Op Amp Balanced Output

Thread Starter

SamR

Joined Mar 19, 2019
5,031
Still working with the TI Op Amp Handbook and was given this.
1644340400063.png
If I interpreted this correctly, using Eo- as a reference GND and inputting Ei=1Vpp, I should see 4Vpp on Channel 4 @ Eo+? So I breadboarded it using TL071CPs with Ei=1Vpp @ 200Hz.
1644341034756.png
And I'm getting 1Vpp, not 4Vpp on channel 4...
1644341350790.png
What did I miss?
 

Jony130

Joined Feb 17, 2009
5,487
Try disconnecting all scop probles from the circuit. And using only one problem only, connect the probe tip to Eo+ and GND clip to Eo-. All other probes should be disconnected from the scope.
 

Jony130

Joined Feb 17, 2009
5,487
Well, I think that the main problem here is that all GND clips in a scope are directly connected and they are also connected to the AC mains PE wire as well.
So you cannot have separate "GND" in a circuit. Also, the external sig gen GND clip, will also be connected to AC mains PE wire. Thus, we can easily do a short circuit.
 
Last edited:

Thread Starter

SamR

Joined Mar 19, 2019
5,031
IMO there should be a negative power supply to the OPA's
There is and it is common to the sig gen. OPAs supplied ±15.

And with only Ch4 connected it is 2XEi but not 4... The sig gen I'm using has no ground and the PSU & scope GND are tied. The sig gen negative is tied to the common GND for PSU & scope. Removing the sig gen negative from the common GND just gives noise.
1644345752205.png
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
They are just what I would expect, Eo+ is the Ei 1Vpp inverted and Eo- is the 1Vpp Eo+ inverted.
1644347153071.png
 

Jony130

Joined Feb 17, 2009
5,487
Yes, indeed everything is working as expected. But your signal gen is 0.5Vpeak, not 1Vpeak. So Vout_pp = ("Eo+" -"Eo-") = 4*0.5Vp = 2Vpp.
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
2Ei I can understand. Wonder if 4Ei is another error?
your signal gen is 0.5Vpeak, not 1Vpeak
I don't think so. I lowered the PSUs to ±10V and ran the sig gen up to 17Vpp without any clipping on Eo+. Then connected Ch1 to Eo+ and it's GND to Eo- and got this. 34Vpp without clipping from a 17Vpp input with ±10V rails. The 4Ei must be a mistake...
1644348571272.png
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
What was I doing wrong with the multiple channels and the common ground? I can see that because I did there were 2 different reference grounds for the scope, but... I can't follow it...
 

Jony130

Joined Feb 17, 2009
5,487
Because I don't know any better? Why should I be using 10X?
You should always be using 10X


have the sig gen set on 17.000V @ 200.00Hz and this is what I get for Ei. 17Vpp
Sure, but the peak value is 8.5V. And in the Ti note, they said that you will get the peak to peak output voltage will be equal to 4-times input signal peak value.

What was I doing wrong with the multiple channels and the common ground?
But if all GND clips are connected to the same common ground in a circuit, you will not encounter any problem.
But if you for example connect one GND clip to Eo-. And the second GND clip to "real circuit GND", you will get a short circuit. Because scope GND clips are connected together.
 

Jony130

Joined Feb 17, 2009
5,487
bsolutely impossible. Your TL071 opamps are not rail to rail. With a ±10V the output will be a little less than ±9.V.
The sig gen peak value was around 8.5Vpeak So it was possible to get 34V peak to peak at the output because of a "bridge" connection (not reference to GND).
 

crutschow

Joined Mar 14, 2008
34,280
I'll keep the probes on 10x from now on.
Yes, unless you need the sensitivity that a 1X probe provides, always use the probe at 10X since it significantly reduces both the resistive and capacitive loading of the probe (less likely to distort the signal you are measuring).
 
Top