• State TI Thinks Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 5 answers
  • Subscribers 62 subscribers
  • Views 591 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UCC28951: Current sense

Ahmed Eldistawy
Prodigy 100 points
Part Number: UCC28951


The datasheet recommends using a current transformer to sense the supply current to implement peak current mode control. The switching frequency is programmable until 1 MHz according to the datasheet as well. But the best current transformers have a bandwidth in the range of 1 MHz. Is that enough to implement peak current mode control for switching frequencies in the range of hundreds of kHz?

Second question also related to the current sensing. If the converter is operated in voltage mode control. Does sensing the current help in preventing transformer core saturation problems? or is a blocking capacitor necessary?

  • 0
    TI__Guru* 86286 points

    Hello,

    In regards to question 1.  The bandwidth of your current sense transformer should be greater than the switching frequency of the CS signal that you are trying to measure.

    If your design is using voltage mode control and you are using the peak current limit to protect the power stage, it will not protect against transformer saturation and the DC blocking capacitor is required.

    Most design choose to do peak current mode control to avoid transformer saturation.  This also removes the need of a DC blocking capacitor.

    Regards,

     

  • 0 in reply to Mike O'
    Prodigy 100 points

    Hello,

    Thanks for your reply.

    But to be more specific, by how much should the bandwidth of the CT higher? Is a CT of 1MHz bandwidth enough for measuring a signal of that has a frequency of hundreds of kHz?

    Kind regards

    Ahmed

  • 0 in reply to Ahmed Eldistawy
    Prodigy 100 points

    And one more question regarding the CT in the datasheet. How does the CT reset with the schematic in the datasheet?

  • 0 in reply to Ahmed Eldistawy
    TI__Guru* 86286 points

    Hello,

    I would select the CT so the gain is above the - 3dB point. Ideally you want 0 dB at the switching frequency.

    Regards,

  • 0 in reply to Mike O'
    TI__Guru* 86286 points

    Hello,

    When it comes to resetting the CT generally there is reset resistor (RRE) across the CT winding, on the secondary of the CT. The following link will bring you to an application note that describes how to setup the CT with a reset resistor so the CT does not saturate.

    https://www.ti.com/lit/an/slua560d/slua560d.pdf

    Regards,

  • 0 in reply to Mike O'
    Prodigy 100 points

    Hello,

    Thanks for your reply.

    I am planning to operate UCC28951 at 400 kHz while using a CT of 1MHz bandwidth (the highest bandwidth I can find) and I was curious whether this is correct. Because the signal comes a bit distorted on the other side of the CT because of the limited bandwidth.

    Would that cause a problem in peak current mode control?

  • 0 in reply to Ahmed Eldistawy
    TI__Guru* 86286 points

    Hello,

    This looks like a simulation.  If it is you may not be sampling enough points.  If this is a  simulation by increasing the sample points you should be able to smooth out the waveform.

    It also looks like you may have excessive filtering on the CS pin.   Generally I would set the pole of the CS filter to 10 times the switching frequency or greater.

    Regards,

  • 0 in reply to Mike O'
    Prodigy 100 points

    Hello Mike, 

    It is a simulation. I am just trying to get to the point about how far I can increase the switching frequency and at the same time implement peak current mode control. Knowing that the best CTs have a bandwidth of 1MHz which would result on its own without any filtering because of its bandwidth in a signal like the one shown in the simulation when operating the UCC28951 at switching frequencies of hundreds of kHz.

    My question simply:  Is a CT (which can have a bandwidth in the range of 1MHz at best, the CT itself will filter the signal, that is what I am afraid of) really capable of accurately sensing the current when UCC28951 is operated at frequencies in the range of hundreds of kHz?

    Kind regards

  • 0 in reply to Ahmed Eldistawy
    TI__Guru* 86286 points

    Hello,

    If you operating at 400 kHz on the primary.  The affective switching frequency would be 800 kHz.  If the graph you are showing is the CT's frequency characteristics.  It looks like you might get some filtering.  At 800 kHz the gain looks to be 1/4 with a phase shift of -40 degrees.

    I do have an idea you could use two CTs in each leg of the PSFB and diode or them to the CS pin.  If you do this each CT will be operating at 400 kHz and then you will not have filtering issues caused by the CT.

    Let me know what you think?

    Regards,

    Mike

  • 0 in reply to Mike O'
    Prodigy 100 points

    Hello Mike,

    Thank you for the idea. 

     But even if I operate the IC at 100 kHz, which means 200 kHz effective switching frequency. When using a sensor that has a bandwidth of 1 MHz there will be a phase shift. 

    That's why I am asking, I'd usually pick a sensor that has a bandwidth of 15 times or more the frequency of the signal I am sensing. But from what I see is that CTs don't exceed the range of 1 MHz which would cause a phase shift even if the switching frequency is 100 kHz. 

    Is the switching frequency used when implementing peak current mode control with a measurement that has isolation actually limited because of the sensor?

    Are there sensors that provide isolation that have a high bandwidth capable of sensing the current without distorting the signal? because I can't find any.

  • 0 in reply to Ahmed Eldistawy
    TI__Guru* 86286 points

    Hello,

    There will be no attenuation of the CT at 200 kHz according to your graph.  Does the CT really have a phase shift with no loss in gain?  I would not have thought that there would be a phase shift in the current in the CT.  You might want to double check this.

    Your other option would be to use current sense amplifiers and a shunt.  However, most of these devices have a band with of around 1 MHz.  The following link will bring you to what TI offers in current sense amplifiers.

    https://www.ti.com/amplifier-circuit/current-sense/analog-output/products.html

    Regards,