Wilba

Very unlikely you have damaged anything. Even while the PICs had their Tx pins connected (with four shunts in J11), that shouldn't have damaged the PICs. If you had another MIDIbox Core module, you could test your PICs and MIDI In/Out with that, but I assume you don't. So... start with the PC. Connect MIDI Out to MIDI In of your MIDI interface, run MIOS Studio, setup the routing and test that notes you send with the keyboard appear in the MIDI In window. Don't just assume it works, TEST IT! :) Here's the MIDI Troubleshooting page: http://www.ucapps.de/howto_debug_midi.html While doing these tests, work just with the master PIC (ALWAYS leave out the other PICs) i.e. put shunt in J11 position 1, do all tests with the far right IC socket, either with the PIC in or out, depending on the test. The MIDI In/Out circuits/parts are near the sockets. There are two traces going down the right side of the PCB to J11, the Tx track (which splits at J11 to each PIC) and the Rx track (which splits and connects ALL PIC Rx pins). You should be able to follow the MIDI Troubleshooting guide with MB-6582, the parts are labelled the same. I suppose most of the time the problem is related to bad solder joints or a short between adjacent pads. I suggest checking the solder joints for the MIDI sockets, the resistors near it, and J11.

loL

White acrylic paint, some people use other kinds of paint.

No, you cannot use 6581 (12V required) and 6582A/8580R5 (9V required) at the same time in sammichSID (without making modifications to the PCB). There is only one voltage selection header/jumper for both SIDs. Technically you can run the 6581 on 9V but the filter will not work. I decided that nearly everyone who doesn't buy two 6582A from me will be using two of the same kind of SID, since this is a MIDIbox SID designed for a "stereo pair" of SIDs and not two SID engines with one SID each.

:santa: There's been a cancellation in batch #3 and in the spirit of Christmas aka. End Of Year Shopping Festival, one lucky person can be the first to post in this thread and say "IT IS MINE, OH YES, IT IS MINE!" and in doing so, jump the queue and sneak into batch #3, which might ship on the 28th or something like that. This lucky person will also receive the limited edition sammichSID keyring valued at over one dollar! W00T! :frantics: DO NOT PM ME! DO NOT EMAIL ME! The whole point is to avoid even more emails! :santa:

Sorry, they were there, then they weren't... this new forum stuffed up some attachments and even fixing the problem sometimes leads to losses. Damn. Maybe fussylizard can add them back again if he sees this :)

vented black one FTW... I haz 6.

To be perfectly clear... SO should not be connected to SI (on base or CS PCB). Similarly, any pin with a label (like RC, SC, Vd, Vs) should connect to other pins with the same label, and NOT to anything else. SO should be connected to PIC pin 19. SI should be connected to PIC pin 20. RC should be connected to PIC pin 15 21. SC should be connected to PIC pin 16 22. The above tests are hard to do between the CS and the PIC (i.e. with the PCBs attached), but if you validate base PCB J8/J9 carefully and also CS PCB J1, you can safely assume there's nothing wrong with the actual headers "mating" and thus the PIC pins are being connected to CS PCB correctly. Then you could move on to other places, like solder joints of IC sockets on CS PCB. I suggested taking ICs out of CS PCB just to be sure... it is more important for the PIC because there is some conductance inside the PIC between its pins... I didn't know you had continuity mode on a multimeter. Take both PCBs apart and retest each pin of base PCB J8/J9 and CS PCB J1, making sure it is not connected to adjacent pins. You can put some spare header pins into J8/J9 for easy testing. You really should do all the tests and list them here so I know what you have done already, it's hard to help if I don't know exactly what you have definitely tested is NOT a short or break.

So you found shorts between SO and SI on both PCBs after they were disconnected? Did you take out the PIC first? And also, when testing for shorts on the CS PCB, did you remove the ICs? Seems very unlikely that you have shorts on both PCBs between the same two pins/tracks. How are you testing if they are shorted? SO and SI are connected to pins 19 and 20 of the PIC, you can refer to the Core schematic, headers J8/J9 are the same on sammichSID: http://www.ucapps.de/mbhp/mbhp_core_v3.pdf

Also ich bin ein Klugscheisser mit einem riesigen Eisenpenis! :w00t:

owns face lol... Rubber feet: I was going to include them but couldn't find a cheap source. Heatsinks: Try the ones fussylizard suggested... you could cut the fins down a bit with wire cutters, maybe... I haven't tried putting them in sammichSID. Maybe some low-profile ones meant for PC RAM or GFX card chipsets would be cool. I'm OCD about cooling too (note the overkill ventilation and "as big as can be" heatsink) but I don't worry about the SIDs in a sammichSID with all that ventilation. Heatsinks don't really help here... the heat is definitely getting out of the SIDs, and doesn't need to get out any faster IMHO YMMV OMGWTFBBQ

Can you confirm you don't get MIDI Out from any of the PICs in any of the IC sockets? Take out all PICs, put PIC #0 into each socket and set J11 jumper to suit, power on, watch MIDI In window. You should get an upload request. Repeat with the other three PICs. You're trying to find out if the problem is the PIC, the socket it is in, or the MIDI Out connections (i.e. something wrong between J11 and the MIDI Out socket). No upload requests means it's probably either a) PICs don't have bootloader firmware burned or b) something wrong between J11 and MIDI Out... you seem to have ruled out c) PICs have good power. BTW what PICs are you using? Ones from SmashTV with a sticker? Or fresh ones from somewhere else which may not have the bootloader burned yet?

