ImAlive

You should propably not use any electrolytic caps in AC situations, the can "take" only 1 polarity and may get damaged or even explode if voltage is applied in the wrong way. A ceramic cap in series (parallel wiring might give you plenty of DC and be pretty pointless - Lowpass) with the signal line should block DC good enough. If touching the ground changes stuff, something else might be touching the ground when it shouldn't, so check for shorts. When checking with an osc., be sure to have DC adjusted to zero before checking, and that all grounds are wired properly.

Wattage on Pots means max. loss power generated in the part itself. You get this by multiplying current through the pot with the voltage over the pot (this depends on the resistance of the bulb AND the pot) - good old P = U * I. So, if you drive 2 Amps through a Pot which eats away 1 V out of 12 (2W generated), and the pot is rated 0.6W, it would probably not stand the heat.

From what I've heard, this should be possible. You'll need 1xAIN, 1xDINx4, 1xDOUTx4 (not quite sure about that, depends on the way to drive the LEDs), and obviously 1 CORE. For instance, the MIDIBOX64-Application should handle this ("ghost controllers" will always stay at zero when grounded). There might be some work waiting, depending on what you want these controls to do ;) But maybe you should give us more details on your plans, so that your question can be answered in a better way, and maybe by someone who has more midibox-knowledge than me 8)

It might be better to employ fuses. Like 20 Amps output current can melt quite some stuff...

If your voltage on the PC-PSU is too high, you should put more dummy load to this line by means of connecting a power resistor, if you run the PSU while not installed in a PC (most PSUs won't run without load anyways). Another issue from PC PSUs is noisy voltage which might create AIN jitter, be sure to use bypass caps.

MIOS ist das Betriebssystem, die jeweilige Anwendung die dann auf MIOS läuft bestimmt, was der PIC letztendlich mit seinen Ein/Ausgängen tut. Beides wird über MIDI (per PC) aufgespielt, den Brenner brauchst du nur für den Bootloader, der ja schon drauf ist - also benötigst du den Brenner gar nicht :D (was besser ist, denn die Dinger funktionieren nie wirklich zuverlässig...). Zum Aufspielen empfiehlt sich http://miosstudio.midibox.org/. midibox64 ist hier: http://www.ucapps.de/mios/midibox64_v2_4.zip Und es ist immer ganz gut die Hilfeseiten von ucapps.de zu lesen, denn da steht das eigentlich alles drin ;D

Die Fader werden als Spannungsteiler benutzt, daher ist der exakte Widerstandswert nicht soo wichtig, empfohlen sind 10kOhm. Du musst erst noch MIOS und deine Anwendung auf den PIC spielen. Der Bootloader "erwartet" einen MIOS-Upload per MIDI (näheres dazu auf der entsprechenden Seite von ucapps.de). Wenn MIOS aufgespielt ist, kommt deine Anwendung, in deinem Fall vermutlich MIDIBOX64.

Oops, I was wrong wrong wrong!! (Damn memory...) The caps go between pin 16 and one out of 6,7 or 8! So that would be the topmost RIGHT and lowest LEFT pins, viewed from the top. Here's an image, as viewed from the solder side:

Just stumbled across this when completing my core. I use only one AIN module, so I thought that the other muxed AIN inputs on J5 of the core (A4...A7) should be clamped to ground, too. Maybe this helps... and if this is the case, then it should be mentioned on the core page in the J5 description, right? Right now I'm about to power up and check those cores... so excited :D - aaand, the cores rush to life, all voltages fine 8)

Yeah, most commonly it's just diagonally between the topmost left and the lowest right pin (top view). You can also buy sockets in which these bypass caps are already contained. And as illogik posted, there are pics on the module page. (If I may add this, soldering stuff on the bottom side of the board sucks ;D )

Normally, this is done by cutting into the copper trace with a sharp cutter blade until it is through at the location where you want to separate the trace. You should check that the traces are isolated with a multimeter to be sure afterwards, though...

This sure is easy, but you won't be able to do a CORE module layout with this ;D You just "draw" your traces directly onto the copper-plated PCB with a resistant marker-pen. Then you throw the PCB into your etching solution (like FeCl3), wait, take it out, clean it, remove the ink with alcohol, and ta-daa ;) I did my modules with the "Iron-on" method, as mentioned in the wiki.

