Wilba

As nebula said, the 7805 you mounted will not be connected to anything because that is not supposed to be mounted in PSU Option B. Basically you need to take out all ICs (except 6N138 optocoupler) and only put in one PIC on the board at one time. Since you know the PIC is good in Core 1, you can then test what is different when you put it in Core 2, 3, 4 etc. Once you are able to put the "known good" PIC (with ID #0) into Core 2, you can test MIDI In, change the jumper in J11 to test MIDI Out, test LCD etc. Get Core 2 going first, then move on to Core 3 and then Core 4. Good troubleshooting involves testing what you know works in something that doesn't, in this case, moving the working PIC into other Cores. Then you can find the problems, in this case, it is probably bad solder joints in that Core, but it could be other things. It also makes things easier for people helping you because you only need to report what does and does not work with one small chunk of the problem (i.e. one Core). What you learn fixing one Core will probably help you fix the other Cores by yourself :)

http://www.voteearth2009.org/home/

OK a few things: There are no parts called IC14 or IC25... you're using this to describe a pin on a part but that's not how you do it... the real part I assume is U1_CORE1, U1_CORE2, etc. If in doubt, refer to the PDF of the PCB (http://www.mb6582.org/plans/MB-6582_Base_PCB_R2_Color.pdf) so you can tell us exactly which part and pin you are testing. Pin 25 is the Tx pin. Depending on the position of a jumper in J11, that switches which of these pins is connected to MIDI Out. When the PIC is not transmitting it will be 5V, when it is transmitting one "bit" it will be changing between 0V or 5V. Pin 14 is one of the two pins connected to the crystal. I don't know that much about crystals, but I would assume the voltage there oscillates somwhere in the range 0V and 5V and is never constant. What you measure there will be an average, and probably affecting the oscillation (clock) so the PIC might not work. You may have been confused by the MB-6582 testing instructions... Note I refer to testing pins 25 and 14 on the SID, not the PIC. The PIC pins to test are 11 and 12 (also forgot to add 31 and 32). My advice: Take the PIC that you know works in Core 1 and put it in the other Core "slots", and move the jumper in J11 to match (route that PIC's Tx pin to MIDI Out). Power on, and look at MIDI In in MIOS Studio. If you get the upload requests, then that PIC is booting up OK and MIDI Out is working. Assuming it has MIOS installed already, if an LCD connected to that Core shows "READY." then that PIC is happy. Try sending it LCD display commands from within MIOS Studio to prove MIDI In works (all Cores are connected to the same MIDI In, but you can change "device ID" in MIOS Studio to address MIDI to a particular PIC, NOT a particular Core. So it makes no difference if PIC ID#0 is in Core 1 or Core 2, apart from what's connected to that PIC I mean...) Run through those tests and report back. And probably paranoid but please check you know pin ordering on an IC, so we're clear which pins you really are testing :)

Crash

6582A SID MEGA-SALE #9 will happen in April WILBA HAZ MOAR SIDS SOON. THERE IZ NO WAITING LIST EMAIL YOUR ORDERING DIRECT TO ME YOU CAN HAZ INFORMATION HOW TO EMAIL THE ORDERING BY CLICKY THIS LINK ----->>>>> [url=http://www.midibox.org/dokuwiki/doku.php?id=wilba_6582a_sid_mega_sale]HERE <<<<<----- LAST SID MEGASALE EVAR!!!!!!!!!!!!!!!11111 DON"T DELAY!!!!!!!!!!!!!!!!!!!!!1 ONLY FIVE HUNDRED FOR FAST SELLING RIGHT NOW!!!!!!!!!!1111 WEN PEOPLES FIND OUT THEY WILL BE SELLING LIKE THOSE THINGS THAT SELL REEL FAST I FORGET WOT THEY ARE, SOME KIND OF PANCAKE BUT SOME PEOPLES THINK THEY ARE CREPES I DUNNO LIKE THEY ARE REELY GONNA SELL SO JUST HURRY TFU AND DON"T MISS OUT ON THE SALE OF THE SENCHARY!!!! BUT YOU GOTTA PAY WHEN I SEZ SO OR I CUM ROUND TO YOUR PLACE AND BEAT YOU UP, DIG? I KNOW WHERE YOU LIVE COZ YOU TOLD ME ALREADY, DIG? MAYBE I GETS SOMEONE ELSE TO BEAT YOU UP< COS LIKE I"M A VERY BUSY MAN< SO MANY SIDS SELLING ALL THE TIME..... CYA!! WILBA

Of course it's OK, it's not like you're printing MIDIbox logos on T-shirts or something... :rolleyes:

In the theme of the other forum game, let's start a new one, this time using only images and videos. Each new post must be related or a response to the previous one and must contain an embedded link to an image or a YouTube video. Title is optional (but might help us understand how you get from A to B). In honour of the original game, I'll start it with the same thing: Beer :) Have fun!

