I performed my first snes digital audio mod today and it works fine for about 15 minutes and then the audio cuts out. I'm guessing it's a ground problem, but I've triple checked my wiring and nothing is shorted.
When the audio cuts out the light from the spdif connector shuts off completely and if I reset the snes I get garbled audio. I need to disconnect the snes from the ac adapter and turn it on to discharge the machine before the snes sound will work normally again. I followed the cs8406 wiring diagram exactly at gamesx. I have heard that there should be a capacitor between ground and I don't have one. Could this be it?
Just in case it relevant, this snes also has a svideo mod so that I can use whatever svideo cable I want with the snes. It's just a straight through connection from the multi out to a female svideo connector. Also, I had to bend the the metal case around the s-apu so that I could feed the wires to the ic's inside. Strangely, when I ran the snes without the metal case the audio problems seems to occur sooner.
I also grabbed my ground for the cs8406 from the 7805 vreg leg pin, is that ok?
Ren01 wrote:
I followed the cs8406 wiring diagram exactly at gamesx.
Context:
http://gamesx.com/wiki/doku.php?id=av:snes_sp_difQuote:
I have heard that there should be a capacitor between ground and I don't have one. Could this be it?
Certainly could be, but usually that would pose a problem sooner than 15 minutes in...
Does anything get hot once the CS8406 has stopped working?
Quote:
Strangely, when I ran the snes without the metal case the audio problems seems to occur sooner.
That vaguely implies you've got a crappy connection to your power rails.
Thanks for the help! No harm in running more or larger wire to the 7805 5v output right? I'll try that tomorrow. Right now I'm trying to isolate the problem by checking if the analog audio is still working when the optical cuts out. I'll also try a different spdif connector since I got these for cheap on ebay.
Ren01 wrote:
No harm in running more or larger wire to the 7805 5v output right?
Nah. You might try making your own twisted pair (throw two wires in a drill chuck, hold the other end) of something higher-ish gauge... although 13mA really shouldn't need much at all. Probably worth double-checking how clean your solder joints are, or if you're using stranded wire maybe they're not all intact. Worth trying the capacitor, but you shouldn't need more than 100pF-100nF.
I twisted a pair of wires for the 5v and the ground, but it had no effect. audio still drops out after 15min to an hour. I tried a different optical connector and it had the same problem. I cleaned all the flux on my TSSOP adapter as well. To be honest, my wire job is a mess and pretty sloppy, so I wonder if noise is a problem.
When the audio cuts out from the optical, the board optical out connector still produces light and then eventually turns off. However, the analog audio from the multi out still works.
I have a set of computer speakers with optical in I can try to see if it's the receiver causing issues. The receiver is a onkyo 876.
Here's my mess:
Ren01 wrote:
To be honest, my wire job is a mess and pretty sloppy, so I wonder if noise is a problem.
Given that you previously said that it's worse without the RF shield? Assuredly.
Also, with that wiring, you should really have a bypass cap on the SPDIF board.
Nothing's warm though? Odd.
I'd probably add the bypass cap, shorten the jumpers, and if it's still touchy, add shielding (either coax or twisted pair) to any signal faster than ~1MHz.
lidnariq wrote:
Ren01 wrote:
To be honest, my wire job is a mess and pretty sloppy, so I wonder if noise is a problem.
Given that you previously said that it's worse without the RF shield? Assuredly.
Also, with that wiring, you should really have a bypass cap on the SPDIF board.
Nothing's warm though? Odd.
I'd probably add the bypass cap, shorten the jumpers, and if it's still touchy, add shielding (either coax or twisted pair) to any signal faster than ~1MHz.
I shortened the wires and ran separate wires for all the different grounds on the chip and it's still doing it. I'll try the bypass cap as soon as I can find a capacitor of appropriate value. I don' think there are any signals over the 1mhz range in this circuit. Nothing's warm though! Thanks for the help!
ISCLK, OMCK, and maybe SDIN are all fast enough to count as "fast enough to conceivably be a problem"
What capacitors do you have on hand?
I desoldered the smallest caps I could find on a dead power supply, they're 2.2uf. I tested the snes for for nearly two hours while I watched a movie and the optical light stayed on, but there was no sound. Tempted to rip the whole thing apart and start from scratch. I will try the snes with a different DAC tomorrow.
Also, I don't know if I'm just off, but the snes sounds better with the bypass cap in place. The deeper bass frequencies are a lot more detailed. After one more test today, when the sound stops playing on my receiver the optical spdif connector is still blinking, which I can only assume means it's still trying to send date. Eventually, the light from the spdif connector turns off completely.
I tried 4 2.2uf capacitors in series as a bypass capacitor and I'm still having problems.
Lacking any better ideas at this point, without test equipment ... try hooking up a button to the /RST input? If nothing else, that will hopefully localize the problem into the CS8406 or something else.
Do I just hook it up to ground and the reset on the 8406?
You probably want something like
+5V—1 kΩ resistor—/RST input—switch—Gnd
Good news! The snes has been running for 3 hous straight without problem. After you mentioned the reset button I reread the datahsheet for the cs8406 and wired the reset directly to the 5v rail. This could cause some some distortion on startup, but since my receiver delays audio for a few seconds after the snes turns on there's no issues
thanks for all the help!
Had you left /RST floating?
I'm not really sure, but according to my ohmmeter it was wired correctly. Snes ran overnight and the audio never dropped.
Ren01 wrote:
I'm not really sure, but according to my ohmmeter it was wired correctly. Snes ran overnight and the audio never dropped.
If reset was floating before, then it explains all of your previous behavior. No inputs should be left floating unless a datasheet specifies that an internal pull-up or pull-down resistor is present.