Only 1 of 4 SIDs have audio out Solved

[Guest Thomas_Holley]

Yet another delema for the novice builder. I have audio out from only SID 1 in my 4 SID box.

I rechecked my connections from CORE to SID and they appear fine.

I checked MIDI connection between CORE modules. Out from master to in of slaves with common ground.

I loaded MIOS in all CORE modules again and reloaded SID application 1.6. I only loaded SID 1.6 once, to the master CORE module. Is this correct?

Display indicates modules are selected OK but audio out is zero.

Thomas

[arumblack]

you need the sid program on each core. also be sure each core has a unique device id.

[Guest Thomas_Holley]

Thanks for the reply. Can I program the SID 1.6 into each core by putting the address of the CORE, i.e. 00 for master, 01 for slave 1 02 for slave 2 etc., or do I just load unmodified SID to core?

Thomas

[TK.]

It's more comfortable to modify the SID device SID in main.asm, to build and new .hex file and to convert it to a new .syx file (don't forget to specify the MIOS device ID with "hex2syx.pl main.hex -device_id 0x<id>" in order to address the right core) -- so long as the MIOS device ID is equal to the SID device ID, this shouldn't be too confusing

Otherwise you would have to change the SID device ID via SysEx, and this would require that only one core module is connected to the MIDI Out of your PC.

Best Regards, Thorsten.

P.S.: seems that it makes sense to implement some kind of automatic device ID configuration into the SID application...

[Guest Thomas_Holley]

I programmed the SID v1.6 with unique id to each PIC. if I change parameters with that SID selected I get MIDI message via MIDIOX so I think it is working.

Connected to audio out and nothing except SID 1 has output. I rechecked wiring again and it looks OK. I am missing something but I can't figure out what yet.

Thomas

[TK.]

Is the Link enabled? The Link button has to be pressed, otherwise the MIDI events won't be forwarded to the slaves.

(see also http://www.ucapps.de/midibox_sid_csB.html, search for "link")

Best Regards, Thorsten.

[Guest Thomas_Holley]

Thanks for your quick reply. The few problems that remain in my SID are frustrating but you have been very helpful.

I have tried with Link enabled, disabled, all SIDs selected, one at a time selected, nothing seems to make the output work. I cheched voltages, continuity and ground and all check fine. I will sleep on it and try to start from scratch in the morning. I think I will start by reprogramming all the PICs from Bootstrap loader up.

Thanks

Thomas

[TK.]

Have you ever tried to swap the PICs in order to ensure that the firmware has been uploaded correctly? Such tricks can save a lot of time ;-)

Best Regards, Thorsten.

[Guest Thomas_Holley]

Here is an update on my progress. I reprogrammed each PIC from bootstrap loader, mios v1.6 and SID v1.6 making sure nothing else was connected while reprogramming. They all work now. I can get audio out os each of the four audio outs.

The SID functions don't work in any but the first one. I plugged each SID into the module and each SID chip works. Furftermore, each PIC can control a SID so the PIC/SID modules work. The problem, I believe is related to my Matrix not working properly as stated in another thread.

[Guest Thomas_Holley]

I fixed the matrix but still no SID functions are evident in audio out of any but first SID. The display indicates data is sending, lights change with appropriate settings. The only thing is audio out does not change on the boards.

I'm stuck. Where do I look for this problem?

Thanks,

Thomas

[TK.]

Check the wiring to the slaves, ensure that the slave boards are not stuffed with an optocoupler, press the link button.

If this doesn't help, please describe exactly how your hardware looks like

Best Regards, Thorsten.

[Guest Thomas_Holley]

OK, I rechecked my wiring again and it appears OK.

I have 4 CORE and 4 SID modules. J10 of each CORE goes to J2 of the respective SID.

The master CORE J11 has all four pins of the connector (ground M1 M0 +5v) hooked up to a LCD input with no MAX323 stuffed. Master J11 M0 and ground go to M1 and ground on each slave CORE J11.

All CORE J3 are jumpered.

Power to J1 is the only other common CORE hookup.

SID modules have power to J1, audio in to J4 and audio out to J3.

Master CORE has J15 to display, J4 to Bank Stick input, J8 to DOUT, J9 to DIN, MIDI in to J13, MIDI out to J12, 6N138 optocoupler stuffed.

All lights, buttons and rotary controllers seem to work as designed. When I activate Link and plug an amp into audio out 2, 3 or 4 and press a key on my keyboard controller I get audio out of that note. When I activate any SID function (OSC, LFO, FIlter, etc.) the note does not change. The display reflects what action I take pressing buttons or twisting knobs correctly but sound has no corresponding change.

I get all functions with SID 1. I can plug the jumper from any SID J2 into the master J10 and it works fine so SID modules are OK.

When I reprogrammed the PICs earlier today I placed each in the Master CORE module to program and tried it with SID 1 before placing it back in it's respective module. Each was programmed with ID corresponding to it's position (00 master, 01, 02, 03 Slaves).

I am not sure how to proceed from this point to determine my error. I have not pulled the circuit boards for the slave CORE modules yet.

Thanks for the help.

Thomas

[TK.]

Hi,

you could connect J11:MI of the master to J11:MI of the slaves so that all incoming MIDI messages are directly sent to the slaves. This is just for testing. If it still doesn't work, you have to check each slave seperately like it has been done for the master...

Best Regards, Thorsten.

[Guest Thomas_Holley]

I hooked J11 MI Master to the J11 MI slaves and get no audio out at all.

I hooked up each Slave as the master by swaping cables and each worked fine.

