Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

[SOLVED] CMFB for fully differential OPAMP

Status
Not open for further replies.
P

Phoibus

Junior Member level 3
Junior Member level 3
Joined
Jul 29, 2010
Messages
26
Helped
3
Reputation
6
Reaction score
3
Trophy points
1,283
Location
Cluj
Activity points
1,472
Hello,

I have designed a fully differential Miller OpAMP with CMFB. I need it to create a library of elementary filtering stages (first order LP and HP, second order LP, HP and BP, the biquad topologies etc).The amplifier appears to work well. in an amplifier connection and in lowpass. However, for highpass and bandpass the CMFB doesn't work and consequently the filtering characteristic is all ruined.

My guess is that because of the longitudinal capacitance, the common mode Vref voltage doesn't reach the OpAMP input (in DC). What could the solution be to my problem?

Any suggestions are welcome.
Thanks,
Paul
 

I'm also uploading the OpAMp schematic.

Low-pass topologies work well with this OpAMP. However, when there is a longitudinal capacitor in the signal path to the OpAMP input, as is the case for low-pass or band-pass filters, the common mode feedback malfunctions and the filter behavior is compromised.

I'm also uploading the high-pass filter topologies.

If you have any idea where the problem lies, please let me know. Any suggestion is extremely valuable.

Thank you for your help.

Paul
 

Attachments

  • OpAMP4.pdf
    20.2 KB · Views: 197
  • Derivator.pdf
    12.9 KB · Views: 197
  • dualpole_hp.pdf
    13.2 KB · Views: 197

Phoibus said:
... My guess is that because of the longitudinal capacitance, the common mode Vref voltage doesn't reach the OpAMP input (in DC). What could the solution be to my problem?
Click to expand...
Use an additional parallel DC path, s. the foll. PDF: View attachment CMFB4fully-differential_OpAmps.pdf

And I think the paper from Jian Xu et al. "Low Voltage Low Power Current Mirror OTA for Sigma-Delta Modulator" (IEEE 2007) might be helpful.
 
Thank you for the suggestion. I did so, but it didn't work as expected.

The additional parallel resistance you suggested introduces a transmission zero. So:
1. According to hand calculations, the additional parallel resistance needs to be infinity in order for its effect to be neutralized.
2. Very high value for the resistance (i.e. grater than 1 meg) still causes the cmfb to malfunction.
3. With a smaller resistance value, the hpf actually behaves like a circuit with two different amplifications (for different frequencies).

Once again, thanks for your help.

If you have some other ideas, please tell me.

Paul
 

Phoibus said:
1. According to hand calculations, the additional parallel resistance needs to be infinity in order for its effect to be neutralized.
2. Very high value for the resistance (i.e. grater than 1 meg) still causes the cmfb to malfunction.
...
If you have some other ideas, please tell me.
Click to expand...

You'd need resistor values in the order of GΩ (1e9, s. the PDF above), but I know this is impractical for ICs.
Did you check out the paper I suggested above?
 

I've tried resistor values in the order of GHz, but the CMFB won't work. Actually, if there is no DC path to the OpAMP input, differential input stage isn't biased. So, it is explainable that it doesn't work.

In this case, the question translates to, "What is wrong with my biasing".

Unfortunately, I don't have access to IEEE, and I couldn't find another copy of the article over the net.
 

OK, I still didn't fix the problem. But at least I know where the problem lies. Indeed, the capacitor breaks the DC biasing path and the OpAMP input stage isn't biased (at 1.5V common mode).
The solution would bi to bias the input stage. Now the question is, how can i bias the input?
If you have any suggestions, please let me know.

Thanks!!!
 

OK, I've got it!!!

Single-pole filters, Sallen Key and Multiple Feedback Biquads are impossible to implement with fully differential OpAMP. The problem is teh one stated in my former post, theer is no DC path for biasing. As such, you need either a Tow Thomas biquad or a KHN biquad.

I wanted to have this written, maybe it will help someone in the future.

I want to thank you very much for your help!!!
 
  • Like
Reactions: erikl

    erikl

    erikl

    Points: 2
    Helpful Answer Positive Rating
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top