Post Filter Feedback with Average Current Mode Control

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
These things are known.
 

Attachments

  • Screenshot from 2018-10-22 14-28-41.png
    Screenshot from 2018-10-22 14-28-41.png
    60.1 KB · Views: 271
  • DraftTri_03.asc
    7.1 KB · Views: 28
Current sensing could be improved a bit by sensing both rails instead and deriving a push-pull output thereby facilitating bridge operation as well.

Yes these things are known in the trade but seldom used by the fraternity due to lack of skills in the art.
 
Places the inductor in an internal loop to control its current reducing the second order LC filter response to a first order current driving a capacitor response. The filter resonance is 'hidden'.

In the case of a triangle wave modulator achieving slope matching defines the crossover frequency as being Fs/Pi, both edges are sampled.

The voltage loop is closed around the current loop. Say Fs is 500KHz the current loop crosses over at 160KHz. If the current loop has a sensitivity of 10A/V crossover would nominally occur when the capacitor has an impedance of 0R1.

1uF would be 1.59MHz. Closing the voltage loop at, say, 80KHz gives the voltage error amplifier gain as being 0.05. Pole capacitors are added to the amplifiers as required.


Unfortunately achievable bandwidth is still limited however it is still possible to get something that seems to get the job done.

https://e2echina.ti.com/cfs-file/__...-Mode-Control-of-Switching-Power-Supplies.pdf

Here is a rehash... Switching model and Linear model including an effort to control things when clipping occurs. Plot V(a)/V(b) to see the loop gain with VDEM set to AC 0V and VAC set to AC 1V.

Do the same with E1 set to -0 and R9 to 8K2 to see the loop gain when limiting occurs.
 

Attachments

  • Screenshot from 2018-10-25 11-14-34.png
    Screenshot from 2018-10-25 11-14-34.png
    56.2 KB · Views: 193
  • Screenshot from 2018-10-25 11-15-18.png
    Screenshot from 2018-10-25 11-15-18.png
    60.5 KB · Views: 189
  • fulla.asc
    7.8 KB · Views: 31
  • vfba.asc
    4.6 KB · Views: 20

Thanks for the link.
I haven't seen Current mode control discussed for Class-D amplifiers, usually it is associated with Switch Mode Power Supplies.
They are similar, of course, but have a different context.
Since you seem to have power supply expertise perhaps you could comment on my thread State of the Art for Tracker/Down Converter? in the Power Supply sub-forum?
In your first attachment there is no earth reference for the Vdd Vss power supplies, I assume there should be?

Best wishes
David
 
Last edited:
The method and location of the current sense element is sometimes, always, a problem.

TroelsM suggests a differential location. Presumably two resistors with a diff amp to sense supply current which is in effect inductor current.

It would still mess up or otherwise influence system ground especially if you were trying to use multiple amplifiers on the same power supplies.

Ideally you might/would wish to sense inductor current directly. Perhaps that is what TroelsM meant. Sense in series with the inductor with a differential amplifier to reference the result back to ground.
 
The method and location of the current sense element is sometimes, always, a problem.

Ideally...sense inductor current directly.

Have you considered a Hall effect current sensor?

Best wishes
David

Maybe hard to find the frequency response you need.
The Hall effect itself may not be a problem but many of the ICs seem slow.
Are any made for this application?
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.