I ordered ¾ but got 1 PCB! Bonus! Thanks nILS!

DIN 1451, 1-stroke :) At 2.5mm text height, 0.2mm engraving width, it looks pretty good. If you increase the text height, you could increase engraving width. Just be careful because paint-filled engraving comes out just a little bit thicker than it looks in FPD, and you want to preserve gaps between the strokes in letters. In FPD, if text looks thin at 0.2mm and just right at 0.4mm, then set it to 0.2mm and it comes out great.

20x2 is only the minimum LCD size. 20x4 is also supported with extra features (TK has put extra info in the 1st and 4th lines, the 2nd and 3rd lines are the same as you would see on a 20x2). Also, 40x2 is a good option if you want to see more parameters per "page". (good if you are not building a full control surface with lots of buttons and knobs).

what about single in-line (SIL) female connector: and solder the other end to a SIL header: (pics courtesy of SmashTV)

Also consider those IC sockets with machined pins. They would sit above the PCB a bit more than an IC, giving more room to do a proper solder joint on the top side.

OK I've waited far too long for people to reply. I'm starting the orders now for the people who confirmed: j00lz Futureman 3rdordertrauma SubContractor jimbob madox andyj2 goldbug Thirstmonkey Widdly bron84 Son Green bron84 casius Scipher drboom This bulk order is now closed.

Protuberant LEDs are easily done without blindholes. I'm not even sure how blindholes would solve it. All you need to do is solder the LEDs with the panel attached to the PCB, and panel side down on a flat surface, so the LEDs do not protrude past the panel. This will work with flat-top LEDs either manufactured or self-made... and why I said getting them exactly the same height wasn't that important, because they would align with the panel anyway.

Yes, easily done with an emery board, or sandpaper. For best results, use a LED in a diffused package, not a waterclear, although the sanded top surface of a waterclear will diffuse a bit, it's not as good as a fully diffused (semi-translucent) package. Also, getting them exactly the same isn't that critical if you're setting them at the same level as your panel anyway... within 1mm difference would be fine.

if it's REALLY low (barely audible) it might be the 9V (or 12V) supply to the SID itself. Using the testtone app, can you test just the audio buffer circuit by taking the SID out and bridging pins 8 and 27 with a 1K resistor? This will generate a 1kHz square wave tone (same as the testtone app makes with the SID) by toggling pin 8 at 1kHz, which is thus connected to the audio buffer circuit at pin 27 (the SID's audio out pin). If the volume of this test tone is the same as the other SID module (you can test both at the same time), then the audio buffers are probably not the problem at all, and it is something wrong with the SID in the socket, like the voltage at pin 28 is not right. I recall someone else having this problem, they didn't put the jumpers to select voltage for each SID, and so the SID only got the 5V supply and not the 9V (or 12V) supply, thus the internal audio amplifier wasn't working. *** if you are feeling really brave, you can try to temporarily touch a wire between pin 28 of a working SID and pin 28 of the not-working SID, but be careful not to touch any other pins while you do this!!!! *** Do that with testtone app already running and all your audio ready to hear a difference. You are probably using 6582 SIDs ;) so the chance of frying the other SID is low, as 9V inserted into other pins are unlikely to do any damage.

Good to hear! Have fun!

Congratulations! Does Osc waveform Saw LED work? Test continuity between Osc 1 LED (left pin) and Osc waveform Saw LED (top pin). Refer to the PDF of the PCB layout to see what I mean... note top layer track joining these LEDs. This is the only track leading to the faulty LED which could be broken, since you do not report Osc 2 or Osc 3 being faulty (which share same track on cathode) or that the Mod Matrix Target buttons (O1 pitch, O2 pitch, etc) are faulty. You could try replacing the transistor in the audio buffer. It might also be something strange about the passive mixer. Check R70-R77 (the resistors below the audio sockets). They should all be 10K.

Do you check the LED is good with external power by touching the LED leads on the top? Or touching the pads on the bottom? Try also, supplying power to the pads at the bottom of the board (with resistor of course and no power to the MB-6582). i.e. try lighting up other LEDs by controlling the matrix at the JDx pads. This will test the tracks to the LED. If those LEDs do not light up with power, check if swapping +/- makes it work (test wrong polarity) My guess is the LED might be in the wrong way, you've perhaps tested it works with external PSU but applied +/- differently to how the matrix applies +/-. Alternately, maybe the solder joints are not good, and you're testing by touching the leads and not the pads. I don't think it is breaks in the tracks as that would mean more than a single LED in a row would not work (matrix tracks go straight through each LED pin, it's practically impossible to have a break that conducts around the LED pin... only a bad solder joint could explain that kind of problem.