So, a little overheating does not kill the PIC, which would mean that chances are good my 2nd one ist still alive. And yeah, I usually unplug the programmer and short the capacitor before inserting a chip. I tried out some combinations, but not really every one imaginable. I switched to a 220 Ohms resistor like you did, tried different caps (as a next step, I would have tried a "bad ass" cap). I just used 0.7 on Linux, 'cause under Win readcfg gave back all FF. I don't know, will the port mode change something? It was set to EPP/ECP in my previous tries, maybe I should try SPP? Perhaps some geek at my university owns a PIC burner, I'll try asking around, my motivation with self-burning is somewhat down... Regarding free samples, I think these are not intended for end users, and I'd feel bad about lying ;) besides of that, I have no idea of what hassles customs would produce.

Sorry for doubleposting... One of my PICs is dead. It died a violent, heatful dead and I'm not quite certain why this happened. It just went *snap* *smoke*. At that moment, I was checking voltages with portdiag with PIC stuffed in (the voltages were OK, but can you do this with PIC in place?). I had done this before without problems. My suspicion is that I might have reversed the polarity of the 9V supply, but I can't be sure. Now you can see 6€ of dead silicon with a burn stain lying around here. Before the PIC died, I tried different combinations of batteries, capacitors and resistors, but it never got recognized, even after 100 tries... Right now, I don't have much more motivation left to fumble around with this burner. My case is built, the faders and pots are about to be mounted, all modules are built, all I need now is a working PIC (and I'll have to go and buy a fresh one. Yeah, real cool. Although it is almost certainly my fault, a swabian guy hates to waste money :P) and all I would want is to go and tweak those knobs, but I'm stuck in a really annoying phase >:(

You really tried like 30 times without changing the setup between the tries? Well, that's something I could do 8) Do you have to run erasepic before you try readcfg? I didn't do that. Whew, everytime I plug that thing in, I'm afraid of frying the PIC or my PC :-\ Getting it out isn't that hard, just use a screwdriver as a lever and some caution. EDIT: Damn... my voltages are dropping like hell! I connected everything, and measured the parallel voltages of all pins with the PIC inserted. Well, my LPT supply drops to 0.8V with PIC in, and the batteries go down to 1.7V (not sure if they are all brand new). 9.5V stays stable. But when changing voltages with portdiag, high is just 0.8V when low is 0.3V! My PIC sucks voltages :-[ I even thought the PIC might get *slightly* warm, but this could as well be imagination. Hm. I checked the circuit for shorts with the multimeter, everything seems to be in order... maybe as a next step, I'll measure the current through the PIC... EDIT: I am such a d*mb idi*t ;D soldered Vdd to pin 10 and Vss to pin 11 *smashes head on table* and if that wasn't enough, the power pins on the right side are shifted up one pin, too. Man, what was I thinking... I hope the PIC survived this... luckily I don't trust myself and did only very short power-on phases... EDIT again: Fixed that wrong pin business. Now, powering the PIC from LPT pulls the voltage down to 3.37V, which isn't so good. Plugging my battpack on the pic results in a voltage of 4.7V - great! When I'm back later, I'll try readcfg... (managed portdiag to work under windows, but readcfg gives all ff, lets see what it'll say with PIC in place).

Well, well... regretfully no instant-success-story here. First try: Serial resistance 330 Ohms - Foil Cap 0,1uF - MCLR from wallwart, 9.5V - Vdd from batts, 4.7V - 10k omitted - Linux Can you imagine how hard it is to chain 3 AAA batteries when you don't have a socket for this? ;D The PIC socket from Reichelt didn't fit, the legs of the PIC were too far apart to fit nicely. So I had to bend them a little. Even with this done, I think I didn't insert it correctly in this first trial, the legs were more "loose", which is why this may have failed. I'll try again tomorrow. All the voltages were switchable and ok with that portdiag. But PIC was not recognized. Second try: Serial 220 Ohms - Elko 220uF - MCLR 9.5V - Vdd via LPT pin 36, 4.5V - 10k omitted - Linux This time I snapped the PIC in correctly, though it was pretty hard, I thought I just might break it every second, so much force was applied. But the results were the same as before.

