• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 2 answers
  • Subscribers 62 subscribers
  • Views 129 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.

LM7480-Q1: Turning on and off again with blip on Vout

Iain Milnes
Prodigy 100 points
Part Number: LM7480-Q1
Other Parts Discussed in Thread: LM7480,

I am using the EVM and monitoring Vin, Vout, and Dgate. The load at the time of the test is 2 A.

Here are two examples when I create a blip on the Vout. All signals are 5 V per division and the time is 10 ms per division.

The signals are:

  • Yellow: Vout
  • Blue: Vin
  • Purple: Dgate

The blip on the 24 V is caused by using an extra power supply fed through diodes connected to Vout and controlled by a momentary switch. The blip caused by my secondary power supply makes the voltage go to about +27 V for about 20 ms.

The gate to the MOSFET is turned off immediately and raised when the blip disappears. This is good; the Vin power supply rises slightly, but not enough to kill the primary power supply connected to Vin.

But here is another example:

In this case, Dgate is immediately taken low, but then Dgate is driven hard on. It then takes the LM7480 a full 3 ms to realize it should be doing something and drives the gate off.

Fortunately, although the primary power supply (at Vin) got a bigger blip, it was insufficient to kill it.

Can you suggest how the circuit can be modified to avoid this long delay in turning off the controlling MOSFET (Q2)?

  • 0
    TI__Guru 57737 points

    HI Iain,

    In case 2, when the Vout starts to go above Vin, the DGATE is immediately turned OFF but we can see that it is turned ON once again. This could be because the Vout may have momentarily  gone below Vin forcing DGATE to turn ON again.

    Adding decoupling caps close to A and C pins  to GND should filter out any noise on these pins and should help mitigate these DGATE switching. 

  • 0 in reply to Praveen GD
    Prodigy 100 points

    Praveen,

    I am not entirely convinced because I can see little physical reason for the voltage to dip. I'm using a pushbutton switch to connect 27 V to the output. Here is a better capture of this, which seems to be reproducible about one time in 20 (5%):

    The signals are:

    • Yellow: Vout
    • Blue: Vin
    • Purple: Dgate

    If you zoom in, you can see a blip of noise on Vout but it has to be <10 µs. But it then takes the LM7480 a further 20 ms before it turns off again, despite the Vout continuing to rise. The switching characteristics of the IC are listed in µs, but here we have a hysteresis (Vgate on to Vgate off again) of 20 ms!

    These tests are done with the EVM. Your suggestion about placing decoupling capacitors close to the output is noted, but C5 and C8 are pretty close to Vout on the EVM and it might be hard to improve on that:

    I guess we have to live with this because our power supply doesn't shut down (as it would without the LM7480 in circuit), but it sure looks like a bug to me!

  • 0 in reply to Iain Milnes
    TI__Guru 57737 points

    Hi Iain,

    There is no Hysteresis (Vgate on to Vgate off again) of 20 ms in the IC. You can see the AC super imposed test behavior of LM7480-Q1 at 30 kHz below. The DGATE turns ON and OFF every 33us. 

    LM74810-Q1 which is very similar to LM7480-Q1 can be seen switching at even 200 kHz below.

     

    From the above waveforms you can see that the switching delay is indeed in 'us' range and not in 'ms' range. 

    In your case, you are measuring the input voltage and output voltage with single ended probes. These probes generally higher error (offset, gain) > 10's of mV. Since the reverse voltage threshold (V(AC_REV)) is only few mV's it would be hard to get accurate information of Vin-Vout using single ended probes. To get a clear picture on what is happening, please probe input current. From input current we can calculate the voltage drop across the FET (using Rdson) and check if there is enough reverse voltage to trigger the turn OFF of DGATE.