kokoon

sorry i haven't posted - it all works now. i didn't have the bootloader on the pic. when i was burning i didn't know what to burn and i thought MIOS would be okay. it wasnt. i built the broccoli18 and burnt the bootloader and now all is good. thanks for help guys!

yeah i figured it on my own (searched the forum), thanks anyway TK, since you're in here already - any idea about the inital problem? (first+second post)

uh okay i got it, had to convert application ID too :) now i gotta figure why it doesn't work in slave (problem from the beginning)

now i'm trying to set the core back to master (id 0x00) but i can't do it with the change_id application. looks like... i don't know what it looks like because there's no response. i connected the midi cables to the slave (0x01), put the optocoupler back in and recompiled the change_id for an id of 0x00. and converted it into syx. then when i try to upload it with midi-ox, nothing happens. midi in works though cause i can play notes. and some controllers (but not all - see above - that was the inital problem) maybe the change_id application only targets devices with id 0x00? if that's the case - can i fix this in the code? i'm just trying to figure what went wrong and i thought putting this slave back to single/master would show me if there's anything wrong with the hardware. i'm especially worried about the filter cause i read the filter just stopped working for someone. my case is similar - filter doesn't work via CC. anyone care to help me out or at least suggest something? please if possible not "re-burn your PIC" cause i put all that junk in the closet and hope to never put it out again for next few months :)

great! i'm really glad to read this. i'll finish the original design first and then go look into the code :)

correction: it responds to some CCs, but filter doesn't work... but filter on slave worked 1hour ago when i was running the same pair (core + sid) as master. also JSynthLib doesn't do anything for the slave. ANYTHING not even main volume. but if i send CC 7 it affects both - the master and the slave.

master does, slave doesn't. merger is enabled - the slave plays notes. what to do?

from what i read - if you're using PIC18x broccoli 18 should work.

i'm running MIOS 1.8 and SID 1.7a. i'll try with the new SID app. but - in short - what happened: i have this M-AUDIO radium49 and it has a terribly slow pot readout refresh rate - if i turn a pot quickly i only get 5 or so values. even only 1 if i try really hard. so i made a patch in pd which interpolates incomming midi controller values so i can use my radium. the patch works fine. i thought until now :)

pure data is a free max/msp cousin for windows and linux. pretty similar to reactor but more low-level. and you don't get that much fancy interface widgets. go check http://www.puredata.org/.

i put flooded in quotes because i don't think it's really a flood. if i twiddle a knob on a midi controller too furiously i get the whole thing to stop responding. last note plays on and nothing changes. have to turn it off and back on. since it's not like 10000000 midi CC messages per second i'm assuming something is wrong. now my question would be - is this supposed behaviour? will it be the same with the cs? thanks guys, i just built a mbsid and i LOVE it. but this really bothers me.

and another advice (now that i have such experience) - DON'T USE JDM. make broccoli18 instead (see thread in testing/troubleshooting). it will save you a day or two.

i can make this in pd (pure data) if anyone really needs it.

YAY!!! got it!!! ;D ;D ;D at the end i had to put both 0.1uF tantal and 0.22uF elko in parallel. and a 220Ohm resistor. and use linux. but DAMN was it hard. frst i tried it on 3 different machines on winXP. then with all the possible IO-allow drivers. nothing. changed the caps, tried the combinations, ... changed the resistors, ... nothing. then i tried it in MS-DOS 6.0. nothing. at this point i was pretty much sure that the pics are broken. and i went to sleep. in the morning (actually in my dreams) i remembered that i never got the measurements right with the portdiag.exe. so the fault had to be in LPT communications. i tried in linux then... a few more combos with caps and ... SUCCESS!!! thanks to you guys i have a mbsid running and working HAHAHA YAY YAY YAY i'm so happy!!!

i'm building my first MBSID and i'll probably build it the normal way (like TK's original) BUT i'm thinking about making another one and desing it like an old analogue synth. that would mean: - NO lcd display - no lcd means no menu - lcd + no menu means separate interfaces for all 3 OSCs and all 7 LFOs. and the 2 ENVs. - all pots, no encoders - rotary selectors (switches) instead of buttons+leds i'm aware the control panel would be much bigger but i think it could still be designed compactly enough. anybody else thought about something similar? the question is wheter MIOS can handle that many AINs. i'd really love to have such a synth and if it's possible i WILL build it. any input/comments appreciated!

what about 0.1uF tantal? i don't exactly understand where should i put 9V and where 3V... or do i only need 9V for erasing and use 3V for programming?

thanks alot. i'll build this today. by the way - it's hard to get a 0.1uF elko here, but someone mentioned trying 0.22 or something. did that work? would that work? about the lcd - tk wrote in his first post that there will be some problems with the lcd because of one of the PIC pins is used that shouldn't be. or something. but disconnecting that pin from the programmer solves that?

and does the lcd work if the pic is burned with this?

slovenija, ljubljana. moj midibox sid je v nastajanju, trenutno mi najvecje tezave dela programiranje PIC-a. lep pozdrav vsem slovanom!

i assume it routes midi notes + all the rest from one midi channel to separate channels for 4xmbsid to achieve 12note polyphony. or maybe not?

so - let me get this straight - i only need: 1 220ohm resistor 1 0.1uF electrolyte 1 9V battery / wallwart adapter 1 socket 1 connector ?!?!? if that's true then i'm wondering why this isn't on the mainpage already... you really don't even need the protoboard for this...

what's the status now on this? what's the hardware anyway? is it this? anyhow, i'm looking for something useful cause JDM i just built just won't work no matter what. i'm trying to program the 18F452.

here you can figure out which has which. http://www.zimmers.net/cbmpics/ yesterday i just got one 6581 for free. i asked on a few forums if anyone is selling a c64 and this guy had a broken one and i just took it. and a psu too. a good one, with easily replaceable fuse. yay!

i'm just having the same issues but i think i just got it this time: 1. get the (blank) PIC 2. look up what you're gonna use it for and determine the needed ID if i got this right then i'll need an all-zeroes ID for a regular lcd master core of the multi-sid, and for each slave core an ID with 1, 2, 3, 4, ... at the end of the zeroes 3. get ahold of a PIC burner of some sort eg. build yourself a JDM burner, go to a friend who already built/bought one 4. plug your PIC into the burner 5. check/correct the PIC ID in the burner app. apparently the values of what you need to change (according to the example i'm following) it's located "in bitfield ID[57-63]" - as written on the ucapps. i guess that means you have to edit hex value on offset 63 but i'm not exactly sure. the screenshot on ucapps shows a text field where you can just enter the ID and that's that. i hope it's that easy. 6. load the bootloader hex into the burner app and burn it. now i'm not exactly sure if all those restrictions for the buffers and delays apply here also but let's say you just burn it the easiest way. 7. plug the PIC with burned bootloader into the core and connect it to your pc via midi. from here on follow instructions on http://www.ucapps.de/mios_bootstrap.html someone that's done this before - PLEASE tell us this is correct. or wrong. and then correct us. the info about this is really lacking.

okay i'll build another core and then if it still won't work i'll build a programmer.