It seems as though the MIDI instructions are getting lost between modules. I measured voltages on all J11 with a multimeter and they all match at just under 5VDC so the wiring seems to be able to pass signal. Since the modules work when hooked up as Master it seems the Module is OK.

Do I look at another reprogramming? When I did it yesterday I made sure each bootstral loader had a unique ID and I reread each PIC after programming to double check. I did the conversion with perl for the mios and SID application to give them the ID. Could I have gotten the conversion wrong and that causes the module to not function correctly?

Thanks again for your help on this. It is really helpful to have something like this forum when doing DIY projects.

Thomas

[TK.]

Hi Thomas,

since all cores are working standalone, the application has been uploaded correctly - a slave will run regardless of the device ID, this ID is only required to address the slave when changing parameters with the control surface or JSynthLib, so no need to reprogram the chips.

Maybe I should repeat my question: are the optocouplers disconnected? The 6N138 should only be stuffed on the master module, the MIDI Ins of the slave modules have to be driven by the master.

It is really helpful to have something like this forum when doing DIY projects

I've the hope that you could write a short report about your experiences (pitfalls etc.) with the project for the MIDIbox portal in order to help other DIY people who are maybe also confused about all the informations on my website (which are sometimes not very detailed, sometimes incomplete, sometimes outdated :-/)

Best Regards, Thorsten.

[Guest Thomas_Holley]

I plugged in the optocouplers to test modules as master. They were out before and are out now. I will measure the J11 wiring and the boards to ensure their propper hook up.

Thanks,

Thomas

[TK.]

You could use the LEDs of your LTC module to ensure that the MIDI signals are going into the slave: just connect on of the inputs of the LTC module directly with the Rx input of the slave PIC (pin #26)

Best Regards, Thorsten.

[Guest Thomas_Holley]

I discovered one difference between the Master and Slave Core modules that may be the cause of my problem. PIC pin 28 (RD5) goes to J11 MD (top right). On the Master CORE this supplies 5VDC. All of the slave modules have 0VDC here. The only other place this PIC pin goes is to J15 which is not connected in any CORE slave. I doubt this makes a difference since I found the corresponding J2 pin on the SID is dead ended.

I will pull the CORE boards and look for faults even though it seems unlikely that 3 of 4 have exactly the same fault.

Thanks,

Thomas

[Guest Thomas_Holley]

I get about 5VDC constantly at PIC pin 26 whether MIDI is sending or not. I unplug MIDI and still get 5VDC. This could be the root of my problem. I will see if I can find where it orriginates.

I spoke too quickly. Of course it is supplied directly.

I took a look at the MIDI signal with an Oscope. Input is about a 1V pulse to the optocoupler. Output from the optocoupler is extremely small, maybe a milivolt and is burried in noise from a dirty signal.

I changed the optocoupler but no change in the signal. Some signal must be getting through because SID 1 works. The signal at J11 MO is very small if it is there at all.

Thanks again,

Thomas

[Guest Thomas_Holley]

I used the light on the LTC module as you suggested and it lights with incomming MIDI signal on all modules so the signal is getting to the CORE.

I am building this for my son who is a synth afficionado. I know virtually nothing about them and have never worked with MIDI before. I must be doing something wrong with the MIDI controller (keyboard) sending incorrect commands.

I swapped CORE cards and put another card in the master position but the results are the same. I get audio out of all modules but the SID only does it's tricks with the master.

I believe now that the wiring is OK. Maybe it is my ignorance of MIDI function that is causing the problem and it may not be a hardware/software problem.

Can you give me any advice on this theory?

Thanks,

Thomas

[TK.]

Oh - you never wrote that you own a scope. This makes debugging much more easier! :)

The Rx and Tx pins are high (5V) when no MIDI message is received/sent.

At the master Rx pin you should see proper digital pulses when a MIDI message is received. Suggested time base: ca. 250 uS/div - trigger on falling edges!

Once the Link is enabled, you should see the same pulses ca. 300 uS to 1 mS later at the Tx pin. This is the signal which goes directly to the slaves. Means: this signal should be visible at the Rx pin of each slave. If not, you've definitely a short between the Master Tx pin and the slave Rx pins. Remove the connections to the slaves, connect them again step by step and check the Tx signal of the master between each step

Best Regards, Thorsten.

[Guest Thomas_Holley]

I can see the signal now at each J11 and at pin 26 of the CORE. So, my MIDI instructions are getting there. The signal out of the optocoupler is spikey though. I changed it again but signal looks the same. I must be giving it incorrect instructions.

Here is what I do, I just turn on my keyboard, hook up the MIDI out to MIDI in on the SID press Link and all 4 SID's and hit a key. The first module follows SID input changes but all the others just stay at the note I hit. Should I be doing something different?

Thanks,

Thomas

[TK.]

Which MIDI channel is assigned to the slaves?

Note that the master directly controls the MIDI channel - means: if your slave is running standalone, the channel could be different

The channel is displayed in the main view, but also in the CFG menu

Best Regards, Thorsten.

[Guest Thomas_Holley]

All are channel 1. Should they be different? My keyboard is set to cahnnel 1 also.

I tried changing channels but got no output. If I unplug the J11 from a slave the audio out stops so some signal is getting to the slave to tell it to play the note I hit on me keyboard. The instruction to add SID functions does not appear to be sent to the slaves because they never change with SID changes. Maybe my problem lies in the area of this signal path.

Thanks,

Thomas

[TK.]

Moment - stucking notes? This remembers me on a bug in an older MIOS version... are you using the most recent release? MIOS V1.6

(Don't mix it with MBSID V1.6...)

Best Regards, Thorsten.