Does the Sonica get unstable when unloaded?
Mar 3, 2004 at 3:37 PM Thread Starter Post #1 of 30

fiddler

Headphoneus Supremus
Joined
Jun 24, 2001
Posts
1,930
Likes
30
I just modded the Sonica as per Wodgy's suggestions (replacing the output caps with Black Gates, and getting rid of the rest of the output stage) and I'm really happy with the results... very much more detailed sound.

But I have one concern:

When the Sonica is disconnected from the amp, it outputs A LOT of DC, while it doesn't seem to when it sees a load. I measured over 100 mV (!) right after being disconnected, then it gradually went down, settling at around 20 mV.

When the Sonica is connected to the amp, I measured across the inputs and ground of the input jacks, and saw no significant DC, and of course the output of the amp also had negligible amounts DC. That's why I'm suspecting the Sonica is oscillating or something when there's no load connected. This might also explain why the Sonica kept getting jammed up after switching around interconnects-- eventually it would refuse to play unless it was unplugged/replugged from the USB port.

I just removed the input caps on my META42 the other day, but I'm contemplating putting them back in, because every time I unplug the Sonica I get a very loud scratching/poppping noise, probably due to a DC spike? Could this be enough to damage my headphones, or should I not worry about it too much?

Otherwise I'm very happy with the mod.
biggrin.gif
I also changed the output jack to a switchcraft panel-mount type, becaues I was unhappy with the default plastic one. It didn't quite fit, so I had to file down the area a bit... ehehe.. Ah well, it works.

So yeah, could Sonica users (who have a meter handy) measure the DC offset for me? Damn, I hope mine's not defective considering I just majorly voided the warranty!
biggrin.gif
I was noticing the pop before the mod, but didn't think much of it until now.
 
Mar 3, 2004 at 3:45 PM Post #2 of 30
Man.. ok, I tested the DC output of the amp when the IC's left connector is touching the ground on the Sonica (i.e., the way it would touch the instant when you plug it in) I measured a bit over 600 mV from the META42's output.
eek.gif
As I said... the DC goes right back down to normal once the plug is fully in.

Also moderator, could you please move this over to the DIY forum? I just realized I'd probably get more help over there.
 
Mar 3, 2004 at 3:57 PM Post #3 of 30
you are getting 600mV ???????!!!!!!!
eek.gif


that is over half a volt and no way that is good for your cans

Put a DC blocking cap either at the input of your amp ot the output of the source

Can you point me to wodgies mods so i can check the changes out ?

rick
 
Mar 3, 2004 at 4:00 PM Post #4 of 30
BTW-not knowing the topology off hand of the sonica output amp it does sound like an unbalanced output , one that does not negate DC output by cancellation so if there was once a blocking cap there it was for a valid reason

the only time you can safely eliminate caps is either with fully balanced topologies , a dc servo or if the next stage has a cap already

Locjup issues : did you make any power supply cahnges to the sonica ?
 
Mar 3, 2004 at 4:08 PM Post #5 of 30
No, there always was an output coupling cap, and all I did was swap it with a higher quality one (Black Gate). The 600 mV was the offset after it had been amplified by the META42, so the input offset musta been around 100 mV or so. Yes, 100 mV, even with an output cap.
eek.gif
I heard the popping-when-disconnecting noise before the mod, so I assume the problem existed before I did anything. As I said, it's only when the IC is being disconnected does it become an issue. No changes to the power supply (yet
wink.gif
).

I think I'm putting the input cap back on, just in case. Imagine the cable getting partially unplugged and staying in the DC-outputting positoin all night! Sizzle sizzle.
 
Mar 3, 2004 at 4:11 PM Post #6 of 30
stupid question ,did you take out a non polarized cap and insert a polarized cap in its place ?

or

if both were polarized are you sure you got the polarity correct ?
 
Mar 3, 2004 at 4:15 PM Post #8 of 30
and BTW -if the popping was always there then it is a design thing and since it is being magnified by the DC issue it sounds far worse

actually it is worse due to the already present 600mv DC componant which is probably ramping up much higher with the pulse and then "amped" by your META

you really are at risk of destroyoing your cans here
 
Mar 3, 2004 at 4:17 PM Post #9 of 30
Yes, they were both polorized, polarity is correct.. The other change I made: there was a voltage divider right after the electrolytic cap... I removed that. Also, there was a small ceramic cap going from each output to ground, those were removed, as well.
 
Mar 3, 2004 at 4:17 PM Post #10 of 30
oh yeah , turn the volume all the way down before turn off to eliminate the thunp/pulse

good practice with all headphone amps is to mute the output at turn on/off unless an output muting relay is a built in feature
 
Mar 3, 2004 at 4:18 PM Post #11 of 30
Rick rick rick, the DC is not there once the cable is fully plugged in! No DC from the Sonica, no DC from the amp.
 
Mar 3, 2004 at 4:34 PM Post #12 of 30
easy dude !

hehe ,OK man-

follow me here

Unloaded sonica DC present

sonica hooked to the amp-no DC ?

did i not read there was 600mv DC "amped" with the meta ?

the "cable' you seak of is the USB correct ?

so once the usb cable is plugged in the DC goes away but when unplugged to go portable the DC comes back ?

Is this a correct statement ?

If not then you have me totally confused man
 
Mar 3, 2004 at 4:46 PM Post #13 of 30
Yes, i've confused you.

The USB cable is connected at all times.

The sonica has a 3.5mm analog out. I'm talking about that connection. When the cable is connected in such a way that the left channel connetor (the tip) touches the sleeve of the output jack, I see the offset from the amp. I.e., every time the cable is unplugged from the Sonica's output, *insert Rice Krispies motto*.

... you know, snap crackle and pop.
rolleyes.gif
.


OK, i put the input cap back in, and in the same situation, the cap blocks *some* of the DC, but I still see about 200-300 mV at medium volume level.
 
Mar 3, 2004 at 5:36 PM Post #14 of 30
Did you remove all the SMD parts between the two electros and the output jack mounting point? I believe there are a total of size parts, 4 resistors and 2 capacitors.

I've replaced the electros, but I haven't removed the SMD parts. If you're saying it works well without them, I'll take them off tonight and try it.

BTW, I used Black Gate NX parts to replace the stock caps. Anything wrong with that?
 

Users who are viewing this thread

Back
Top