• State Resolved
  • Locked Locked
  • Replies 19 replies
  • Subscribers 64 subscribers
  • Views 2764 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.

LM22674MRX-ADJ - Some questions

Hirotaka Matsumoto
Guru 27585 points
Other Parts Discussed in Thread: LM22674, LM2267X-22680DESIGN-CALC, TINA-TI

Hi all,

Would you mind if we ask LM22674MRX-ADJ?

<Confirmation1>
When we use WEBENCH with LM22674MRX-ADJ(LM22674), it only shows output voltage which MAX range is 5.0V. 
Our customer would like to use output voltage 12V or 10V, so do we have to use LM2267X-22680DESIGN-CALC
for caluculation, don't we?


<Confirmation2>
Condition: Input DC24V, output DC12V, Max current 0.3A 

We sumilated the schematic of LM22674MRX-ADJ using LM2267X-22680DESIGN-CALC as following.
1300.20140930_ LM22674_DC24V-OUT12V calculator.xls
And then, there is no Bode plot(because of simple switcher).
So we could judge whether phase margin is enough or not, if you know it, could you let us know? 


<Confirmation3>
Condition: Input DC12V, output DC10V, Max current 0.3A 
8561.20140930_ LM22674_DC12V-OUT10V calculator.xls
Output will be no stability.
About "the maximum duty cycle before the output voltage will "dropout" of regulation" P8 on the datasheet,
we got the Vin(min)=11.21V. So, compared with Vin=DC12V, there is no margin.
Is our recognition correct?


<Confirmation4>
Condition: Input DC24V, output DC-12V(negative voltage), Max current 0.3A 
After LM22674 operate,  the ground leve of LM22674 will be -12V.
3348.20141001_ LM22674_DC24V-OUT-12V schematic.pdf
When LM22674 turns off, is EN pin only connected to GND, right?

We need your help.

Kind regards,

Hirotaka Matsumoto

  • 0
    Guru 27585 points

    Hi all,

    Could you please let us know your circumstance regarding this question?

    We would be grateful if you could reply as soon as possible.

    We need your help.

    Kind regards,

    Hirotaka Matsumoto

  • 0 in reply to Hirotaka Matsumoto
    TI__Guru 66575 points

    I will pass this to the Webench guys.

    FD

  • 0 in reply to Frank De Stasi
    TI__Intellectual 2930 points

    Hello,

    To answer the WEBENCH question, WEBENCH is going to use the 5.0 version of this device to do designs over 5V as recommended by the datasheet.

    If you need help verifying the stability of an ADJ version, send me the schematic and I'll run a simulation for you

    Best regards,

    Tommy

  • 0 in reply to Tommy Jewell - WEBENCH Design Center
    Guru 27585 points

    Tommy san,

    Thank you for your reply. We add some schematics.

    We'd like to confirm the stability followings;
    5482.20141002_ LM22674 Schematic.pdf

    1. In DC24V - Out DC12V 
    2. In DC12V - Out DC10V 
    3. In DC24V - Out -DC12V(negative vol) 
    4. In DC12V - Out -DC10V(negative vol) 

    So LM22674 is "simple switcher product", we guess that we cann't adjust phase compensation(stability).

    Kind regards,

    Hirotaka Matsumoto

  • 0 in reply to Hirotaka Matsumoto
    TI__Guru 66575 points

    Regarding the schematic of the negative output regulator.  I am not sure what the purpose of R104 and R111 ??

    If you do not need the EN function you should be able to leave it float.  If you want to pull it up with those resistors, I would increase the values to lower the current into the EN pin.  The EN pin has a 6V Zener connected from EN to the GND of the device; so you must limit the current.  To turn the device off, the voltage between EN and the GND of the device must be brought to <=1.6V.

    FD

  • 0 in reply to Frank De Stasi
    Guru 27585 points

    Frank san,

    Thank you for your reply and your comment!
    First of all, we apologize that our schematic made you confuse.
    We mistaked to attach the file.

    As you mentioned, we suggested that customer use following schematic, when the customer does't use UVLO function.
    1513.20141003_LM22674 Schematic.pdf

    -In case of Enable: EN pin floats
    -In case of Disable: EN pin connects to GND or <=1.6V

    And then, if LM22674 starts to output the negative voltage, the GND level of LM22674 will be -12V.
    So, if our customer would like to use ULVO function, it needs to be calculated as 36V at EN pin.

    Kind regards,

    Hirotaka Matsumoto

  • 0 in reply to Hirotaka Matsumoto
    TI__Guru 66575 points

    OK; thank you

    FD

     

  • 0 in reply to Frank De Stasi
    Guru 27585 points

    Tommy san,

    We change some values about the schematic.
    (In case of Input 12V and Output 10V, according to calucation method(4) on the P8 datasheet,  
     we got that Input voltage must be over DC11.7V.)

