• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 61 subscribers
  • Views 682 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LM26003 Frequency Foldback behavior

Masatoshi Oguri
Expert 4780 points
Other Parts Discussed in Thread: LM26003, LM26001

Hello,

One of my customer found the below behavior with LM26003 at light load. I guess the Frequency Foldback is wroking here. However, there are a lot of things unexplained. Please make an answer to each question below, or if my guess is totally wrong, please give me some insight. 

In the below figure, an experiment of VFB pulled to GND is shown. Yellow shows Vout while red shows VFB. Vin is 13.5V, Vout is configured 5.0V, and the load is 1kohm resistor.

At the left end in the figure the device is in stable operation at Vout = 5.0V and Vfb = 1.236V.

When VFB is pulled to GND at the first grid, Vout changes from 5.0V to somewhere above 6.0V. 

Q1: Do you think the Frequency Foldback limits the Vout at around 6V?

Then, some noise injected to VFB triggers Frequency Foldback to discontinue making Vout close to Vin.

Q2: Does an increase in Vfb release Frequency Foldback?

Another strange behavior not shown in the figure is that Vout settles to 0.0V long after pulling VFB to GND. You should see in the figure the descending Vout. It eventually goes to GND after a long while.

Q3: Do you think this is a valid behavior?

Thanks.

Regards,

Oguri (TIJ automotive FAE)

  • 0
    TI__Intellectual 1775 points
    Hi Oguri-san,

    It is difficult to interpret the behavior in this figure. VFB is pulled low then is allowed to go above the reference voltage before being pulled low again. The frequency will be severely fold-back if the feedback voltage is held low but the switching period should not be as low as 12ms. Is the load constant? If the load changes, the low frequency could prevent the output from rising until the load is reduced below a certain level.

    Increasing The VFB above 0.9V should move the part out of frequency fold back. If something is holding COMP low, that would also result in cycles being skipped.

    One thing to note is that you are likely pushing VBIAS (and maybe PGOOD) above its Abs Max value during this experiment although I do not think it is affecting the waveform here.

    Best Regards,
    Florent Boico
  • 0 in reply to Florent Boico
    TI__Expert 4780 points

    Hi Florent,

    Thanks.

    The load is fixed at 1kohm. Thus, the load current is 5mA.

    Can the Frequency Foldback be triggered without the current limit being active?

    Regards,

    Oguri

  • 0 in reply to Masatoshi Oguri
    TI__Intellectual 1775 points
    Hi Oguri-san

    According to the block diagram, it is triggered as soon as FB goes below 0.9V, regardless of the current. Another form of frequency foldback happens when the part is in sleep mode (COMP voltage is low). I cannot fully explain the behavior you are reporting from the scope shot and I don't have an LM26003 around but I can verify the behavior of the LM26001, which is a similar part, in the lab. I'll let you know what I find out.

    Best Regards,
    Florent