Congrats tarzan boy ;D All right, in that case I'll just keep the ~9V. And I mounted the 10k-pulldown on RB5 too, time to remove it if it hinders success ::) Ah, and since this was mentioned before, in my setup, I omitted the D2 and D3 lines. We'll see if this works later.

Just stumbled upon this, and maybe, I won't have to build a JDM ::) The board's ready, just measured my port, it reads 4.5V. The manual of broccoli suggests to use the same voltage for Vdd, which makes sense, hence the logic out of the PIC cannot go higher than what it gets by the source. For 4.5V, I calculated 300 Ohms (and used 330). Now, 9V is only needed when erasing the pic, right, since it's MCLR? Won't it be better to remove that voltage when actually burning? And for wallwarts, I tried out some lying around, they suck :o one even delivered 8.5V when it should have 4.5V :o ok ok, without load, but still... For the other, is 9.45V a ok voltage to erase? I think so, since any up to 13.5V should be ok for the PIC. Let's throw in some live-distro and see if we can burn this...

Pull-up-resistors are mounted between your logic "high" supply (e.g. +5V) and the pin you want to "pull up". This puts the pin at potential +5V and ensures a defined "high" signal. You want to keep the resistors at high values to have only small currents go through them, so that your supply won't drop in voltage. Same goes for pulldown resistors, just that they connect the signal pin to ground (0V). For analog inputs like on the AIN, as well as the multiplexed lines to the pic, you just clamp unneeded inputs to 0V since there won't be no current to worry about, it's just that they stay at 0%. So, if you want something to be always "high" (inside the defined high voltage), you add a pull-up, at least that's what I think.

Just made a mistake when soldering my two cores... since I'm also building a JDM, I accidentally soldered the BC547C for the JDM on the cores instead of the BC337 :-X Now my question, can I just leave the BC 547C on the core, or do I have to remove it (which will be difficult, because I already bent the legs of the transistor :-\ ). In the schematic, the 547 is mentioned alongside with the 337, this is why I thought this could be possible. Which would of course mean that I would have to go and buy another two 547 ::) EDIT: When we're on the subject: Could I even substitute the 547 on the JDM for 337? 8) EDIT2: Solved. I googled up datasheets, and since hFE and current ratings aren't compatible, it's "Go buy a desoldering pump".

Just etched the 1st PCB in my life... pretty crappy, but I think it'll work ::) I did it by the "Iron" method: Printed the layout (AIN) on Reichelt catalogue paper with a laser printer, and ironed it on the copper board (well, I still have to improve my skill with that technique, the tracks are a little fuzzy, and I had to scratch off quite a few short ciruits, but now I think I'm fine - the toner did not stick to the copper very well and came off partially when removing the paper. But luckily, water-resistant pen worked fine for masking. Maybe I should prepare the copper better before ironing, and apply more heat?). Then, I threw the board into FeCl3 (250g into 500ml of water, in a plastic bowl contained in a warm water bath, a little too cold in the beginning so I had to pour in some more hot water), and after 20 minutes, the layout looked pretty decent (Ferric Chloride: awful stuff. I looked pretty funny I think with that goggles and gloves and stuff, but you better be safe...). Removed the toner with nail cleaner (didn't have pure acetone around), but now it sticked pretty hard to the surface, heh. And done! The small letters (MHBP_AINX4...) are well readable, so with some practice (and a masker pen), I think this method is suitable. Learn from your mistakes and have lots of fun, that's what DIY is all about, right? 8) Okay, now that this worked, I'll iron my next 3 boards (2x core, another AIN, since I'm building 2 boxes in cooperation with a pal). EDIT: The 2nd AIN came out really good. The problem is the cores. I had quite some trouble with those, hence the toner didn't stick good now as well. Bigger boards are harder to make, using this method, I think. Also, someone mentioned that HP toner doesn't seem to be suitable for this, and as I used HP toner, I might agree with that. But for now, the cores at least seem usable (no disconnected traces, but quite a few shorts and fuzzy lines).