    3022.20141006_LM22674 Schematic.pptx

    Kind regrds,

    Hirotaka Matsumoto

  • 0 in reply to Hirotaka Matsumoto
    Guru 27585 points

    Hello Tommy san,

    Could you please let us know your circumstance regarding this question?
    We guess that you are busy day.
    However, we would be grateful if you could reply as soon as possible.

    We need your help.

    Kind regards,

    Hirotaka Matsumoto

     

  • 0 in reply to Hirotaka Matsumoto
    TI__Intellectual 2930 points

    Hi,

    I just realized that there is a Bode plot in the xls sheets you attached, its in the bottom left corner. And is titled "Complete Loop Transfer Function".

    I ran a sim with your schematics and am getting the same as what the excel sheet shows, which is that this design doesn't have any phase margin. The big cap and inductor make your resonant peak right at crossover and so you lose a lot of phase.

    Unless you need the bulk cap for big transients from whatever you are powering, you might be better off without it.

    Best regards,
    Tommy

  • 0 in reply to Tommy Jewell - WEBENCH Design Center
    Guru 27585 points

    Tommy san,

    Thank you for your reply!
    Is this result in case of Input DC24V output DC12V?

    And then, if we change Cout=100uF, Inductor=150uH, do we get much phase margin?
    (The datasheet shows that it is recommended to use 100uF over cap for Cout.)
    Or, if you have some recommendation value about Cout Inductor, please let know us.


    Kind regrds,

    Hirotaka Matsumoto

  • 0 in reply to Hirotaka Matsumoto
    TI__Intellectual 2930 points

    Hirotaka san,

    Yes that was the Input 24V output 12V. The Input 12V output 10V was very close to the same.
    I do not currently have a model for the inverting configuration.

    In this case I would go against the data sheet. The quick start xls notes that good starting values for Cout for the LM22674 are between 2.2uF and 47uF. Values in this range will give you a much better Bode Plot.

    At your current 2.2uF is plenty to cover a steady state ripple of below  .5% of your output voltage. So your cap will be decided by the transient spec.
    20uF will completely cover 4 switching periods of full load current within a 1% spec of your output voltage. Depending on what you need values around here seem better.

    Do you have access to Pspice? There is a PSpice model for this on the product folder that you could use to simulate your schematic. Both for Loop and transient.

    Why do you want to use the ADJ option instead of the 5.0 option? The 5.0 Option is better for the higher voltages and supported in WEBENCH.

    Best regards
    Tommy

  • 0 in reply to Tommy Jewell - WEBENCH Design Center
    Guru 27585 points

    Tommy san,

    Thank you for your prompt reply and update.
    And, I am sorry if I cause any inconveniences.

    I don't have access to Pspice.
    I only use Tina-Ti, but there is no TINA-TI Spice Model about LM22674-ADJ on the site.
    If you have it, could you send me?

    Our customer needs output over DC10V using LM22674-ADJ.
    The 5.0 Option is output DC5.0V only.

    Kind regards,

    Hirtoaka Matsumoto

  • 0 in reply to Hirotaka Matsumoto
    TI__Intellectual 2930 points

    Hirotaka san,

    I know this seems counter intuitive, but actually the 5.0 Option can do more than 5V. 

    5V is the regulated voltage at the Feedback pin, but if you put a resistor divider between Output and FB you can increase the regulated Voltage at the Output (see Pic below)

    This is actually preferred over the ADJ option because the internal compensation parameters for the 5.0 Version are better for higher voltages. 

    Best regards,
    Tommy

  • 0 in reply to Tommy Jewell - WEBENCH Design Center
    Guru 27585 points

    Tommy san,

    Thank you for your reply.
    We got your saying.

    And then, if we can confirm using Webench as you showed us, we calculate our side.
    Please let us know

    Kind regards

    Hirotaka Matsumoto

  • 0 in reply to Hirotaka Matsumoto
    TI__Intellectual 2930 points

    Hirotaka san,

    Yes. That is correct. You can use WEBENCH for this and see simulations.

    Also there is TINA-TI for the 5.0 Version

    Best regards,
    Tommy

  • 0 in reply to Tommy Jewell - WEBENCH Design Center
    Guru 27585 points

    Hello Tommy san,

    As the result, because our customer has had experience to use -ADJ version over 40k amount.
    And then, our customer decided to use -ADJ version, so we'd like to confirm one point.

    In case of Vin=12V, Vout9V, how much of capacitance value should they use?
    From 2.2uF to 47uF as same as Vin=24V, Vout=12V?

    Kind regards,

    Hirotaka Matsumoto

  • 0 in reply to Hirotaka Matsumoto
    TI__Intellectual 2930 points

    Hi Hirotaka san,

    Yes it should still be in the 2.2u-47uF range. 

    It may vary from the 24V design though depending on what you chose for the inductance. Since Vin is lower  than 24V they could use a smaller inductor. Use the xls to make sure the capacitance picked is good with whatever inductor they pick.

    Best regards,
    Tommy

  • 0 in reply to Tommy Jewell - WEBENCH Design Center
    Guru 27585 points

    Tommy san,

    Thank you for your comment!

    Kind regards,

    Hirotaka Matsumoto