MidBox SID Core Module schematic

MIDIbox SID
1

Hello all,

 

 

i am finally after 5 years coming to the point of trying to finish off my SID module :slight_smile: Since i got a lot of nanoSIDs and now trying to build a full 10 SID chip device, yeah :slight_smile:

 

I have a question, since, in my humble opinion, the documentation is quite not good, sorry to say so, but I am missing a lot of documentaiton on the schematics, for me the best of coourse would be to have some eagle pcbs files or at least double layed brd and sch files as gif or pdf, to controll all the values.

 

But what i am really missing big time is some real schematic of my mb6582 modules.

 

Here is my first question:

 

in each core module there is a Diode on Pin 35, it gives 5 V thru the pull up? R80 which should be 1k.

 

Where can i possibly find that in any schematic?

Or to put it another way, does anyone have a full schematic of the modular addict mb6582 PCBs?

 

 

Would appreciate any help, otherwise i will maybe continue to post my built process.

cheers,

seb

2

Hi Seb - I too am building the MB6582 pcb from Modular addicts. I would love to see the schematic also, but I did some digging and found this:

http://ucapps.de/midibox_sid/mbsid_v2_communication.pdf

It shows the CAN bus between the four PICs. Unless I am in correct, this is what you are asking about.

Regards,

Justin

3

Exactly, thanks a lot for the quick advice, I already thought about this a little.

 

Also interesting choice i think with the can, but I think those PICs ought to have a CAN bus module integrated.

 

Also I have to flash 4 new PICs for my built, maybe someone from Germany would like to help me out there, i dont yet own a Microchip programming tool, I know i can get one for 30€ but right now i will exclusively use it for the midibox sid built seems like a little overkill to buy one.

 

Feel free to post your newset builds and progresses, maybe I will too :slight_smile:

 

Right now I build up the Core and SID module, i still have to solder all the pins and headers and then poer up the unit.

I choose PSU optionB, for 12V and 9V capability, strangely I do own 2 different PSUs form old C64, one has the required 9Vac and 5Vdc, the other one has only DC outputs, maybe from a newer C64 batch.

Cause i got 1 6582 ( 9V) and the nwer 8580 12 V if i am right, ah also a sidequestion, does anyone have a datasheet of the 8580 SID?? my basic research on the webs did only provide the 658x variants..?

 

The rest of the SIDs I do not own originals, but I got my self a batch of swinSID nanos, I will test them, they are suppose to be Atmel SID emulations, running on a single 5 V supply, direct replacement of the original SIDs, I am very looking forward how and if this all works, and which one sounds better, as I then got 3 different SID chips to test.

 

 

Ah also the front, I call it control PCB, is not populated yet, I want to test the hight of incremental rotaries and the buttons and LEDs first, havent decided yet what i am going to use, simple question here:

Do I have to use rotary knobs with switch ( the ones you can press) on the PCB it seems that you can or must use those, probably you can activate something in the menu later by clikcing it, but somehow i did end up ordering the things without the switch, maybe I did read something about this somewhere, the project is rotting in my place for quite a bit now :slight_smile:

 

Thanks for reading.

 

 

 

 

 

 

4

So far, I have the control surface built. There are some pics here:

I originally built a stereo Version 1 sid, with two SID modules and two cores. It only had an LCD and 4 buttons. Now I am upgrading to V2 with the complete control surface. 

I’m a little further on the case, but don’t have pics yet. I have about half the bottom board soldered, but I am waiting on parts. I have 2 6581’s and just got 2 swinSID Nano’s. I haven’t hooked them up yet. I thought I had an 8550 somewhere, but can’t find it.

You don’t need encoders with push buttons. I know there is a mod for the MBSeq where when you push the endocer, it speeds up how fast the values change, but it is not required. My encoders do not have buttons.

I have a PicKit2, so I could program your PICs for you, but you would have to mail them to the US :frowning:

Good luck!

5

Haha had to laugh with your 8580, because the exact same thing i had in my case, gladly I found it in another old C64 after while, but I also thought id lost it :slight_smile:

 

I think i am going to buy a pickit or some china ripoff and then flash it myself, seems faster and cheaper overall :slight_smile:

 

btw did you see, synthtopia.com there is a little article about the midibox sid.

 

So my base PCB is almost finished, from a solderstandpoint, right now I only got a pre flashed PIC, old one from smash tv back in the days but its the flahs intensive one P18F462 or something like that, it aint got no CAN on it, but maybe I hope I still can test the first 2 SIDs in my setup with it.

 

I did sample the other PICs from microchip, but it might take some weeks before they arive.

 

 

Last question for now, as I come to the ultimate seciton of my base PCB, thepower supplies.

 

In my option B I have to solder 3 LDOs, the 7809 twice and one 12V device.

 

Why are there 2x +9V LDOs in it, I couldnt find out:

V2 is +9V and it will supply all the SIDs, but what is V1 for?

6

Have you gone through this thread: ?

I think I am going to go that route. I like using the old C64 brick, but who knows how long it will last.

 I got a bunch of parts yesterday, and have everything done on my base board except the power section and the current limiting resistors for the LEDs.

7

Quick update on my behalf.

I dont know where i got the rotary encoders from, but it seems i can detent them by opening, so i will do so for 14 parts besides the main menu encoder, I hope this i sright, and nothing else will be used as an menu encoder. btw i think i got the one from the netherland voti.nl.

 

What stops me a bit is the missing frontpanel, maybe i will lasercut it in my local fablab, but i do not know if thta can happen in the next days, so i am a bit stuck with the led matrix, i would like to solder them when i know the front panel measures.

 

Today i orderedthe Pics and some more Amegas88 for the swinSID, at first i got some Atmega88V turns out the V variant is for probably less than 32MHz, the PA is also only 20MHz but will be driven with a 32MHz clock anyway, I hope this all works out, it is not my designs, someone else did those, so I hope it will work :slight_smile:

 

For now I am probably trying to lasercut the fronts tomorrow, and will get the PICs flashed, when they hopefully arrive tomorrow morning.

If that does not work out :slight_smile: whihc it probably does not, recalling murphy, i got one PIC18F452, the one without CAN onboard, and will try try to work 2 SIDs from that one, just to get a proof that the whole thing works.

Really looking forward to producing the first sounds..

 

Ah and btw I did recycle the original PSU socket and power switch from the original PCB by destroying it today :slight_smile:

Maybe i will try fitting in the midiBOX into the original case C64 as well, but I will decide on that later, making sound is more important right now, btw I saw your milled pictures, really looks beautiful.

cheers

8

just a quick updateon behalf of the lcd connection:

 

I can only detect D7 to D4 connected on all the 4 PICS to J15, the LCD port, is that true or some defect? I guess right now maybe the LCD is working in a different mode and those connections are not needed, did not power up yet to see if it is working.

 

Just noticed and thought hm this seems strange, will maybe come back to it when it doesnt work.

On a sidenote on my original mike smashtv PCBs those pins were all correctly connected according to the uCapps website documents.

 

we will see how it works out. btw all the PICs did arrive , but i do not know if i can flahs them otday i think rather not, so i will test the thing with my old PIC and only 2 SIDs

 

 

PS: Edit#1

Just saw in the wilba design files, somehow it loads very unstable on my PC, ist a huge pdf or my pdf reader is broken,

the baord file, and there it can be seen where those ocnnections go, so no probrlem after all everythign is fine

9

Just beeing able to flashing my pics..

 

i dont know exactl yif it worked, but the pickit kit says so, i cannot verify this, but anyway i gotta trust something.

 

Now i did want to update the mios on the pics in my PCB setup, but sadly the MIOS is not recognizing my midi PICs yes.

Does someone know what the hell J11_CORE is for? Is it just a bridge from MIDI(UART) TX to RX? I had that jumper set all the time, to it was kind of a short, i hope i cannot damage the output pins with that.

 

I can test a feedback loop just liek stated in the troubleshooting, but my PIC MIOS is never recognized yet.

