Sammich Issues

[braintentacle]

Hi Everyone,

First, I want to thank Wilba again for this amazing project. This is my first post just wanted to say hi and ask for some help with my sammich. I soldered everything and loaded the midibox application, but the sids only play very muffled sounds. All the voices respond, they are quiet, the pitches are all the same and the noise floor is quite high. Also, the encoder does not respond and the shift button introduces more noise when pressed. The up and down buttons do not seem to work as well. The voltages all checked out except for the 12V rail which was only 11V. Maybe this is the issue? Do any of you midi masters have any suggestions?

[Wilba]

What kind of SIDs are you using?

Is the shunt in JP correct for your SID type?

Do you have the filter capacitors (C1L, C1R, C2L, C2R) installed in the machine pin strips? Are they correct for the SID type? Are they in the right holes?

[braintentacle]

I have 6582A's in mine and the shunts are correctly aligned. I have 4 22nF caps in pin 1,3 and 4,6. I am getting 8.85 volts on my sids. Maybe they are in the wrong holes?

Thanks.

Edited June 15, 2010 by braintentacle

[Wilba]

Take off the control surface PCB. You say pressing buttons has some audible effect, this shouldn't happen.

There are two test apps you can use.

http://www.ucapps.de/mios/mbsid_interconnection_test_v2.zip

Testprogram for the interconnections between Core and SID module of MIDIbox SID (read the README.txt file for the usage)

http://www.ucapps.de/mios/mbsid_testtone_v3c.zip

Just plays a 1kHz triangle wave on the SID - nice for testing the audio output without the need for a MIDI sequencer or keyboard. With the new v2 version, both MBHP_SID modules connected to a core are initialised.

In addition, a 1kHz square wave will be generated on the CS pin (#8) - see xxx for additional informations.

[braintentacle]

I cant really tell if the test tone is working because there is a high pitched, crackling and low cycling sounds are constantly playing, but it is 1khz when I run the test. I did the interconnection test. I am guessing that you measure from GND on the sid to the different pins? CS and A0 are always 5v no matter what setting the test is set to and every other pin tested is 0v ,but on the second sid CS is 0v. I checked the voltage on the sid from GND to VDD and it is 8.91V now. Not really sure whats going on. Also, when I press the shift it seems everything get 2-3 louder not noisier.

[Wilba]

I think you may have multiple issues.

You will need to sort out either the control surface issues, OR the SID issues. Both might be related to the same problem.

I would probably say the SID issues are easier to resolve, because there's less things to check, only one PCB where there might be a problem.

You can try the testtone app with the SIDs removed. This helps identify if there's a problem with the audio buffers (i.e. the noise comes from them). Take out the SIDs and put a 1K (or similar) resistor between socket pin #8 and #27. Make sure you connect it in the right pins! Pin #27 will NOT have 9V! Connecting 9V to pin #8 is BAD

You should get 1kHz out of the audio outputs, this will probably be at max volume level too. If you don't, there's a break between pin #8 and the PIC, OR the audio buffer is the problem. You can narrow this down by connecting pin #8 from the left SID socket to pin #27 of the right SID socket (and vice versa).

BTW are you sure your audio is working? You are using a stereo plug?

The interconnection test is supposed to validate that the PIC->SID connections are correct, as well as connections to the two 74HC595 ICs below the SIDs. If you're not getting the right voltages when you send the Mod wheel message to change which pin is high (for the pins connected to the 74HC595), then there's some issue with the three serial data lines between the PIC and the 74HC595.

Check for bad solder joints, shorts, breaks, etc. between the PIC and the SID sockets, 74HC595 sockets.

AFAIK a high noise floor might result from either the audio input not being grounded (some issue with the audio input socket, or tracks between there and the SID pins)... OR if setting the volume registers in the SIDs isn't working, because the data lines aren't working. So until you get the interconnection test working properly, I would say not worry about high noise floor, that might be fixed by fixing the interconnections. Also, if you are still observing high noise floor when SIDs are removed, then it obviously has nothing to do with SIDs or the interconnections... some issue with voltages like the 9V supply, or ground, or the audio buffers. Try to identify if the noise is on one or both channels perhaps.