To be sure it's definitely on the CS PCB, remove it from the base PCB and check. If there's a short while it's removed, then it's obviously on the CS PCB. If there isn't a short when it's removed, it could be on the base PCB (between pins of J8/J9 or near the PIC pins they connect to) so check for the short on the base PCB with the PIC removed. If there's no short on either when it's disconnected, then the short is happening by something else touching those pads while the PCBs are attached. Unlikely but possible. On the CS PCB, it could be on the bottom of J1, although that's unlikely as you'd need a lot of solder to fall through the holes and build up under the PCB. SO and SI are only close together at J1 and also at J2 (top right corner) but J2 is an unused header and you wouldn't have soldered there. Confirm where the short is (CS or base) and I'll come up with some more ideas.

Fixed attached images in first post.

I have used 1K before with the red ones (TK's MB-6582) but you could use 220 ohm, it will not be too bright. Definitely use 220 ohm for SmashTV's green and yellow LEDs.

I don't know SmashTV's exact parts, but these are good LEDs for sammichSID etc. "LED Standard Green Diffused 570nm 20mcd" Mouser: 638-204GD "LED Standard Hi-Eff Red Diffused 625nm 12.6mcd" Mouser: 638-204ID Don't be put off by the low mcd rating. These things glow nicely even in a matrix, using standard 220 ohm resistors, they have good diffusion and have the common package like what most 3mm LEDs look like... i.e. just like SmashTV's LEDs.

I often wonder whether it's worth making different revisions of PCBs (and fixing past stuff-ups) or just be consistent. :rolleyes:

You can't measure current at J25... that's just an unregulated output of the rectified 9V AC coming from the C64 PSU. To measure current you need all the power to go through your ammeter or multimeter. In this situation, you probably have a C64 PSU which is failing to deliver enough current on the 5V supply. In order to test actual delivered current/load, you'd need some way to break the 5V supply "track" and make it go through the ammeter/multimeter before going to the rest of the PCB. There's a good place to do this... at J73. Regardless of whether you are using PSU Option A or B, all 5V supply passes through J73. It was a "backdoor" header intended for people to do workarounds etc. so the build guide says: So remove that bridge... if it's like mine, and just a U-shaped resistor lead, then cut it and desolder each half separately, soak up all the solder and put in a 2-pin header. Then a shunt/jumper can be used to connect it like it should be, OR you can take out the shunt and connect your ammeter/multimeter there, so all current will pass out of pin 1, through your ammeter/multimeter and then back into pin 2 and the rest of the PCB. Then you can measure the current, although the load is fairly well estimated anyway and the problem is most likely the PSU unable to deliver 5V at ~1500mA or whatever it really is... I did the math once... but from memory, each SID draws 100mA on the 5V supply, plus you have four PICs and an LCD backlight. The simplest fix MIGHT BE to move the LCD backlight to the 9V supply (or maybe better, the unregulated supply at J11). I've written about this in another thread... I think it was m00dawg or fussylizard's MB-6582 build... in short, you replace the 5V supply to the J15 B+ pin but NOT to the other pins of J15 or the P1 trimpot. Swapping the supply like that might shave off 250mA of load on the 5V supply, which is enough for two SIDs! This idea works... I've tested it in sammichSID with high-power LCDs that draw >250mA with all current being supplied by the input 12V DC and not the 5V on the PCB. Ideally you should replace P1 with 50K so you have more control over the actual current, and ALSO keep a break in that supply to J15 B+ so you can measure the real current delivered to the LCD backlight and trim P1 to tweak the current, buy you could get away with just replacing P1 with 50K and not going past 50% brightness.

J11 is used to set which ONE of the FOUR PICs is connected to the MIDI Out. I thought this was pretty explicit when I wrote: So using four jumpers in J11 is errr... *cough* completely wrong. :whistle:

It's OK to trim the headers on the top of the CS PCB, but you shouldn't need to do this... there is plenty of room on either side to put the soldering iron tip. The fact that you're getting a different LED pattern and buttons don't work suggest that something is wrong with the connection between the PCBs. So perhaps you should check there are no shorts between the pins on the J1 header of the CS PCB. Note that most of the pins are in pairs connected to the same track (RC, SC, Vd, Vs). It's quite possible that shorts between RC, SC, SO and SI might be causing your problems.

There's no "navigate to DIN". The text under DIN should change as you press buttons or turn the encoder, it's showing the state of the DIN modules (the 74HC165 pins). Ignore the AIN and DOUT. The AIN bit is showing you the state of analog inputs which are not connected and thus random. You might have some bad soldering in the J8/J9 header (the header that connects the two PCBs). Check the solder joints on both sides. Check that there are no shorts between the pins.

The slave isn't around :( In the meantime, you can use this app: http://www.ucapps.de/mios/ain64_din128_dout128_v2c.zip It will show you the state of the switches under "DIN". You should see events when you press and release switches... i.e. pressing the first select button will show "5*" when it's down and "5o" when it's released. DIN assignments for sammichSID look like this: (13/14) 5 6 7 8 1 11 9 4 3 2 12 10 [/code]

I will get my slave to write a custom test application which will show the switch states on the LCD. Then you can know if the problem is really hardware or just specific to the MIDIbox SID app. SLAVE! Come here!

Woot! :frantics: Great photos! I like seeing sammichSID in a group shot! :thumbsup:

Still sounds like a corrupted upload... although it's possible it could be hardware related, the fact that you had earlier troubles uploading suggests the firmware isn't 100% good. Please confirm you got zero errors uploading using MIOS Studio with "Use feedback from core" and all the settings from the build guide. I suggest you switch now to using the latest release (RC34) so we all are on the same page. (Let's rule out your email app corrupted the hex file I emailed you). Check that the heatsink isn't touching any of the pins under the control surface PCB. It's best to rule that out now :) Then start searching for hardware problems... some bad soldering near the 74HC165 ICs might be the cause... incorrectly aligned resistor networks, etc.