Just spent 2 hours of my night to get it working, also tried my older PIC18F4620 setup, doesnt seem to recognize it neither, but i know that that one worked like 4 years ago :slight_smile:

 

THen something odd happend, i was pulling and stuffing the J11_CORE in and out while querying to see if i can can loops, and indeed it happend, and surprise in the beginngin one time he also stated MIOS recognized, but of course i cannot work with that, as it vanished when i pulled the J22 again or stuffed it back in.

 

So the journy goes on.

BTW brilliant written troubelshoot guide, really something to read acutually helpfull and sensible stuff, great work.

 

Right now i am suspicious of my noname midi interface, i got that from amazon a while ago for 7€ or so, but it did work for the last 'PIC and MIOS so im am not that sure, well maybe it works with notes but not with sysex, do not know that yet, i cannot recall if i bought a mios PIC back in the time or just a preburned one, but i think  i got it from the german site so probably only bootloader installed, which means mios upload did work once before, well maybe it is dependent on my win10 now, who knows..

 

Second thing, as i am not yet that familiar with PICs and i got myself of course the cheap diamex substitue instead of the pickit flasher, I am a bit suspicious about it. I am using the official microchip software for the pickit2 kit, and it states that my code is on the device, but i cannot see that , cannot verify it,the button simply says its not possible, i guess somehow the pic is not readable, at least that would make a bit of sense.

So if the upload of the bootloader did not work, then of coursethe MIOS wouldnt work.

 

Which leads me to something new..

Is the pic supposed to be MIOSable if only the bootloader is flahsed the first time?? I would assume maybe not, hmmmmmm.

 

anyway it is kind of late, i am passed my sleeping window for 2 hours, so will inform you as soona s i squeeze the first bit of sound out of it.

 

cheers

10

Hi Seb,

J11 controls which PIC is connected to the MIDI In/Out. That way if you have multiple PICs stuffed, you can upload MIOS and the SID firmware to each. I think only one jumper should be set at a time. If you only have one PIC, then it should bridge the two pins labelled 1.

Looking from the top of the board with the midi jacks to the rear, the port on the left is the Midi Out and the port on the right with the optocoupler is the MIDI in.

If you are in MIOS studio and you power on the pic with the bootstrap on it, you should see a sysex message pop up in the left window. Also it will tell you in the info window that the bootstrap is up and running. Then you load MIOS and hit start. 

You can reload MIOS or other hex like the MBSid as often as you want. You only need to program the bootstrap once.

One tip - if you are using the MB6582 control surface, look for the HEX that is for that version. I had the plain MBSid loaded at first, and it wasn’t reading the knobs/buttons correctly.

Here’s where I am. I just got some sound out of it this weekend. This is just one of TK’s Bass patches. The sequencer is really cool. These are two SwinSID nanos.

 

 

11

Sorry, I wasn’t quite correct:

Quote

What is the function of J11 (directly below U1_Core1)?

J11 (not J11_CORE1) is a 4×2 pin header which controls which Core (PIC Tx pin) is connected to the MIDI Out port. You would only need to use this once for the first upload of MIOS and the MB-SID V2 firmware, thereafter connect the master PIC (1) to the MIDI out and after uploading new MB-SID V2 firmware, the master PIC can clone itself to the other PICs across the CAN bus. NOTE: All Cores (PIC Rx pin) are connected to the MIDI In port. The different device ID (0,1,2,3) on each PIC determines which PIC receives an upload, J11 is used to get “feedback” from that PIC during the upload.

 

12

hi great news with your deivce, congrats, looks and sounds pretty convincing.

 

And thanks for all those information, I kind of already have read over that, and I did not mean the J11 jumpers but the J11 Core jumpter, i dont really get those as one can short the tx rx from the PICS.

 

Anyway, did oyu mean this hex? setup_mb6582.hex ?

I will try now to flash the MIOS again and then upfinally recognize it with my MIOS on the mb6582 board, will also probe around witht he scope if it wont work, but i am pretty sure MIDI is working just fine until the PICs.

Since my hardwareloop is recognized and I get the sysex message out:

[300922.170] f0 00 00 7e 40 00 01 f7 every second or so

 

edit1:

short update, I did get another badge of fresh PICs yesterday via the nice microchip sampling service, cause I thought, maybe i messed up with some configurator bytes? is that even possible with the pickit programmer and the diamex???

