c0nsumer

Ah, here's the filters I have: http://www.grainger.com/Grainger/wwg/itemDetailsRender.shtml?ItemId=1611746862

I'll continue this conversation here. But yes, I believe that all you have to worry about is the epoxy and fiberglass dust. Try looking up the MSDS for FR4 PCB material for more information, but I don't think you have too much to worry about. -Steve

Check your PM. I just replied. I use a simple dust filter in replacable cartridges. PCBs are made from epoxy and fiberglass, with a laquer coating on the outside. It's really not anything too harmful unless it's burning. In my experience I've found that simple dust cartridges stop all the dust and irritation to my lungs. It's weird to take the mask off and suddenly be hit with a wall of bad smelling PCB dust after cutting. The pads which go into the dust filters are pretty cheap too, so you can replace them whenever they start getting a bit blocked up. Also, be sure to wear safety glasses. The dust will also get in your eyes, and the little chips can be nasty.

By the way, thanks for mentioning this rate. I had been curious about it, but I never bothered to actually check into how often it scans the encoders. -Steve

Eagle is pretty easy to use once you get into it, and the non-profit license isn't too pricy. The smaller board non-profit version is completely free, and many things can be done with that. -Steve

By the way, how's your 9090 going? Got any photos of it? I started stuffing the boards last weekend but took a break for some brewing and to finish the MIDIbox SID-NUXX. What are you doing for an enclosure? -Steve

I don't have all the final info up yet, but I'm pretty excited about this so I thought I'd post now then add more later. My project, the MIDIbox SID-NUXX is finally complete. What you see there are my two MIDIbox SID-NUXX's, one with a 6581 and the other with an 8580. The red box in the middle is a friend's 6581-based one, but he hasn't yet brought the SID by for me to plug it in. Everything works great, and they sound really nice. I'll be finishing the documentation at my site tomorrow, with any luck. Then it's on to the next project. It'll be either the 9090, Casio SK-1 MIDI Mod (which I received the PCBs for today), a PIC-based Nixie clock, or maybe simply taking a bit of a break for a few days. While I learned a ton on this project, it was a lot of work. Almost comparable to back when I built a full stand-up MAME cabinet in 2000, only smaller. Anyway, that's all for now. I'll add more info once I have it, but for now it's bed time. -Steve

Thank you, and you're welcome. The only restriction on it is that boards made from it cannot be sold commercially. The the MIDIbox SID thread later tonight or tomorrow. I finished up all my MIDIbox SID-NUXXs tonight and I'm taking photos now. :D -Steve

Yes, on both counts. It's well within spec. It passed just fine. I figure that if that were the problem there would be more errors than that, especially around vias and some of the finer traces. I've actually got a total of ten other PCBs here using that same part footprint, and for all but two of them I've confirmed that they are made properly. The other two haven't been confirmed yet simply because I haven't had a chance. I really think this is a defect. As soon as I'm absolutely certain that the other two PCBs are good, I'll contact PCBFABEXPRESS about a replacement or whatnot. -Steve UPDATE: For what it's worth, the rules they publish are here. They say 6mil and I checked everything to 8mil.

It would be better to get things fixed so that if you need anything changed in the future you can compile it yourself. What about it isn't working?

Have you tried to do it yourself? It's very easy. Just install MPLAB, fire up MPASMWIN, point it at main.asm and click 'Assemble'.

Here you go: MIDIbox SID-NUXX - Mainboard#Design That's Core, DINX1, BankStick, LTC, SID, and IIC MIDI modules and a power supply all in a single schematic, with full board layout. And yes, these boards have been tested to work. It's what is used in the MIDIbox SID-NUXX. -Steve

Anyone want to see what a rather frustrating PCB manufacturing error looks like? That. Right there. This is almost too small to see with the naked eye and was located on the top side of one of my MIDIbox SID-NUXX's mainboard PCBs, beneath the power socket. (Look here for a scan of a good PCB, the power socket is in the lower left.) Unfortunately I found the error after one of my 12VAC wall warts mysteriously stopped working while I was doing testing / final assembly on my 8580-based device. With wall wart power supplies being relatively cheap I figured the one had simply died, so I tested a second, hooked it up, and it too died. A bit of poking later and I found that one of the AC lines were being connected to ground. This had to happen at either half the bridge rectifier, the power switch, the power connector, or something somewhere in between. Since no part of either the power switch or the power jack (SW1 and PWR1, respectively) connect to ground, it had to be a defect. Nothing was visable, so I lifted the power connector. A little blurryness was visible between the groundplane and annular ring, and the above macro photo confirmed the defect. Of the five PCBs I had made two are known good, this one has this error, and I've yet to check out the other two. This error should be readily fixable with a Dremel and a dental bit. I just hope there aren't other, more hidden errors.

