Wilba

I can't understand why the non-conductivity of the heatsinking compound causes so much chatter. I should have just specified "white silicone goo". Actually, I should have just packed a squirt of the stuff in every kit, but I don't know how to pack it. Suggestions?

LOL At least you got to do some troubleshooting... I've not needed to on any board I've soldered myself in a long time... which is slightly boring. I don't even bother with voltage tests before plugging in ICs just hoping for a fried IC to locate and replace. *sigh* I suppose it's a good thing because troubleshooting aka. debugging aka. software development is my day job. The last time I had some head-scratching WTF moments was with the OPL3 module: Comment #7 is lulzworthy though. BTW that OPL3 module is totally redundant now. Free to a good home, inc. free shipping.

Why does nobody put up a sammich "unboxing" video? Likesuchas: :rolleyes:

a) Polarity doesn't matter if you are NOT bypassing the bridge rectifier via headers JBP. However, if you are using a regulated DC power supply that is tip positive, you can optionally bypass the bridge rectifier (like you must for 6581 use) so you get 12V DC going into the 9V regulator. b) 12V is OK. I use either a 12V AC 500mA or a regulated 12V DC. It should work fine. c) Of those two choices, I would not choose the 2nd one because it is switchmode. The 1st one will work and is preferred because it is regulated and appears not to be switchmode. Switchmode power supplies typically have an input voltage range (like 100V-240V) and aren't so chunky. Also it's confirmed "linear" in the questions tab on that page. You might need to use shunts in JBP as mentioned above, but probably won't... 12V - 1.2V (drop of 2 diodes in brige) = 10.8 V, should be enough for 7809. YMMV. Please update the wiki with a link to this one if it works well for you: http://www.midibox.org/dokuwiki/sammichsid

POIDH!

Try swapping the PIC with another one. Scratch it with the knife first to clearly mark it as suspect #1. (Note to other builders: I can suggest this to him cos I know he has lotsa sammiches) The "every nth patch doesn't get uploaded" could be due to too short time between SysEx blocks, try increasing this to some magic number which I can't remember, but mine is set to 1000ms which is overkill.

You need this enabled in the setup asm: ;; if 0: pin RB2 and RB3 not used for CAN bus ;; All MBHP_OPL3 data pins connected to J15 ;; if 1: pin RB2 and RB3 are used for CAN bus ;; MBHP_OPL3 data pins #2 and #3 have to be connected to PIC pin RB1 and RB2 ;; (available at J5:A6 and J5:A7) ;; Note: this option should only be enabled for PIC18F4685 (PIC18F452 doesn't contain a CAN) #define ENABLE_MBNET 1 [/code] sammichFM uses a PIC18F4685, so two pins of port B get reserved by default for CAN bus/MBNET, and two pins from port E get used instead. So maybe this is your problem. Try using setup_pic184685_mbfm_v1.asm as a base and change to "#define ENABLE_MBNET 1"

I blame nILS then... :whistle: sammichFM was entirely his idea, I'm just the monkey that puts parts in boxes.

Unfortunately yes... :huh:

Update the wiki please :)

I confirm that I sold him new 6582A SIDs and sammichSIDs, in case anyone was doubtful.

TK, can you reproduce the issue with your MB-FM by enabling the extra instructions that write bits to port E? It always writes the full byte to port B, so you would still get sound, this would just let you reproduce it and prove the extra three instructions cause the problem (the two blocks wrapped by #if ENABLE_MBNET in MBFM_REG_Write). If I'd known this would be a problem I would have scrapped the future support for MBNET and just stuck with port B connection to OPL3... :blush:

MB-SEQ-CS_combined.zip

Interesting... DIN inputs 14 and 15 aren't used. Well, technically the 14 one is connected to a pad under the encoder in case you use an encoder with switch, which you didn't, so they both are unconnected to switches and thus you should never get inputs from them. So... Check RN4 is correctly oriented. Yes, again. :) You should be able to see the dot on the resistor network from that side. Resolder the joints on RN4, and check there are no shorts between pins. Check RN4 pin 1 (next to dot) is connected to 5V. Resolder the IC pins connected to RN4. Look at the PDF of the control surface PCB and check there is good continuity between the IC pins and RN4, where there are top layer tracks in red. It's possible (but unlikely) that the resistor network in RN4 is faulty, but I would try other things before replacing it. You could double-check the part number, just in case you got the wrong part (maybe it was put into a bag of the right part at the factory? I dunno, just guessing. I only buy and pack one kind of resistor network!) You could try to connect those "faulty" pins of IC2 to a 5V pin, using a 10K (or anything between 1K and 10K) resistor. It's safe to do this... you're just replicating what the resistor network is doing. The pins in question are pin 5 and 6 of IC2, which are connected to pin 4 and 5 of RN4 (counting from the common pin with the dot marker). You can connect to 5V at C2. I'm not suggesting you do this as a permanent fix, but if you do it and the tester doesn't show those "unused" events anymore, it confirms what is at fault, and you can then move to a more permanent fix, like replacing RN4 perhaps.

Is the 5th item (i.e. K#5 or K#V) value blinking while you are changing it with the encoder? Probably not. If it was blinking, then that item would be selected and turning the encoder should change that parameter. However, I suspect that somehow the code is interpreting the encoder as the 5th encoder in a full control surface, with the "Knobs" layer active. i.e. If you have a full control surface with the "Knobs" layer active, and turn the knob labeled "Release"/"Pulse Width"/"Knob #5", then it should jump to the Knobs page and change "K#5". Try this app: http://midibox.org/forums/index.php?app=core&module=attach&section=attach&attach_id=6823 which tests buttons and LEDs and the encoder. Turn the encoder and tell me what you see on the screen. You can try all the buttons too and tell me what you see.

Please add links to good power supplies to the wiki for sammichSID and sammichFM... I can't recommend any for people outside Australia because I can't test them here. I've added placeholders for the info. Just post your link like this: [[ http://etc... | Name of Store, Power Supply voltage/rating/type etc, Price ]] The "|" separates the URL from the text description. Preview the page before saving, don't leave a mess, etc.

Hmm... I don't know how shorts could cause the effects you are seeing. Look at the PCB: http://www.mb6582.org/sammichSID/sammichSID_CS_PCB.pdf The 5 select buttons and 3 function buttons are handled by the 74HC165 on the left. The other 74HC165 handles the other four buttons on the right, and the rotary encoder. All I can suggest right now is check that there are no shorts between the pins connected to switches and the rotary encoder. Maybe try this app: http://midibox.org/forums/index.php?app=core&module=attach&section=attach&attach_id=6823 which tests buttons and LEDs.

Button "F2" should jump to the Knobs page, but it should then show K#1 to K#5. Confirm that "F2" jumps to the Knobs page and the first parameter is "K#2" not "K#1". If that is true, I don't know how it gets scrolled one parameter. Perhaps if F2 switch and an encoder pin are shorted, then maybe that's what is happening... turning encoder jumps to Knobs page, and then handles the scroll of that page.

I meant what shows AFTER that "MIDIbox SID boot" screen, i.e. if it jumps into a menu straight away, or you see the main screen with patch name etc.

No, you're right... I'm paraphrasing TK's MIDIbox FM specifications here: http://www.ucapps.de/midibox_fm.html Maybe it should be written as "4 simultaneous instruments of up to 6 voice polyphony each, plus drum kit. Each voice is a 4 operator waveform." Or you can propose an alternative.

Check you are installing setup_sammich_sid.hex (and not any other .hex file). Check the resistor networks are oriented correctly, the dot on the resistor network must match the dot on the PCB. You could have a short between the pins connected to the rotary encoder and pins connected to switches, but this seems unlikely, as they aren't close together. Tell me what you see on the screen as it boots up (what screen is showing after the "MIOS" screen).

It is a lot less current draw than sammichSID. nILS did some adding up once and I forgot the total. I put an extra heatsink in there to be sure, and kept the ventilation the same as sammichSID, because people will probably try to run it on 12V AC 500mA plugpacks.

I got a parcel today but alas it was only 3000 Re'an P401 knobs. WO IST MEINEN DREHKNÖPFEN!!!!!!!!!1111111111einself

Yeah it will work, and might cool the voltage regulators by an extra 1°C compared to cheap silicone compound.

Wiki updated: