• State Resolved
  • Locked Locked
  • Replies 12 replies
  • Subscribers 63 subscribers
  • Views 390 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.

LM2611: Large current draw and possible instability

Mark M.
Intellectual 255 points
Part Number: LM2611
Other Parts Discussed in Thread: ALLIGATOR

We have the following LM2611 circuit below that is powering a circuit that has a current consumption between 20mA and 80mA (verified by removing the LM2611 and using a bench supply). When we use a bench supply set to 5V and 1A max as Vin for the LM2611, the current limit is instantly hit. The circuit draws 1 A with the bench voltage sitting around 2 V (since the bench supply current limit is hit).  

Below is from the LM2611 circuit when it is is disconnected from the circuit and instead has resistors loads.

Here are the 5V bench supply currents at different loads:
no load = 0.6mA.
1kohm load = 22mA
120 ohm load = 180mA
Below are scope captures under these load conditions.
Ch1, yellow = D1 anode
Ch2, green = Vin
Ch3, blue = Vout
Ch4, red = pin 1 sw 
No Load
No load, zoomed out
1k load
120ohm load
Also, this may be a side point, but the output voltage measures around 8.9 to 9V but the resistor divider should have this set to 8.65V 
Can anyone provide some insight on why this circuit may draw excess amounts of current driving our circuit but appears to work ok with resistor loads? Is this a stability issue?
  • 0
    TI__Genius 16515 points

    Hello Mark,

    Thanks for reaching out to us via e2e.

    Is there a way you can tell how - in case of your circuit - the load looks like?

    It seems that it is not mainly resistive and somehow confuses the feedback.

    Thanks and regards,

    Harry

  • 0 in reply to Harry Hofner
    Intellectual 255 points

    Hi Harry, 

    I can characterize this load on Monday. I'll look at the start-up and steady state current. Let me know if there is anything else you would like to see. 

    Mark

  • 0 in reply to Mark M.
    TI__Genius 16515 points

    Hi Mark,

    I assume that you first connect the load and then turn on the bench supply.

    Indeed it would be interesting to see, how the system behaves (switch node, output) before the current limit and associated voltage drop comes in.

    The combination of parameters where it ends up during the current limit is pretty unlucky:
    The current is probably still below the current limit, so it does not turn off.
    VIN falls below the lower limit for proper operation, but remains above the shutdown.

    The question is: why does it get into that state?

    Thanks,
    Harry

  • 0 in reply to Harry Hofner
    Intellectual 255 points

    Excellent advice. I'm glad I asked. I'll get you this information on Monday. 

  • 0 in reply to Harry Hofner
    Intellectual 255 points

    Hi Harry,

    Here are some captures:

    Bench supply powering the load circuit (no LM2611)

    LM2611 powering the load circuit (two captures, one with current from bench supply, the other with current out of LM2611 regulator circuit)'

    Any thoughts on what may be going on? It appears there is no current going to the load circuit and that the 1A is contained within the LM2611 circuit. 

    Thanks,
    Mark

  • +1 in reply to Mark M.
    TI__Genius 16515 points

    Hi Mark,

    The Voltage of the bench supply is coming up slowly and when it has reached about -2.5V the inrush current of the load is kicking in.
    With the converter inbetween it seems that this happens even earlier.

    Below a stable 2.7V supply the LM2611 will not work properly.

    Can you please try to hold the SHDN pin low until the bench supply has reached a stable output voltage?

    Will that make it work?

    For your final application a switching transistor with a simple RC circuit in front, which charges up slowly may be good enough to control the SHDN pin.

    Best regards,
    Harry 

  • 0 in reply to Harry Hofner
    Intellectual 255 points

    Below are captures from manually (ie disconnecting an alligator clip) changing the shutdown pin state. The LM2611 does not power up. I am going to try with the RC switching circuit you suggested in case the alligator clip noise is causing an issue. I'll post those captures soon. 

    All the images below are from the same capture, just different time scales and portions of the traces.

    Disregard Vout in the images below, it was not connected. The circuit still fails based on the SW pin.   

  • 0 in reply to Mark M.
    Intellectual 255 points

    Good news. Controlling the shutdown pin with the RC circuit appears to work. The first few attempts with this circuit failed but now it appears to work. I am not sure what changed. I'll do more testing to confirm this solution is robust. 

    Also, in the previous post, the Vout probe was not connected, so please disregard that trace in those captures.

    Below are two captures with different RC constants (same trace designations as the previous post). I think the long time constant is probably best since Vin is fully stabilized before turning on the LM2611.

  • 0 in reply to Mark M.
    Intellectual 255 points

    I spoke too soon. I moved the RC circuit from the breadboard to a protoboard and now I am having startup issues again. See captures below.

    Any thoughts on the possible issue? 

  • 0 in reply to Mark M.
    TI__Genius 16515 points

    Hi Mark,

    For my understanding, can you please clarify the following::

    On your breadboard you saw the following behavior.
    - Which load was connected ?
    - Was it your target device or just a resistor ?

    On the screenshot of the signals on your protoboard I can see the following things happening:
    - When the shutdown is released, the converter starts up until a certain output voltage has been reached (marked in red).

    - In fact, that voltage level has even been exceeded a litle bit
    - Presumably the load has already decreased the current consumption at that point of time, so the output capacitor gets charged too much.

    - The converter cannot just reduce the duty cycle, otherwise the capacitor would get charged even more.
            So it stops completely, waiting for the output voltage to come down to the desired level.

    - As the load has reduced the current consumption to 20 mA, it takes a while until the Voltage has dropped enough.
    - When the output voltage is back at the desired level, the converter starts again to keep it stable (marked in blue).

    ->  What exactly is the problem?

    Is it correct that the output voltage does not reach the desired level (-8.66V)?
    Or are you using a 10:1 probe here?

    If the output voltage is too low, please check the voltage divider / capacitor on the NFB pin

    Which output voltage do you see with a resistor as a load ?Is there a difference between the resisor and your tsrget device?

    If you are just talking about the gap, where the converter stops (between the red and the blue marked areas):

    Are you using the exact same components on both boards?

    Is the output capacitor the same (capacitance, ESR)?

    Please let me know

    As per your initial side note: 
    > the output voltage measures around 8.9 to 9V but the resistor divider should have this set to 8.65V

    There is always a difference between the calculated values and the reality (tolerance/spreading of component values, influence of the board, placement of the parts etc.). Three percent are not too far off.

    Best regards,
    Harry 

  • 0 in reply to Harry Hofner
    Intellectual 255 points

    Hi Harry,

    It appears that this circuit is now working. 

    I double checked my setup and made new measurements based on your questions. The Vout probe reading was correct. Vout was around -1.5V, even with no load.  The NFB pin measured around -1.2V which does not make sense based on the voltage divider.

    I replaced the LM2611 and both voltage divider resistors and now everything appears to work as expected. See captures below with the LM2611 driving the target device. I suspect that the circuit had been damaged from all the testing where the LM2611 would not fully turn on and consumed a lot of current. 

    Answer your questions in case they are helpful:

    On your breadboard you saw the following behavior.
    - Which load was connected ? - Target device
    - Was it your target device or just a resistor ? Target device only

    What exactly is the problem? Is it correct that the output voltage does not reach the desired level (-8.66V)? - Correct, voltage does not reach desired level
    Or are you using a 10:1 probe here? Yes, was using 10:1 probe and scope had correct settings. The circuit was outputting -1.5V

    Which output voltage do you see with a resistor as a load ?Is there a difference between the resistor and your target device? - I tested the circuit today (before replacing the components) with no load and it still had an output voltage of -1.5V.

    Are you using the exact same components on both boards? Is the output capacitor the same (capacitance, ESR)? - Yes, all components were identical. 

    Thanks for all the help,

    Mark

  • 0 in reply to Mark M.
    TI__Genius 16515 points

    Hi Mark,

    Good to hear that it is working now.
    And thank you for setting the thread to "Resolved" and "Closed" and even answering the last questions.
    I am glad I could help.

    Regards,
    Harry