I ended up ordering the PIC board for a couple of reasons... One, everything I've done thus far has been PIC related. Two, I've got some spare PICs and programmers and such laying around home. Three... It's what I originally intended. I can always get the AVR board later if need. Hopefully it'll be pretty nice... It looks like a good all-in-one experimentation platform. I also added on the ethernet module, the RTC, and the CF reader.

This is probably a pretty hot topic, and is somewhat off-topic for the board but I'm going to throw it out anyway. I'm looking at picking up one of the MikroElektronika dev boards, either the AVR or PIC one, and I'm not sure which to get. The reason I'm not sure is I really don't know which I should start seriously learning, PIC or AVR. I know that the MIDIbox stuff is all done in PIC, but lots of other MIDI stuff is also done in AVR. I've also been Googling around about this, and I can't seem to find a reason to pick one over the other. So I guess what I'm asking is why do you (whoever you are) prefer one versus the other? Is it just what you are familiar with? Is it the (from what I've been reading) more straightforward ASM in AVR? Is it what your job requires? Or... what? Eventually I'd like a better understanding of both, but I don't want to try to learn both in parallel. That'd be stupid. Thanks very much... -Steve

I still disagree. If you have a properly mounted PCB with switches on them, it will feel incredibly solid. Case in point, the step / keypad switches on the Sequentix P3: http://www.nuxx.net/gallery/v/stuffivemade/sequentix_p3/in_progress/IMG_4962.JPG.html http://www.nuxx.net/gallery/v/stuffivemade/sequentix_p3/in_progress/IMG_4965.JPG.html With proper support for the PCB and proper quality switches things can be as solid as panel mount. PCB mount can also provide closer spacing between devices, as there isn't the need for a flange around the parts needed to attach to the panel itself. -Steve

One other thing to remember, especially when reusing the tin plating solution: Be sure all the crystals are re-dissolved before dropping the board in. If not, one can get what I can best describe as odd results, where certain parts of the board don't have a shiny plating. Also, polish the plated parts with a paper towel and ammonia afterwards. If I remember the reasoning correctly, this dissolves some of the remaining stuff on the tin plate besides the metal itself preventing it from hazing over down the line. -Steve

It also allows for more attractive designs. This isn't a problem if the PCB is properly mounted to the enclosure. If the enclosure is cheap to begin with and things aren't generally well designed, it will feel flimsy. But yes, this is generally the fault of cheaper things. Compare to better designed stuff (as I'm sure you have) and you'll see that it can be done right. -Steve

Here here! Handsome is as handsome does A man's reach should exceed his grasp. Every little helps A proud look, a lying tongue, and hands that shed innocent blood. Corporations have neither bodies to be punished nor souls to be damned

Why wasn't this just deleted instead?

I've actually been thinking of that as sort of a 'next step'. For now I just wanted to use something where I didn't have to do any new software or electrical design. That said, I think that it would be really neat to have a device which would allow MIDI IN and OUT to happen with the SK1, but because the keyboard matrix connections would be used for it, any in would trigger an out, and there'd be a loop. I'm not sure how to work around that yet.

Yes, because this was already done. I just wanted to reuse the existing design. Laying out the board only took a day or so. There'd be no reason to use a MBHP version if all I wanted was remote triggering of the keys. If I were to do more, I'd probably put it in a new enclosure and such...

This should help you out: http://www.nuxx.net/wiki/MIDIbox_SID-NUXX_-_MIOS_Changes#mios_tables.inc I'm using some Alps EC12E24204A8 and Pin B goes to the even numbered input, and mios_tables.inc needs to be set accordingly. Pin A then (obviously) goes to the odd numbered input, and C is common and goes to ground. Hopefully it's the same for yours. -Steve

Why don't you just buy one from a supplier? Samples are generally intended for people who will use more in the future. If a bunch of people request samples and then don't follow through with orders, suppliers will tighten up their sampling, hurting students and those developing open source projects. -Steve

Ah, I could understand that then. What about allowing it to be a more traditional wiki where anyone can edit without an account? Vandalism doens't happen very often, and if it did, self-policing should take care of that. With the essentially eternal history, rolling back changes is trivial.