LM74900-Q1: Tolerance for the Imon:Ilim ratio for corner case calculations

Part Number: LM74900-Q1

Tool/software:

Hi,

The source question and answer confirmed my assumptions as to the source of magic numbers in the VIMON and RILIM calculations.
It also states that ISET is apportioned 9:1 Imon : Ilim

I require the tolerance of this scaling factor for corner case calculations.

On first glance, it looks like it is provided in the tabulated parameters on page 7 as VGE_SET; which looks like it provides bounds on the figure of 90 used in equation 7 (p22)

(the factor of 100 arises from the specified RIMON and RSET values, 5k/50 ).

My assumption does not look to be correct.
When applied to the ILIM equation, the scale factor varies 18/100 ... 10/100 ... 3/100 and provides a commensurate wide range of limit currents;
In my specific case the OCP nominal current of 18.3A swings to 4.52A and 38.22A (Min/Max corners of all resistors, VGE_SET and ILIM_VSNS_TH).

This is so broad as to be unusable, and well outside of the simple bounds calculated by the TI LM749x0-Q1 Excel design tool;
It seems as though my assumption is incorrect or there is some design bound or paired deviation in play that makes the worst case corner selection impoissible in practice.

Can you please provide guidance on my use of the VGE_SET parameter for determining the current division tolerance, and how to calculate the corner cases of the OCP and current monitor paths.

Kind regards,

George Tzanatos.

  • Hi George,

    Please use the following spec. for corner values of OCP comparator.

    Hence, 12 in the equation can be replaced with 10.8/13.2 for min and max OCP values respectively.

    Regards,

    Shiven Dhir

  • Hi Shiven,

    thanks for the response.

    As it stands, it does not answer my question. I am checking corner cases, and not just typical operation on a bench at 25C.

    The input to the comparator is from the current shunt opamp. The op-amp output is divided into two paths.

    At a first order of analysis, there is an error in the gain of the op-amp and an error in the current split ratio. These will not be perfect, as noted in the VGE_SET parameter for the IMON path. The opamp input offset error is documented.

    I assumed that the VGE_SET parameter covered the gain andcurrent split errors for the IMON path as the numbers involved match the ratio provided in the source message. This led to my apportioning of VGE_SET to the ILIM path and the very wide deviation observed. This assumption left the question of gain error open.

    Are you inplying that the entirety of the the errors in the ILIM path (>15% from memory on the Imon path - opamp gain, current split and comparator threshold voltage) have been rolled into the comparator thresholds?

    Thanks for your support,

    George Tzanatos.

  • Hi Shiven,

    Can you please estimate when a response will be provided?

    Thanks for your support,

    George Tzanatos.

  • Hi George,

    I will get back to you by Wednesday EOD. Apologies for delay.

    Regards,

    Shiven Dhir

  • Hi Shiven,

    Can you please estimate when a response will be provided?

    Thanks for your support,

    George Tzanatos.

  • Hi George,

    It is true that if error exists in IMON path, the remaining will be present in ILIM path. 

    Since the current limit increases due to error (38A), short circuit protection will still be present and keep the system safe. Short circuit protection is not referred to current split path and has no effect due to error.

    For lower side(4A), you can scale the system and resistor value according to worst case condition.

    Although, these are very corner values, and I have not seen in systems yet. You can still scale according to the equation stated below.

    Regards.

    Shiven Dhir