Can't you work that out from the PT-10 datasheet? It seems to show the rear "hole" size (i.e. where it overhangs the panel) as 8.474" x 1.960" (215.24mm x 49.78mm). The rear panel size (both the blank one supplied with the PT-10 and the FPD files I supply) are 8.580" x 2.070" (217.93mm x 52.58mm). You do the math. Oh wait, I already did it in my head: 1.345mm on vertical edges. 1.4mm on horizontal edges. Add 0.5mm wiggle room. Any exposed silver border is easy fixed with a black Sharpie. Just looking at the case, even this bezel is perhaps not needed... that channel is a little wider than 1.5mm, I reckon a 2mm panel would fit there. You really should just get a PT-10 case first and have a look for yourself.

I've used a 2mm red panel on one MB-6582 I made, the only difference was the switches were 0.5mm lower, so only protruded 1mm instead of 1.5mm using the standard 10mm spacers (i.e. 13mm-10mm-2mm=1mm instead of 13mm-10mm-1.5mm=1.5mm). Switches were still very pressable, and the extra 0.5mm panel thickness was not noticeable when put into the PT-10 case. IMHO routing the back of the panel is a waste of money, since it doesn't address the switch height issue at all, and the way the PT-10 case is designed, an extra 0.5mm actually brings the panel flush with the case edge rather than being 0.5mm lower than the case edge. So I would suggest just working out how to make 9.5mm threaded spacers :) and your problem is solved. I was going to do this with the one I made, but since the switches were functional at 10mm gap, I just left it as-is... I intended to file down 10mm spacers with Dremel and griding disc. You could in theory route 0.5mm blindholes on the back for the regular 10mm spacers to fit into, but again this is expensive - any extra work on the rear side of the panel adds $20 minimum. For the rear panel, routing the edge to make it fit probably wouldn't be that more expensive since it's a cheap panel anyway, and since the PCB is designed to touch the rear panel, you actually need the bezel routed on the front (artwork side) of the rear panel, so the rear side (non-artwork side) of the rear panel is aligned with the channel edge.

I don't know why you're asking for gerber files... I thought it was pretty clear where to buy the MB-6582 PCBs:

To use MB-6582 as a case example, people may have got annoyed waiting for parts kits, panels, bulk orders for knobs, etc. This time I'd rather people get annoyed at me for teasing them with PCB designs they can't have until other parts like panels/cases are ready. In addition, the actual rear panel layout is still a bit undefined, as I haven't been working with Doug on this (and the prototype case design) for a while. I am still not sure whether to use panel mounted sockets vs. an extra I/O PCB and much cheaper PCB-mount sockets. Each has their advantages (e.g. PCB-mount USB is an easy addition, and I now have 500 of the same PCB-mount rocker switches used on MB-6582!) The impression I got from people replying to my "questionnaire" was that most people want a pre-fabricated and cheap panel/case solution, i.e. they'll wait for Doug's panel/case, and also want the uncommon parts that work with this PCB/panel (switches, encoders, knobs). For this MB-SEQ project, soldering the parts to the PCB is not very time consuming at all! The switches and encoders snap-in, meaning you can do them all at one time.

I don't have any more spares, sorry. Put your name on the bulk order page, but I won't order more until I know they work! (I've been too busy with other projects to test them, and my one OPL3 board already had ICs soldered).

The problem is, 90% of people are waiting for Doug's case/panel, and I will not be ordering PCBs until that case is prototyped and ready for ordering. The alternative is running the bulk order for PCBs and parts before cases are ready and then have everyone getting annoying at Doug for holding up their MB-SEQ project. In theory I could run a "bulk order" for 10 PCBs now, but they would be quite expensive - US$50 each maybe. Add in the cost of Schaeffer panels and an order from Mouser for parts and the price might scare you :o If there is enough interest (10+) from "bleeding edge prototypers" wanting just a PCB and willing to buy all the other parts themselves (inc. panel) at non-bulk-order prices, then I'll do a mini PCB bulk order. These would be people willing to pay the ~US$175+ for a panel from Schaeffer/Front Panel Express. Unless you have a CNC router in your workshop, self-made panels are not an option. Judging from TK and bugfight's success, it seems documented well enough to build it without much help from me. People who bought the Soundwell encoders from me or mc4, and Re'an knobs from nebula, and already have an ultracore, are ideally suited to build it.