Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Deskamp power supply schematic.
New Posts  All Forums:Forum Nav:

Deskamp power supply schematic. - Page 8

post #106 of 112
4/23/11 at 1:42pm

If you change the 1k resistor to 2k and the 2 ohm resistor

to 4 ohms, you should be safe at +-12 volts and maybe

even get up to +-15 volts like you want.

Reply
post #107 of 112
4/23/11 at 1:44pm

ooopppsss...double post...

Reply
post #108 of 112
4/24/11 at 2:43pm
Thread Starter 
So I plugged in the 2k resistors and used two 2ohm resistors in series to get 4ohm. Now on the left channel i'm seeing 150mV DC offset and on the right I see a whopping 3.5V. Obviously something wrong there. I plan to investigate more tonight.
Reply
post #109 of 112
4/24/11 at 2:55pm

With the op-amp feedback, the DC offset should be very low, 1mV or so.

Reply
post #110 of 112
4/24/11 at 10:01pm
Thread Starter 

It's strange, I just rebuilt the circuit, (not sure what I had wrong before but this is always a good way to get a feel for the circuit) and this time the DC offset starts in a reasonable position (about 11mV for the right channel and -12mV for the left.)  However the voltage steadily INCREASES on both channels, quickly going above 20mV on the right channel and eventually (after about 30s or so) on the left channel as well.

 

I'm wondering if I should have increased the resistance of the 20ohm resistors as well, to 40ohms, like was done with the other resistors.

 

Also, for future refference, if I wanted to do a class A bias I would tie the JFETs to the output BEFORE the transistor stage of the amp, right?

 

EDIT: I guess in the case of this circuit the DC offset is greatly affected by the power supply.  Lowering the voltage to +-12V allowed the right channel to settle to -10mV.  It settled fairly quickly but the left channel took some time.  It STARTED at -70mV and settled at -29mV ISH.  The left channel would bounce between -27.8mV and 029.5mV)

 

EDIT2: OK, I got the balls to actually TRY it, even with the DC offset being sub-optimal, and it's clear.  It sounds good, even on the crappy Logitech headset.  Hell, I even plugged my DT990s into it and it sounded great.  Honestly... it's a relief to hear good quality sound from it, finally.  I just need to nail down the DC offset a bit.  Then I get to come up with a board layout for it.  I'm tempted to buy a copper clad board and cut the paths I need.

 

EDIT3: Actually, part of the problem here seems to be how unstable this circuit is on a breadboard.  I attempted to switch Op-Amps from the OPA2134 to the LME49860 to see what changes that would make and must have bumped something sensitive as now the offset on both channels is up to about 400-600mV.  =/  It's getting hard to get a feel for this... or maybe I'm just bad at breadboarding.


Edited by miskatcitnamor - 4/24/11 at 11:06pm
Reply
post #111 of 112
4/25/11 at 5:55am

Keep the 20 ohm resistor at 20 ohms.

This is a Class AB buffer and the 20 ohm resistor sets

the depth of the class A. When it comes time to bias

the op amp into class A, the jfet goes on the output

of the op amp before the buffer. Another troubleshooting

step you can take is to measure the DC offset on the

output of the op amp. As the buffer has no gain, if the

feedback is working correctly, the DC offset at the output

of the buffer should be close to zero and the DC offset

at the output of the op amp should be the value the op

amp is using to correct the buffer. If the DC offset on the

op amp is close to zero and the offset on the buffers is

all over the place, then the feedback is not working.

Reply
post #112 of 112
4/25/11 at 7:22am

Here is a new schematic with some values for troubleshooting.

 

AudioCUBE.png

Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Deskamp power supply schematic.