DC on output

Status
Not open for further replies.
Jump to Latest
udip

udip

Member
Joined 2004
#1
Hello everybody!!

I have a dc problem on my preamp, and I hope u can help me
find a solution:

I have built a mic preamp, something like the one in linear tech
lt1115 datasheet, with a dc servo instead of caps in the signal line

For the transformer I used SOWTER 9820, with
5k1 and 220pf for termination.

Instead of the LT1115 I used the OPA627 from TI.

For the integrator I used OPA134, 470n capacitors and
220k resistor (instead the 100k and 1uf).

Now here is the thing, I checked for oscillations and other errors
with my oscilloscope, and everything is fine except that I have
50-200mv on the output!! which changes with gain and frequency.

I couldn't find why, does anyone know??

Thanks!!

The schematic:
 

Attachments

  • dc.jpg
    dc.jpg
    67.9 KB · Views: 400
#10
It appears that the +-15V output from the DC servo is not enough to correct the offset of the gain devices through the 10k/10ohm attenuator. Try 100 ohm instead of 10 ohm. It may be also a wiring mistake causing the DC servo to produce positive feedback

To measure the DC component from a waveform with a dominant of AC component just insert a lowpass filter before the multimeter [ie: some resistors and capacitors]
 
#11
Hi Eva,


I inserted a low pass filter, butt the scope could also know what
the frequency was (I don't know how because the pole was
at 1 hz!!)

About the dc, i changed the 10r ro 100r but no change,
also i checked wiring and everything is ok.....

the scope reads between 0 and 200mv dc depending on gain,
it takes about few seconds to the scope to settle and show the
exact dc, is this ok??

The funny thing is that i connected the preamp to an amplifier
and could not hear signs of dc as gain is changing...
 
#17
Thanks everyone...

The problem has been solved (at least i think so).

I think that my mistake was checking dc offset with input
signal.
When i checked offset with no input with and without the servo
the offset was about 2mv to 2 volts without it (during gain changing) and 0 - 1mv when the servo was connected...

so i think this should be ok, no?
 
#19
Eva said:
It appears that the +-15V output from the DC servo is not enough to correct the offset of the gain devices through the 10k/10ohm attenuator. Try 100 ohm instead of 10 ohm. It may be also a wiring mistake causing the DC servo to produce positive feedback

To measure the DC component from a waveform with a dominant of AC component just insert a lowpass filter before the multimeter [ie: some resistors and capacitors]
Click to expand...


I don't think it! If the output of the DC servo is enough to keep the ouput on zero, it goes up or down to rail (say +/-15V), and keep it. No moving between the rails. As I see, the servo can regulate 15-18mV offset on the input. which looks enough to me.

So I think, this is low frequency oscillation. I had this problem with one of my amplifier. The typical sign is the slow triangle wave on the output of the DC servo.

Maybe the transformers cause this problem. I would try to use non inverting DC servo and connect it to the pin2 of the LT1115 via 33-47k series resistor.

sajti
 
#20
sajti, I think you are right..

I checked on scope and I can see a triangular wave on the output
of servo with input, but only with low frequency.

That's servo oscillation?

If i put a non-inverting servo to the - input of opamp, should I
use some sort of bias to the + input in order the eleminate the
feedback to ground capacitor?


Thanks...
 
Status
Not open for further replies.
Top Bottom