Forum: Analog Circuits pure Despair: strange OpAmp behaviour.

von Xer0 (Guest)

Attached files:

Rate this post
0 useful
not useful
Ok, here we go:

I am building a SoundCard DAC for controlling my Laser.
The SC outputs from 0.7 to 3.1V whereby 2.55V is the "middle".
An Amplifier should convert this to a differential signal with 10V p-p.

currently not soldered anything to the SoundCard; im using a USB plug 
for power and for testing the opamp's stage behavior, im was using a 
1MOhm potentiometer to adjust any signal voltage between +5 and -5V. 
now, after i thought the high resistance of it could be the error cause, 
i replaced it with 3x 10k R's at which i can get 4 different voltages 
for testing.

I attached my circuit plan, its exactly how i builded it. between 
+5V/GND and -5V/GND are a 100µF elko each for stabilising, otherwise 
currently no capacitors are used, except the ones for the ICL7660 
negative voltage converter.

So i did a row of different measurements on different setups. Results 
are attached to this Post, too.

         1  |  2  |  3  |  4  
-5V :  -4.82 -4.80 +4.84 +4.95
 A  :  -1.55 -1.55 +0.94 +1.88
 B  :  +1.71 +1.71 -1.03 -2.06
+5V :  +4.99 +4.99 -4.74 -4.76
-vref + -0.94 = 3.49
-vref + +1.03 = 1.52
-4.81 + -(0) = -2.41
-4.81 + -(GND) = -4.76
-4.81 + -(-vref) = -2.27
-1.55 + -(0) = -0.58
-1.55 + -(GND) = -1.16
-1.55 + -(-vref) = +1.39
+1.71 + -(0) = +0.64
+1.71 + -(GND) = +1.29
+1.71 + -(-vref) = +3.84
+4.99 + -(0) = +2.50
+4.99 + -(GND) = +4.96
+4.99 + -(-vref) = +4.95
+vref + -(0) = +1.28
+vref + -(GND) = +2.56
+vref + -(-vref) = +4.95
0 + -(+vref) = -2.55
GND + -(+vref) = -2.54
-vref + -(+vref) = -4.76
0 + -(0) = 0.00
GND + -(0) = 0.00
-vref + (-0) = -1.28

In the table, the 1st collumn is just measured with multimeter on the 
Vin 3*10k divider, withoud any load. second one is the divider voltages 
feed into a voltage-follower opamp circuit, 3rd one is fed into 
inverting-amplifier with 1x gain and 4rd finaly with 2x gain.

So this is strange. the "real" voltages on the divider are coming out 
unaltered with a impendance amplifier / follower; but are changing 
extreme when i use something with gain.
Also in the last test setup, the substractor, when i feed in 2.55V i got 
the half out, without any other voltage is going in which could alter it 
(0 in the table means unconnected cable)
On the other side, whenever i do operations with buffer stabilised 
negative n/o directly from LM336 reference circuit Vref ±2.55V Voltage; 
the output never differs from ideal theory values!

What the...?!

I never get "zero-level" signal output on 2.55V "offset-level" input. SC 
is a CM106 so the Vref is set right. OPA's are rail2rail and able to 
deliver Vout = Vsupply - 0.01V !
Here's a bigger scan: 

von MaWin (Guest)

Rate this post
0 useful
not useful
Whats the problem ?
You wonder, why a voltage divider built up from 10k resistors
is not stable if a load of 10k is attached ?

Use voltage followers for Va und Vb, and 3 and 4 will work.

von Xer0 (Guest)

Rate this post
0 useful
not useful
I always got told the resistance of an opamp is very high an practically 
no current i flowing from the inputs?

I tried it with a 1MOhm divider bevore. why it didn's work there?
And will it do when i use SoundCard Oputput's (after Chip) directly?

von Andreas S. (andreas) (Admin) Flattr this

Rate this post
0 useful
not useful
Correct, the INPUT of the amplifier has a virtually infinite resistance. 
But the additional resistors that you connect from the input to the 
OUTPUT of the amplifier draw current from your voltage divider, thus 
changing the voltage.

von Kai Klaas (Guest)

Rate this post
0 useful
not useful

you entirely forget, that the inverting OPamp circuit is creating a 
"virtual ground" at the inverting input. So, connecting an inverting 
OPamp with a 10k input impedance to your "3 x 10k" voltage divider means 
to connect an additional load of 10k to signal ground, which erodes the 
potentials of your voltage divider, of course.

Kai Klaas

von Xer0 (Guest)

Rate this post
0 useful
not useful
So it can even change the potential when i use a 1Mohm Poti for 

because it even then got wrong...

von Kai Klaas (Guest)

Rate this post
0 useful
not useful
>So it can even change the potential when i use a 1Mohm Poti for

Yes, even more, because then the 10k input impedance of inverting OPamp 
is even more loading the voltage divider. The voltage divider must have 
an impedance that is way smaller then the input impedance of inverting 
OPamp. So, connecting an inverting OPamp to a voltage divider is not a 
good idea in your example. But a voltage follower with its nearly 
infinite input impedance could be used.

Show us in a bit more detail, what exactly you want to implement. An 
inverting OPamp could be used, eventually, for your purpose, but not in 
the way you intend right now. Is it a pot you want to buffer?

Kai Klaas

von sssss (Guest)

Rate this post
0 useful
not useful
Is your soundcard able to put out DC? Is there no highpass?

von Xer0 (Guest)

Rate this post
0 useful
not useful
Ok... you all was right :<

Have used an follower as buffer and done the Tests now. its working!

What i want to build? The Soundcard-DAC for Lasercontrolling!

I didn't drawed my full circuit to scan yet, but can do it.
Merely it converts single ended signal from soundcard, which swings from 
0.7V to 3.8V with 2.55V as middle. thus need to be converted to 
differential output with 10V p-p, also +5V to -5V

von Xer0 (Guest)

Rate this post
0 useful
not useful
@ sssss

We desolder the output coupler wires for that


Entering an e-mail address is optional. If you want to receive reply notifications by e-mail, please log in.

Rules — please read before posting

  • Post long source code as attachment, not in the text
  • Posting advertisements is forbidden.

Formatting options

  • [c]C code[/c]
  • [avrasm]AVR assembler code[/avrasm]
  • [code]code in other languages, ASCII drawings[/code]
  • [math]formula (LaTeX syntax)[/math]

Bild automatisch verkleinern, falls nötig
Note: the original post is older than 6 months. Please don't ask any new questions in this thread, but start a new one.