modularkomplex Posted October 12, 2008 Report Share Posted October 12, 2008 Hi,i've just completed my mb-6582 (with 8 6582), yeah!First of all, many many thanks to TK and Wilba for their great work! But now to the problems i've got with this sweet box:1. Sometimes a bassline on the first core gets crazy, there is a continuous bass, even when the sequence is not running. Depending on the sequence i can hear some notes running and i can change the sounding with eg. the filter-settings. This happens only on the first Core and with both SIDs.When this happens and i switch to the drum-engine, noise can be heard. Switching to the lead- or multi-engine leads to a garbled sound but when playing notes on the keyboard it sounds fine and the garbled sound goes away. Also, when i initialize a new bassline-patch the behavior gets normal for a while just to start again after some moments.Could this be a result of a bad soldering? A software-bug? Something else?2. When running a (bassline- or drum-) sequence on any other core than core 1 the matrix-leds of the row and column corresponding to the sequences light up all the time. i can't see the sequence running like on the first core. What could that be?Mike Quote Link to comment Share on other sites More sharing options...
Smithy Posted October 12, 2008 Report Share Posted October 12, 2008 Thats a really strange one.Id say get up some audio or preferably video to give a bit more insight to whats going on. Quote Link to comment Share on other sites More sharing options...
modularkomplex Posted October 12, 2008 Author Report Share Posted October 12, 2008 Hi,thank you for your response, here are some samples:Core 1: No Sequence running!Core 2: Same Bassline running fineCore 1: Drum-engine running with errorAll samples recorded contain only the left channel but the error is on both SIDs connected to the first core. Can't reproduce the garbled lead-sound at the moment.I can't do a video, sorry. But it is not so spectacular with the other bug, just all leds in the row and column of the sequences running are lighting up continously. Mike Quote Link to comment Share on other sites More sharing options...
lukas412 Posted October 12, 2008 Report Share Posted October 12, 2008 2. When running a (bassline- or drum-) sequence on any other core than core 1 the matrix-leds of the row and column corresponding to the sequences light up all the time. i can't see the sequence running like on the first core. What could that be?Not sure about the first problem, but number 2 is just how it works. Leds only trail on the first core.Luke Quote Link to comment Share on other sites More sharing options...
modularkomplex Posted October 12, 2008 Author Report Share Posted October 12, 2008 Not sure about the first problem, but number 2 is just how it works. Leds only trail on the first core.LukeAh, thank you Luke! I thought that could be an error, too. Mike Quote Link to comment Share on other sites More sharing options...
lukas412 Posted October 12, 2008 Report Share Posted October 12, 2008 No problem, I thought the same thing when I finished mine. Hope you get the other issue sorted.Luke Quote Link to comment Share on other sites More sharing options...
modularkomplex Posted October 12, 2008 Author Report Share Posted October 12, 2008 I've reflown the solder-joints which connection the first PIC with the SIDs but no success. The strange thing with this proble is, that it does not happen all the time. Sometimes i power up the box and it works fine. But i have another issue where i dont know if its normal for the Midibox-sid or not (i think not):A scattering bassline.This happens more or less on all cores. With internal and external clock. Is this normal? I think slight differences in the length of notes might be normal for analog equipement, but so much?Mike Quote Link to comment Share on other sites More sharing options...
Wilba Posted October 13, 2008 Report Share Posted October 13, 2008 I suggest doing the following to determine if the problem is software or hardware.i.e. if the problem persists after each step, then make note of it, if it stops after a step, then report it here.This assumes problem is only present on Core 1, and that other Core/SID channels are working fine.1. Upload the MB-SID application again to Core 1. Make sure you "Use Feedback From Core" when uploading.2. Swap the two 74HC595 from under the two SIDs connected to Core 1 (i.e. far right two) with another two 74HC595 that you assume are working, i.e. the ones connected to Core 2.3. Upload the MB-SID interconnection test: http://www.ucapps.de/mios/mbsid_interconnection_test_v1b.zip and run the tests (see asm file). Check all the outputs are working as expected.4. Upload the change_id application: http://www.ucapps.de/mios/change_id_v1_9d.zip to set another PIC (i.e. the one in Core 2) to device ID 0, and put that PIC in Core 1. It's ok to leave Core 2 without a PIC. Upload the MB-SID application again to Core 1.After step 4, you've basically put the chips from the "working" Core 2 and 2nd SID module into Core 1, as well as validated the wiring (with the test app).The strange thing with this proble is, that it does not happen all the time. Sometimes i power up the box and it works fine. Very occasionally, I power up mine and it does strange things, makes strange noises, like stuck notes etc. and that's with fully tested hardware, so I can only assume the PIC gets some noisy power on startup and goes into a weird mode, as if it's memory has some garbage in it... quickly powering off and on fixes the problem. If yours is having problems most of the time instead of only occasionally, then that's a problem that should be fixable. Quote Link to comment Share on other sites More sharing options...
modularkomplex Posted October 13, 2008 Author Report Share Posted October 13, 2008 Hi,Step 1: done - nothing changedStep 2: done - nothing changedStep 3: done - all fineStep 4: done - SUCCESS! Running fine!Thank you so much Wilba!Mike Quote Link to comment Share on other sites More sharing options...
Wilba Posted October 13, 2008 Report Share Posted October 13, 2008 So you now have a PIC which has problems?Try the change_id app with that PIC, make it device ID 2, and put it in Core 2, upload the MB-SID app to that PIC again.See if it works happily in Core 2. If it doesn't, the PIC is at fault somehow.If it works fine, then it's just one of those weird things that happens... you've swapped PICs so in theory the original PIC 1 should not "just work" in Core 2....Sometimes you tinker with stuff and they just work again without really knowing why. Don't question it further, just thank the MIDIbox Gods and start jamming. Quote Link to comment Share on other sites More sharing options...
modularkomplex Posted October 14, 2008 Author Report Share Posted October 14, 2008 Hi,i've tried it with the PIC again and again. But after uploading the sid-app its continuing with upload-requests. Anyway, i have a spare PIC now which runs fine. Someday i will try to reburn the PIC with the bootloader and investigate if it will work afterwards.But now i have a new problem: I'm getting no output from the second SID on core 2 anymore. I've swapped the SIDs and they are fine, same with the 74HC595. It had worked before, but the last two days i've only checked it mono and didn't notice that.Thank you again for all your help!Mike Quote Link to comment Share on other sites More sharing options...
Wilba Posted October 14, 2008 Report Share Posted October 14, 2008 Run the testtone application on Core 2. This should prove it's not the MB-SID app at fault.Since you say both SIDs are working, it's probably a hardware fault... perhaps bad solder joint somewhere, or maybe even the audio output buffer. Quote Link to comment Share on other sites More sharing options...
stryd_one Posted October 15, 2008 Report Share Posted October 15, 2008 after uploading the sid-app its continuing with upload-requests...........But now i have a new problem2 One thing at a time ;) Quote Link to comment Share on other sites More sharing options...
modularkomplex Posted October 15, 2008 Author Report Share Posted October 15, 2008 2 One thing at a time ;)Hey Stryd,i only do one thing at a time, i've put the faulty PIC aside to check it again later. So the only problem left is the one missing channel. I'm going to investigate (and hopefully fix) this further this evening. Tomorrow is the first public presentation of this little beast.Mike Quote Link to comment Share on other sites More sharing options...
stryd_one Posted October 15, 2008 Report Share Posted October 15, 2008 So, you're 100% sure that there is not a short (like Wilba suggested) which could also have contributed to the failure of the first PIC, and may do the same to the next one?What you're doing is not really the point of "one thing at a time". To quote the page I linked, and to paraphrase your reply:“So what? It’s fixed!†I hear you say...well...maybe it is, maybe you just moved the problem somewhere else, maybe the problem will return... you don’t really know! Quote Link to comment Share on other sites More sharing options...
Wilba Posted October 15, 2008 Report Share Posted October 15, 2008 The 1st problem was with the first Core.He's swapped PICs 1 and 2 and the 1st problem is gone. Hooray.The PIC that used to be in Core 1 has (apparently) got some issues with uploading.Maybe it's the PIC, maybe it's because he's trying to upload in Core 2, which has a problem which might be related to the right SID not working.But obviously to know that the right SID is not working, he must have a working PIC in Core 2.So he might have two NEW problems, the order he wants to fix them is up to him.Stop being a troubleshooting Nazi.Oops, Godwin's Law, you better lock this thread too, you forum moderator Nazi.Oops. Quote Link to comment Share on other sites More sharing options...
stryd_one Posted October 15, 2008 Report Share Posted October 15, 2008 the order he wants to fix them is up to him.It is not! He must do it MY way! EXACTLY!Stop being a nazi nazi :P Quote Link to comment Share on other sites More sharing options...
modularkomplex Posted October 15, 2008 Author Report Share Posted October 15, 2008 :o Aaargh! Stop that! ::)That one PIC had problems from the beginning. No short. I stuffed another PIC in there, uploaded all the stuff and it worked like a charm. I changed the two PICs from core 1 and 2 yesterday, changed the ids, uploaded the app... but nothing changed, even both SIDs on the first core are working.I will check further this evening. Mike Quote Link to comment Share on other sites More sharing options...
modularkomplex Posted October 15, 2008 Author Report Share Posted October 15, 2008 Ah, i found it!Checking the J3 and J23, both SIDs work as expected. It was the audio-jack!The jack had no good connection to the pcb. Perhaps this is because i had first soldered other jacks and realized too late that these did not fit the back-panel. The unsoldering of these jack may have weaken the pcb-traces at this jack. I have bridged that with a short piece of wire from J23 to J22 and it works again.Many thanks to all of you who helped with your hints and ideas!Mike Quote Link to comment Share on other sites More sharing options...
freddy Posted January 5, 2009 Report Share Posted January 5, 2009 Not sure about the first problem, but number 2 is just how it works. Leds only trail on the first core.Can this be "fixed" (i.e. extended to other cores as well)? I can not imagine a reason why it shouldn't behave the same for all cores. Quote Link to comment Share on other sites More sharing options...
TK. Posted January 5, 2009 Report Share Posted January 5, 2009 Can this be "fixed" (i.e. extended to other cores as well)? I can not imagine a reason why it shouldn't behave the same for all cores.No, this cannot be "fixed". Displaying realtime informations from the slaves would produce too much CAN bus traffic.Best Regards, Thorsten. Quote Link to comment Share on other sites More sharing options...
freddy Posted January 5, 2009 Report Share Posted January 5, 2009 No, this cannot be "fixed". Displaying realtime informations from the slaves would produce too much CAN bus traffic.Sure, makes sense. Thanks! Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.