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.

SMPS circuit showing an oscillation in the output current...current monitor problem?

Status
Not open for further replies.
T

treez

Guest
Hello,
The attached SMPS circuit is showing an oscillation in the output current. That is, the output current as seen at the output of an ACS722 current sensor, as in the attached.
Vin = 30v, vout=48v, iout = 3.2A (Av)

I got kicked out of the lab by the cleaner before I could get my teeth into it (going back in tomorrow).

However, I suspect the problem is with the ACS722 current monitor itself, and that the VIOUT pin is not seeing a min load of 4k7, would you agree?
The ACS722 is setup for 20kHz bandwidth.

I see no oscillation on Vout.

I got kicked out too soon, but the oscillation was around 60kHz……and was about 0.8Apkpk.

ACS722 current monitor (10A, unidirectional)
https://www.google.com/url?sa=t&rct...tasheet.ashx&usg=AOvVaw3JPQuBETUeMAFE0wXfP4gr
 

Attachments

  • DCDC SMPS circuit with oscillation.pdf
    77.5 KB · Views: 95
Last edited by a moderator:

I wouldn't be too quick to blame the sensor, I have
seen similar-sounding "sloshing" in POL DC-DCs
when pushing up against stability limits in search
of best step response.

Obviously you ought to address any discrepancies
in the sensor application vs specs.

You don't mention the fSW. From the control chip
pinout I'm guessing this is a hysteretic controller
and maybe the frequency is not fixed. But a question
is, how does the 60kHz oscillation relate, if it relates,
to @OP switching frequency and perhaps subharmonics.
Is 0.8A just the baseline designed ripple current? Or
ripple exacerbated by cycle-cycle asymmetry?
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
The ACS722 is setup for 20kHz bandwidth.

I see no oscillation on Vout.

I got kicked out too soon, but the oscillation was around 60kHz

According to the ACS722`s datasheet, and from the attached schematic above, by pulling up pin 6, you have selected the 80KHz bandwidth mode, not the 20KHz.

Anyway, why did you selected the CL capacitance ( C5 in your design ) much smaller the value that manufacturer has used for characterizing device specifications (actually, 1 tenth)? Increasing this one would decrease bandwidth.
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
sorry yes, there are two fittable resistors to select ACS722 bandwidth.....on the board i have it set up for 20kHz...(sorry the schem indicates otherwise)
The Fsw of the DCDC module is some 750kHz.
The ACS722 datasheet seems to be asking for a 4k7 resistor (or lower value) to be connected to pin VIOUT?
 

The ACS722 datasheet seems to be asking for a 4k7 resistor (or lower value) to be connected to pin VIOUT?
No. Datasheet says minimal 4k7.

ACS722 has very limited capabilities to generate a 60 kHz oscillations. It's much more likely that it's originated by the switcher module. There should be no problem to locate the 60 kHz source by measurements.
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
thanks ive looked at the vout and vin under AC coupling, and i cannot see the oscillation there at all....but i see your point.....the noise afflicting the ACS722 could still be coming from the vicor module......the freq of the oscillation is 80khz...i got chance to measure it now.

- - - Updated - - -

ive RC filtered the supply of the ACS722, but to no avail.........setting the ACS722 to 80khz BW doubles the oscillation peak-to-peak (compared to 20kHz)

- - - Updated - - -

No. Datasheet says minimal 4k7.
..thanks...so it really does need a resistor to be connected from VIOUT to GND?
...and regarding minimal.....thanks, i wonder what is the maximal needed.
It seems a little bizarre that it needs a resistor from VIOUT to GND.?
I added 4k7 from the VIOUT pin to GND and it didnt help or improve things.
 

I added 4k7 from the VIOUT pin to GND and it didnt help or improve things.

Did you review what was mentioned above regarding C5?
 

Thanks, changing C5 makes no difference......i had C5=3n3 and it made it no better.

I removed R23 and shoved 3.2A through the ACS722 current sensor from a lab PSU….with the vicor dcdc module not running, and there is then no oscillation problem…..so it looks like this is the vicor DCDC module causing noise which makes the ACS722 current sensor oscillate in its signal output from its VIOUT pin

The oscillation at the signal output of the ACS722 Hall sensor gets twice the magnitude when I connect the Hall sensor for 80khz bandwidth, instead of 20khz.
.....................*.................*.....................*.................*.....................*.................*.....................*.................*.....................*.................*.....................*.................*.....................*.................*
Ive a terrible feeling that since the ACS722 is a Hall sensor, it relies on having a precise magnetic field inside it…and very possibly, with the sensor being just 1cm away from a powerful DCDC module (as shown attached) , the magnetic field inside the ACS722 Hall sensor chip is getting disturbed….producing the oscillation. Would you agree?
We wanted to use the Hall sensor because it is isolated…..and the control ground , where the current measurement will be used, is not the same ground as the output ground of the DCDC module….so an isolated Hall sensor seemed like the best choice…no I don’t believe it is.
 

Attachments

  • PCB with hall sensor.jpg
    PCB with hall sensor.jpg
    246.9 KB · Views: 234
Last edited by a moderator:

Do you agree we will have to shield the Hall sensor from the DCDC module?..ie, with some metal strip
 

Do you agree we will have to shield the Hall sensor from the DCDC module?..ie, with some metal strip
These open path sensors have a certain sueceptibility to external magnetical fields, from conductors as well as from transformers and inductors with air gap. But there's considerable distance to the module. To test the effect, you would operate the sensor with disconnected primary terminals.

I rather guess, you are measuring real current.

If it comes to shielding, are you considering AC or AC + DC fields?
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Yeah just rework to lift the allegro out of the circuit and short around it. Now you’ll only see the result of noise pickup.
 
  • Like
Reactions: treez

    T

    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