used just one DIN module (just a single 74HC165) as this would give you the option of more buttons if you need them, or attaching a rotary encoder (preferable for changing patches to up/down buttons), and no special coding required. I would also suggest you try to put more buttons on the part with the LCD... especially if you're considering running MB-SID V2, where you may also need to switch the synth engine (Lead, Bassline, Drums, Multi) and the BankStick.

A bit of advice for those building things into 1U racks... after a short while you will get really tired of switching between the menus, i.e. jumping back and forth between LFO, Filter and Mod matrix, etc. You obviously don't have enough room for a "step C" control surface, but while designing, consider adding some "shortcut" buttons to jump between the menus (i.e. like Osc, Filter, LFO, Env, Mod Matrix, Arp, etc.).... in addition to adding five rotary encoders which can be assigned to parameters for real-time control, and the very cool PageUp/PageDown button functionality that TK just added in V2. If I was doing another MB-SID in a rack, I'd go with 2U minimum, a 2x40 LCD and 10 buttons underneath, five "assignable" knobs and a button to jump to most of the sub-menus. But that's just me ;D I would suggest to everyone designing a CS for MB-SID V2... build the "step B" on a bit of prototyping board with some cheap switches and a rotary encoder (you don't need the LEDs) and try it out... you'll come up with your own ideas about how to extend it into a "step B with extras" to fit your panel. (I suggest "step B", if you've only got one SID then "step A" is enough). No amount of thinking about what you will need can replace actually using it and discovering it for yourself.

You might want to try local suppliers of Winstar LCDs... judging from the part numbers and datasheets, Crystalfontz just sells LCDs made by Winstar.

Not that I know of... With some basic side-by-side testing, I've noticed a difference between 6581R4AR, 8580R5, CSG6582A (d/c 89) and CSG6582A (d/c 92) The 8580R5 and both kinds of 6582A have a MUCH lower volume click, everyone with them in a MB-SID would notice the difference on program changes. I've even modulated the volume register and if you're modulating between 50% and 100% volume (i.e. tremolo effect) you can't hear any clicks. What was interesting to observe is that the 8580R5s and d/c 89 6582As have identical output amplitudes, but the d/c 92 6582As peak a tiny bit lower. It's not an audible difference, but it is fairly consistent across the frequencies. I'm using the Spectroman in FL Studio, so I don't have any fancy graphs or figures ;D I just see two overlaid graphs with peak on the y axis, freq. on the x axis, and one line is just a little bit higher.

You can use just one IC, it shouldn't matter that the output (pin 9) of the one IC is not connected to another. Check voltages of pin 10 and 13. Take out IC and test no connections between adjacent pins of the socket, test connnection of pins 11, 12 and 14 to the right PIC pins. Can you explain more how it lights 2 neighbouring LEDs? i.e. what is the output on the LEDs when turning on just D0, just D1, just D2, etc.

Maybe you are grounding the OE pin instead? Usually if things work if you touch them ;D there's something wrong with ground, or something that should be ground.

I can't believe that was the problem... :o

The names of the pins on the port of the SID module are a bit confusing because it was originally used for a different module. I'll explain briefly... refer to the SID module circuit schematic. SID CS pin (port pin SO) is normally high, pulsed low to enable the SID and the SID will read the address and data bits. Port pin MU is the serial data bit (connected to SER). As port pin MD (connected to SCLK) is pulsed high, the state of MU is shifted into the shift registers, you do this 16 times to get 16 bits into the two shift registers, but they're not output yet, that's what pulsing RC does (connected to RCLK). You can see how O8 of one shift register connects to the SER of the second shift register, that's how the bits are shifted into both shift registers, and also will tell you what order you need to send the bits. So that's basically it, CS stays high, write bits in with pulsing SCLK, output them with a pulse of RCLK, then pulse CS low to get the SID to read the bits. Refer to the 74HC595 datasheet as well... but the names of the pins will be different.

I recall raising the sustain level after the gate is on will cause the sound to stop. If you're setting the registers in the order you wrote, then that might be why there's no sound. Set the gate bit last, if you're not doing that already.

I just noticed a bug in my code... the first two MIOS_LCD_PrintChar('.'); should be MIOS_LCD_PrintChar(' '); MIOS_HLP_Dec2BCD is what is used for all the MIOS_LCD_PrintBCD* functions, so it should work, if it does not then something else is wrong, maybe byte ordering? Perhaps try MIOS_HLP_Dec2BCD and then MIOS_LCD_PrintHex to dump the BCD values for a range of inputs, try to find out how it is wrong.

Turn the number into BCD (binary coded decimal) i.e. 12345 -> 0x01 0x23 0x45 then print out each nibble (four bits). This is actually how it is done in MIOS (but in assembler). The "if"s remove the leading zeros. Anyway, if you know what printf is, you don't need any more help than this ;D unsigned int x = 12345; MIOS_HLP_Dec2BCD(x); if ( MIOS_PARAMETER3 >> 4 ) MIOS_LCD_PrintBCD1( MIOS_PARAMETER3 >> 4 ); else MIOS_LCD_PrintChar('.'); if ( MIOS_PARAMETER3 & 0x0f ) MIOS_LCD_PrintBCD1( MIOS_PARAMETER3 & 0x0f ); else MIOS_LCD_PrintChar('.'); if ( MIOS_PARAMETER2 >> 4 ) MIOS_LCD_PrintBCD1( MIOS_PARAMETER2 >> 4 ); else MIOS_LCD_PrintChar('.'); MIOS_LCD_PrintBCD1( MIOS_PARAMETER2 & 0x0f ); MIOS_LCD_PrintChar('.'); MIOS_LCD_PrintBCD1( MIOS_PARAMETER1 >> 4 ); MIOS_LCD_PrintBCD1( MIOS_PARAMETER1 & 0x0f ); [/code]

;D It is, I was just joking... I didn't say they were 2.54mm compatible... only that you can make them fit into the holes of the standard prototyping board (not the stripboard kind which has really small holes).

Yeah that should work... What I was trying to warn about was, for example, using a 12v AC supply and putting that through a bridge rectifier (so you have 12v DC) and then supplying the SID module's 7812 with that, and then ALSO supplying the Core's bridge rectifier with the 12v DC. This would cause the Core's ground to be different (+0.6v) to the SID's ground and can cause logic errors (very wierd ones like what I had once).... and joining the ground between the Core and SID modules won't help... and is a bad idea. If you follow something like this: http://www.ucapps.de/mbhp/mbhp_4xsid_c64_psu_optimized.pdf then you'll see where the (1) label is, all the grounds are connected at this point... you can connect the SID and Core modules in a chain OR use separate +5/ground wires to each module, it shouldn't matter too much... and that's essentially what you're thinking... you're bypassing the bridge rectifiers on all the modules.

I suggested it can be done with one Core and 13 DOUTs (4 DOUTx4 modules) to drive 768 LEDs... but that requires writing your own code, you can make each LED turn on/off based on MIDI messages... While it is possible to do the same thing with linked Cores and lots more DOUTx4 modules, I guess you would still need to change the code, you probably cannot just use the MIDIIO128 application as it sounds like your need for 768 LEDs is more complex... You should explain more about what you want to do, how each LED should be turned on/off with MIDI events.