I am only very very familiar with code composer and the  Texas Instruments controllers, Atmel I had before because of the great work of Oliver at mutable instruments.

What puzzles me a bit is the complicated way Atmel works with its fuse and other settings, with the Texas controllers you just flash a hex file and everything else you dont have to care, or someone else already took care of in the code, Atmel you have a working hex files but the code wont work because someone forgot to tell you about those fuse settings… seems kind of unnecessary ot me.

 

So because of that expirience I wondered or still wonder, is it the same with the PIC, can I possibly mess up a perfect hex file with wrong config bit settings, I read :

https://www.eevblog.com/forum/microcontrollers/pic18f4685-config-bits/

 

and there it at least says so.

 

 

So enough of my rant back to business:

 

I did flash a couple of those PICs I got, btw i also got some PIC18LF devices, are those the same, what does the L stand for, at least i could flash them also.

Now after yesterdays struggle I think something might have changed in my connections ( I am programming on a prototyping solderless board right now, so open connections and too much parasitic capacitance could actually be a thing), because now the hex code seemed to work fine and I can now also verify the program ( I assume the pic program checks if the actual code on the PIC matches the imported hex file from my PC )

 

Then I connected the PIC back into the core module, startet Mios.exe and a couple of times the same as yetserday no MIOS detected and so on.

I also flashed the troubleshoot PIC code on one device, strangly no echo occured, BUT the one pin on J14 I think toggled when I did send a random code over the MIOS, so flashing probably works I thought.

Well after some more fiddling, the first time the upload of the mios hex file did work, it stated MIOS8 detected, that was on device id 0, then i switched on dev 1, it did not detect anything, then back to dev 0, again no mios detected, hmmm.

But anyway I assume MIOS studio got some issues for itself, and i cannot blame it its hard to program communication stuff on pc and midi then having to rely on some external cheap chinese midi gear like i got and so on.

anyway after restart it detected MIOS8 again and so on, then i had to leave home for work, but the hopes are high to produce a sound until the end of the week.

 

And just a headsup, i inserted a pic wrong way around one time ( with 10 pics and on and off troubleshooting and all quick and on a messy workbench, this is 100% sure to happen to anyone :slight_smile: ) just wanted to say i “think” it is still working, at least i could reflash and verify it.

 

thats all for now.

 

13

Ok, if it says MIOS8 detected, do you have an LCD connected? It should say MIOS, and then Ready. Yes, that is the hex file for wilba’s front panel. If it says ready, then you can send the MBSid hex file. You can also send the test tone hex file. For that one, you take out the SID and jumper between pins 8 and 28, and the pic sends a 1khz signal to your audio amp circuit. I found it helpful for troubleshooting. Here’s what the LCD should look like if MIOS is up and running.

20180128075242.jpg.963be0f2386be25bf3ed87aa4dbf623e.jpg
20180128075242.jpg.963be0f2386be25bf3ed87aa4dbf623e.jpg1932×2576 823 KB

14

Also, I think the config values are in the hex file. I just programmed the pic with the hex without changing the config at all.

15

UPDATE #1

< base_url >/topic/18680-solved-debugging-my-mb-6582-part-2-software-upload-issue/?do=embed&embedComment=163383&embedDo=findComment#comment-163383

I followed that one and got it working i think, so cloning is NOT working properly in the wilbas mn6582 I would assume right now, but its easy to fix, just placing the Jumper on J11 on the device you wanna flash via MIOS and then send the mb6582.hex to the PICs one by one, beforehand i switched the device IC, cause this part is not altered by the flash process.

Sweet sweet bootloader and MIOS flash tool, I love it, works like a charm, who needs cloning with that anyway :slight_smile:

 

Ok.

got the whole thing working, did not really test the sid sounds yet, and only the first two sids have been tested for the audio amplifiers, but i am pretty confident it will wokr out, or at least i know where to fix the others since i got some of them working :slight_smile:

 

I thought i could clone whatever program there is on the first sid, stupid me, i assume this only works with the actual midibox sid program, not with the audio test program.

