Wilba

;D I listened to this before I read how you made it, so I laughed my head off when the Beastie Boys jumped in, that's one really cool mash up.

Good news: Smash and I are talking about getting a batch of PCBs made, which he can sell through his shop. It is still early stages yet, as MB-SID v2 is not finished. I will tweak the design a little, add the CAN bus and modify the PSU section so people can mix SIDs if they want. The three DOUT and five DIN modules are enough to support a "step C" control surface with a switch matrix and LED matrix. The placement of the jacks/ports at the back can't be perfect for everyone's current case (if they are upgrading) but as TK is demonstrating, it's good enough for a C64 case. Keep in mind that a single PCB can never be perfect for everybody, but I'll try to make it fairly useful for the majority of people who've built their current MB-SID into a C64 case or rack case. As for full "kits": depending on the interest, maybe a batch of the control surface PCB might get made, I don't know yet... there's no plans to turn this into a complete kit package like the x0xb0x... but people are more than welcome to copy my example, get the same parts I used and make one like mine. I just have to finish mine first... ;D

No. I'm not interested in making kits, this is a one-off project which might be duplicated by others. I will release the Gerber and NC drill files (the files you need to get PCBs made) and then others can arrange their own production. It might cost as little as US$50 per PCB. I got them made with Gold Phoenix PCB, check out their "Special Prices". It can definitely use less than 8 SIDs, but this PCB is for 8580 or 6582 only (only 9v supply). In the next revision, I will replace the space allocated for IIC MIDI modules with a more flexible power supply section, so that 12v and 9v supply tracks run across the top. Each pair of SIDs could then be connected to either 12v or 9v. I found a batch of new old stock. I will sell the spares soon. Not intended, but as I've said before, the design will be available so anyone can order their own "prototypes" made. It's currently using MB-SID v1 firmware, each stereo SID module plays the same sound out of each SID, but due to slightly different filter capacitors, it's not always identical sound, a lot of phasing happens when you LP filter something like "Techno PWM" patch.

For info about "Wilba's PCB" ;D go here: http://www.midibox.org/forum/index.php?topic=8389.msg58396

With my latest MB-SID project in the final stages, and TK posting pictures of my PCB inside his C64 case, I thought I might post a bit of a blog about it. The PCB design is a prototype, only two were made (mine and the one I gave TK). There's lots of technical info in my the wiki pages: MB-6582 Wiki While I designed this PCB just for me, it's quite usable by others with C64 cases (like TK) so I'm happy to make it available for others to order their own made, but this will have to wait until after I do some revisions (added CAN bus, removed IIC MIDI modules). I'm just about to order the panels, and then I can finally finish this off!

Welcome! For a total newbie, that's great soldering. ;D A tip: when you get your desoldering pump, try sucking up some of the excess solder from those joints. A good soldering joint is a neat cone shape, not a blob. I strongly recommend a C64 PSU, this will save you a lot of time and is a "known good solution". Good luck!

WinPic800 doesn't burn PIC18F4685 (yet), so I've asked the name to add it. PICPgm already supports the PIC18F4685, but not the "ID Locations", so I've asked the name to add it. (Actually, I asked him six months ago and just nagged again! He's sent me a prerelease and it works well). Both are incredibly reliable and work perfectly with my JDM burner. http://www.midibox.org/dokuwiki/doku.php?id=jdm_with_winpic800 http://www.midibox.org/dokuwiki/doku.php?id=jdm_with_picpgm

Just to clarify: you have a bunch of holes on a PCB in 2mm x 4mm spacing and need to connect a ribbon cable to it? I would probably just take a strip of single row header pins (or double row, whatever was handy) and pull out the gold plated pin from the plastic and put them in the holes. Then use the crimp pins from a single row header connector without housing them in the plastic connector, maybe heatshrink each one. The other option is soldering cable to the PCB, but only a short length, with a female IDC connector attached, so you can still unplug whatever you're connecting.

My completely biased opinion ;D Start with (invest in) one 8580 and get it going. If you're scavenging from a C64, you're more likely to get a perfectly working 8580 than a perfectly working 6581. You might find the sound and filter of the 8580 sounds great to you, nice squelchy basslines, less noise than the 6581, etc. Then once you're hooked on the SID sound, you might try out a 6581 for comparison. In other words, you're better off getting one of each before deciding the final choice of all your SIDs, so you might as well start with an 8580 and enjoy the "303" mode!

These knobs are the same ones used by Waldorf. I've bought these knobs from ALBS.de, refer to previous discussion: http://www.midibox.org/forum/index.php?topic=8190.msg56745#msg56745

Excellent work, TK! You can programmatically turn on 4-bit mode in the application's initialization, but it's not a configuration bit in the PIC ID header. Can you add a bit in the PIC ID header so MIOS can use 4-bit mode on any PIC, not just the 18F4685? (or use 8-bit mode on the 18F4685, if one didn't need CAN?) This hardware change has just brought to my attention that I could use a cable with four less wires and would like to start doing that now! ;D

Probably not ;) As far as I know, the specs on the DOUT have not changed, "bypass capacitors" have been added to the schematic and newer PCB revisions, but it's still "compatible". There was a change in the wiring of the Core with MIOS 1.7 http://www.ucapps.de/mios_v1_7_hardware_mods.html This probably is unrelated, but you did mention different versions of MIOS, just confirming you are aware of it. It is really strange that one Core resets when it gets note 18, perhaps narrow down why that happens first, i.e. does it happen even when there is no DOUT connected?

I saw that video a while ago and wondered why you had never posted here... not even some photos of your finished MIDIbox, which looks really cool from the video. As for your DOUT problem, start with checking the wiring between Core and DOUT module... test a pin on the PIC is actually wired to pin(s) on the shift registers. Then maybe try this? http://www.ucapps.de/mios/srio_interconnection_test_v1.zip But this sounds like the shift registers are actually working, and when MIDI note 18 is turned on, the shift register output pin that should go on is actually shorting to ground. Hmm... Try unplugging all the resistor networks and testing if the shift register outputs go high/low as expected by your firmware (using a multimeter). See if the resetting problem still happens with the resistor networks unplugged. Also, test each DOUT module individually. I suppose it's unlikely you've put the resistor networks or the shift registers in the "wrong way" ;D but check that too. Finally, you might want to check pins 10 and 13 on each shift register, check they are connected to +5v and ground respectively. If you've soldered a PCB from SmashTV, it's probably not a PCB fault, maybe just a bad soldering joint?

Reminds me of another wish I had... My wish for the V2 patch format is that there is space reserved for future features coming from TK as well as some space that is totally user-specific and usable by people adding their own features. The CMOS switch control is a great example... as what you're switching might be filters or effects or a noise gate... and all people need is a place to store those bits that won't cause any compatibility issues with the preset patches or with other people's patches, with patch editors or patch loading/saving utilities etc. Another good example is a "Volume" modulation target, complete with an 8th row of LEDs in the mod matrix, that's what I really want, whether TK puts it in or not ;D so it would be really nice if there was space in the patch format to store those eight bits.

If it's possible, try mounting the switches onto their own board, and then put spacers between the two boards. However, if you've got a lot of individual switches and not a lot of grouping, this probably won't work so well.... maybe even invert the idea, mount all your switches to one board and then mount the taller components behind it, poking through some holes/slots.

Wilba's MIDIbox Debugging Service Debugging your MIDIbox from across the globe with only the powers of his own mind! If you have found this service worthwhile, feel free to an amount that reflects how happy you are that your MIDIbox now works! (Note: donations are not tax deductible)

The MB-SID should boot up in the "Top Screen": and turning the rotary encoder should change the patch. You don't need to press the menu button. You might have the wiring of the buttons and encoder wrong somehow, or maybe you're missing a resistor on one of the DIN inputs, or missing a resistor to "terminate" the chain of shift registers (see R33-R36) or maybe even forgot to ground the CLI pin (pin 15) on the shift registers. (OK all this suggests you hand-made your DIN module instead of using a PCB).

I've changed patches without an LCD before, the lack of LCD is unrelated I think, but makes debugging harder. It has happened to me before that by plugging in a BankStick for the first time, it got stuck on Bank H, patch 128... that bug should have been fixed already :-) The 6 tone tune might be the Zelda patch which is patch 128, again, sounds like it's stuck on the last patch. The "startup" tune is a very short rising tone, less than half a second. I suggest trying to change the patch using MIOS Studio... the virtual keyboard can change the patch (see "Program" spinner). Also try switching to Bank A, from the spec, sending: F0 00 00 7E 46 <device-number> 0B <bank> F7 will change the current bank (you'll probably just use "F0 00 00 7E 46 00 0B 00 F7") Switching to Bank A will store in the PIC that you're using Bank A, so next time it should boot in Bank A (in case it was doing wierd things and stuck in a non-existent bank!) Same with patch number, it should remember what patch you used last time when you boot it. The "default" patch is stored in the PIC, and it's no longer accessable after you plug in a BankStick.

I don't really know... I suppose in theory you can make the AOUT work with +-9v but the voltage output range is also reduced, so no good if you're driving gear that needs -10v to +10v. I don't think digital noise is a problem because I don't get any with my SIDs, where digital noise is more noticeable than with AOUT... i.e. for the SID box the 9v (analog) and 5v (digital) grounds are joined at one place, close to the power socket. So similarly, splitting AOUT power from Core power at one point as near to the transformer as possible should be fine.

I would also suggest 22nF capacitors for the 8580 and 6582 instead of 6.8nF. Seach for other posts about this topic, in short, 22nF will give a better frequency range with less stepping at the low end.

I will have to disagree ;D Isn't 1 amp of 5 volts enough for two LCD backlights and a Core? My "black" C64 PSU brick is rated at 1.5 amps for 5V and 1 amp for 9V AC. There are two "windings" on my PSU's transformer, though I haven't tested it, I'm assuming they're identical, both delivering 9V AC, one 9V AC output goes direct to the plug, the other goes into a 7805 regulator with a HUGE aluminium heatsink. In my MB-SID v2, I rectify the 9V AC into 9V DC and join the ground with the 5V DC ground, and have had no problems with it at all. So I don't see any problem with using a half-wave rectifier on the 9V AC and getting +/- 9V DC for your AOUT, and the 5V DC to power the Seq, and joining the grounds together at a central point (you want to keep your AOUT's ground well away from the digital ground of the Core and other modules). You can effectively treat the 9V AC and 5V DC outputs as separate, as if they were coming from two separate AC adapters. Something like this ought to be adequate (taken from the PAiA ribbon controller circuit): That's the general idea - feed your AC into two diodes so that your matching pair of positive and negative voltage regulators each get the positive and negative "phases" of the AC power. Each regulator only gets half the power, so the current output is half (probably less) for each regulator, i.e. for 1 amp 9V AC, you can get +9V at ~500mA and -9V at ~500mA. My advice is to increase the capacitors C6 and C7 to be big, like 1000uF - 2200uF perhaps... this will ensure there's enough voltage stored in the capacitors during the gaps in the phases to supply the regulators. Remember to use POWER DIODES not those tiny signal diodes, you want the big black ones that can handle the current. The beauty of this idea is that you can use any old AC adapter that outputs AC (not DC) and get a symetrical DC supply out of it, without needing to buy a transformer and wire it to mains power. I can't tell you if this is better or worse than a symmetrical PSU built like the one presented here, but it should do the job and will be easier to construct.

No, it was US$80 shipping, it is international courier only. >:( Thanks for the offer, nym! Much appreciated. PM me and I'll send you some dough via paypal.

Yeah, cheap, but only ships to Australia by UPS for $80. Bugger. Any chance you can buy some on my behalf and post it to me?

bill: With a step B control surface, you can hold down one SID button and then press the menu button and it toggles playing a note on that SID.

I'm currently trying to make my own ribbon controller but haven't found a good material to use. In my searching, I came across Velostat, and if it's the same stuff in which my PLED display was packed, then it might just work well... but haven't found a supplier.