8580 Audio Output..

[davotron]

Firstly, hello everyone, I'm a newbie here..

Secondly, after much cogitation and listening to the awesome sounds of the elektron sidstation (but being scared away by the price.. ) I decided to build a simple SID + CORE with a minimal control surface and being an electronic engineer I thought "right, it should be straight forward..", but I have run into a couple of problems;

I have built the Core and SID module, and the core is up and running nicely (although I might add I changed the pull up to MCLR to 1k rather than 100R [5mA rather than 50mA I believe flowing out the pin], I found the p18f4585 was getting stuck in a state of constant reset due to a noisy PSU I have, got rid of it though.. and added a reset).  I downloaded MIOS to the core easily and the LCD gave me sensible characters, and I get midi comms and can debug using MIOS studio.

I hooked the SID module up to the Core and using MIOS studio downloaded the sid test tone app to the core, but don't really get any output.  I believe it is there, but very quiet. Using a mic amp with 75 dB's of gain you can hear a tone, albeit with a lot of noise :D and I'm certain it isn't power supply whine or oscillation as it goes away with just MIOS running with no app.  Is there a common problem with dead 8580's giving very quiet outputs?  I'm not sure the best way to test it - when I run the interconnection test does sending different mod wheel values to the core change the states of the SID address and data pins (i.e. the parallel outputs of the shift registers)?!

Secondly, I'm having big problems re-compiling my SID app but I'll go post that in the right place :)

Cheers TK for a great project

David

[nILS]

Firstly, hello everyone, I'm a newbie here..

Welcome aboard!

Secondly, after much cogitation and listening to the awesome sounds of the elektron sidstation...

mbSID kicks the shit out of a SIDstation ;-)

I found the p18f4585 was getting stuck in a state of constant reset due to a noisy PSU I have, got rid of it though..

That must have been incredibly noisey. I've tried some really really crappy and extremely noisey PSUs but never managed to actually get the core to reboot because of that.

I downloaded MIOS to the core

Just me being picky: you uploaded MIOS to the core ;-)

I believe it is there, but very quiet.

So far there's tons of possible reasons for that.

Is there a common problem with dead 8580's giving very quiet outputs?

Not really. There is a tendency for sid chips to die, especially single oscillators or filters, usually not the entire chip.

when I run the interconnection test does sending different mod wheel values to the core change the states of the SID address and data pins (i.e. the parallel outputs of the shift registers)?!

Yes. Do that. Run the test and see if the data gets there correctly.

Also measure all voltages (8580 runs on different voltage than the 6581) and check all wiring.

Secondly, I'm having big problems re-compiling my SID app but I'll go post that in the right place :)

I'll be on the lookout for that thread, but since I'm already at it:

http://www.midibox.org/dokuwiki/doku.php?id=windows_toolchain_core

http://www.midibox.org/dokuwiki/doku.php?id=windows_toolchain_quickstart

[davotron]

thanks for the help!

Of course it is uploaded, my mistake! I've said that again in my other post :D

I'm not sure what was wrong with the PSU I was using, it had insane amounts of ripple or very poor rejection, but changing that pull up definitely fixed it - microchip suggest a 1k - 10k in the datasheet I believe.

All voltages/connections are good - although I'm sure one should measure 4.7v DC on the audio output, that the signal rides on, I see more along the lines of 500 mV which is not good!  Does that 1k pull down just provide a DC path for the ac coupling cap?  I'm not sure what the internals to this pin are and what the output impedance is etc; I THINK it is an open drain output so it probably pulls the drain down.  I'll try disconnecting, then measure it.

David

[nILS]

Even without putting too much thought into the whole thing or knowing anything about electronics I think it's fair to say that the parts and values used in the sid and core module  have been tested and found to be working in a couple thousand cases. If they don't work for you - you've got a problem somewhere. Try finding and fixing the root of that problem (get two decent PSUs for starters) instead of going with some odd workaround.

[davotron]

Ah yeah of course, this I understand, and it's this is what makes it interesting!

Yep got a tidy regulated bench supply in the end.  The SID gets Vdd @ pin 28 of around +8.9 to 9V, current limit at 1500mA so there's plenty of power there if it so requires; Vcc around 4.8V.

David

[Wilba]

This is so deja vu...

This is so deja vu...

http://www.midibox.org/forum/index.php/topic,12111.msg99563.html#msg99563

Check the SID is getting 12V or 9V at pin 28.

If it does not, then the SID's audio section isn't being supplied with power, yet somehow it leaks out some audio signal.

Alternately, the audio output buffer (the BC547 transistor etc.) could be faulty, to test this, see instructions in the linked thread above.

[davotron]

After running the interconnections app. I discovered I definitely have a connection problem between CORE and SID! (which, incidentally means I might not have an audio output problem.. but doesn't explain the weak tone I was getting yet) I'll go over it properly tonight.

Cheers nILS Podewski for that quick software build guide, I was concentrating on doing it in Mac OS X but that sorted me right out for now!

Update: It seems that the silk screened "SC" on my SID V3 board actually goes to the "MD" pin of J10 on the CORE; this is the signal from pin 28 (RD5/PSP5/P1B) of the PIC to the SCLK storage reg clock input of the 595's, have I missed a change log somewhere that I couldn't find?  It has cured it anyway!

[nILS]

Um...

http://ucapps.de/mbhp/mbhp_8xsid_c64_psu_optimized.pdf

[davotron]

I don't 'geddit?  I'm probably being completely stupid.  There is no "MD" silk screened on the V3 board, and the "SC" pin seems to go to where the "MD" pin connects to according to Thorstens (that link) drawing for the V3 PCB layout I have or are you referring to the "SO" to J14 for the right SID?

Edit: I actually have a V4 PCB, I got from smashtv, I can't seem to find the layout data for this though.

[nILS]

I'm not referring to anything. I'm just showing you a pdf that shows you how the wiring should be ;-) It's basically a 1:1:1, only difference being the SO pins.

[davotron]

I found the R5 Din board has a similar thing; the "SO" pin of J1 ("SO" being screened on the PCB) isn't actually connected to the serial out of the shift reg - the one adjacent to it is that has no screened text.

[stryd_one]

the "SO" pin of J1 ("SO" being screened on the PCB) isn't actually connected to the serial out of the shift reg

Sounds correct. Why would you connect a serial output to a serial output? Wouldn't do much, would it? :)

(Actually, it might even short)

[davotron]

Ah that old conventions chesnut... the number of long winded conversations we used to have at ssl over this one, ha!  I totally agree, so why does that diagram have "SO" to "SO" for the core to one of the SID's if I was following that?  And on my Din the actual serial output pin has no screened label.

[stryd_one]

[me=stryd_one]removes foot from mouth[/me]

Looks like TK uses the 'other' convention, where the label is the same at both ends. Once again I rush a post and screw up.

[nILS]

You're chewing on your toenails? Eww. Haven't we noticed that before? (the pin names, not the chewing) ;-)