Well i stuffed in the other 3 PICs, had to search for the menu button, shockingly thought to mayself maybe i should have installed a switch endocer as menu encoder and not the one without switch, but turns out there actually is a menu button right under the decoderto the right.

So then I did clone the mb6582 hex file to all the other SIDs, in the display some strange stuff like 0D!!! was presented, maybe that indicates that something is wrong???

I will find out later, then i just randomly clicked stuff in to test a bit around, at startup i can hear some sounds, after that I do not know yet how to produce sounds or maybe even program a sequence in it, but i will find that out later.

 

Also a minor other thing, when i connect the LCD to the other PICs J15, just the PIC device ID 1, 2, 3 is shown in the LCD and no background is lot, as those pins are not connected according to the board file..

16

Cool - making progress. I can’t help with the extra PICs yet. I only have one stuffed so far. To get some patches loaded, open up MIOS Studio, then go to the librarian. If you go to:

The first post has some of TK’s bassline patches. You can use the librarian to send a patch to the core. If you have the play button hooked up (above the LCD window), it will start the patch playing. Do you have the memory chips stuffed? If you do, you should have gotten the formatting messages for each chip the first time you turned it on after putting in the chip.

 

17

so far so good, yeah sounds playing, can püaly sound over MIOS and the respective MIDI channels.

So far so good, can only say as I got one 8580 9V plugged and one 6581 12V, they both seem pretty similar, so far cannot hear any difference when unplugging or plugginh the filter caps, maybe they will be important in some other modes of playing.

another fun thing, the 12V 6581 I did let run from 9V out of fear that it could be the wrong, it sounds a little bit more weak than with 12V, but works just fine, of course it is probably not recommended to run it with 9V, that was just for testing.

 

And no surprise, the swinSID nanos, just soudn way weaker than the original sids, so no chance to mix them otgether, well over seperated inputs to the main mix it will work out just fine, but right now for testing i got all over the passive on board resistor mixer ocnnected ( well done btw very nice feature for testing purposes) it is really hard to find the right cables these days :slight_smile:

 

So mybe i buy another set of PCBs and hunt down 6 other SIDs during my lifetime :slight_smile:

18

Great - I have two SwinSIDS and two 6581’s. I haven’t had the chance to really A-B test the difference yet, and I don’t have the most discriminating ears anyways. The first thing that jumps out is of course the noise level. In comparison, the SwinSIDs are very quiet. The 6581’s have a ton of background noise, especially if the patch has a wavetable going. I have some audio level issues also. I made a mistake at first and put a .1uf CAP where the 100nf goes. It is cramped in there, but I got the cap out and replaced it. For whatever reason, my right SID is over twice as loud as my left(both SwinSIDs at the moment). I need to take the time to debug the audio amp and make sure I didn’t make any mistakes.

For the feedback pots, I had the idea to install 4 single and 4 dual-gang pots in pairs, with a DPDT switch. Then I could have the option of controlling the feedback for a stereo pair of SIDs, or flip the switch and have a separate pot for each audio output.

19

I’ve been doing some digging, and Altitude added an 8 channel internal mixer to his:

20

yeah i read that one for a bit, though a internal passive mixer is already on board, do you know that?

I use that output for testing, but do not know how i plan for future use, since my 16channel mixer well already is kind of used with a lot of stuff, so additional 8 channels from the SIDs does not make sense to my setup.

 

And a quick maybe solution to your swinSID sound problem. I do leave both jumpers off the SwinSID boards, one should emulate for 8580 sound when plugged, the other one will put an additional 1k to ground into your audio out signal path, you dont need that one, since there is already a 1k on the wilba PCB, if you connect the swinSID jumper in, you will halfe that 1k, since they are paralell.

Effectivly halfing your audio signal too, so there might be the problem, of course easily repairable jsut unplug the jumper.

If it is not that problem, well you have to do some troubelshooting :slight_smile:

 

And yeah i also get some VCA bleeding and backgroundnoise on the 6581, the 8085 is much quieter, and swinSIDs are simply not as high of output level, i thought about putting a small opamp buffer int here with gain of 2, maybe i will do that and post the brd files, but not in the near future :slight_smile: