• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 64 subscribers
  • Views 758 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPS2492 Tina-TI

Ignacio Linares
Expert 1655 points
Other Parts Discussed in Thread: TPS2492

My customer has reported the following problem:

[Problem:
Using transient model for TPS2492 in Tina. Simulation requires that I reduce my sense resistor in half in order to work. Datasheet and spreadsheet calcualtions says 4.5 mOhm for sense resistor, but does not work. Change sense resistor to 2.25 mOhms and device works. Using Optional Startup Method.]

[Steps Needed to Recreate Problem:
Have Tina simulation file that can be sent.  Basic design parameters:  Vin=42V; Ilimit=10A; UV=10V; OV=45V.  Using TI data sheet Figure 19 on page 20, here are the component values:  R1=294k; R2=31.6k; R3=10k; R4=174k; R5=32.4k; Ct=0.15uF; R6=10k; C2=1500pF; Rsense=4.5 mOhms; Rgate=10 Ohms; Rcg=1k; Cg=0.047uF; Co=1000uF; MOSFET=Infineon IPB017N06N3G.

  • 0
    TI__Expert 7990 points

    Please explain what "does not work" in specific terms, and also explain what kind of loading you might have had, for example, output capacitance and loading. 

    Please send the simulaiton file and any calculations (such as the Excel Worksheet calculation tool) you might have used to obtain the values in the circiut. 

  • 0 in reply to Martin Patoka
    Prodigy 20 points

    The output never completely turns on, even with a load less than the expected design limit of 10A.  The output capacitance is 1000uf with a 3 Ohm load in parallel.  There is also a programmable current source on the output to simulate spikes, but it is set to zero.

    TPS2492 Transient Model IPB017N06N3G_10A__45V_Removed Extra Components.TSC
  • 0 in reply to Martin Patoka
    Prodigy 20 points

    Here is the excel spreadsheet.  It appears only 1 file can be uploaded at a time.

    90000125 24V IPB017N06N3 at 10 Amps 42V Calculations.xls
  • 0 in reply to Carl Oleksak
    TI__Expert 7990 points

    Examination of the spreadsheet shows the programmed MOSFET dissipation is 70W.  Running the sim, the system hangs up with Vout = ~5.74V & Iload ~1.91A.

    The MOSFET dissipation is Vin-Vout * Iload ~70W.  So you have a classic load-line latchup.  If you were to plot the hotswap output I/V (for example, the excel spreadsheet plot page and generate a plot based on Vout = Vin-Vds at the same currents) and add the 3Ohm resistor load line, there sould be an intersection at the point found in the sim.  If this is a sim only problem, use the PG signal in conjunciton with a switch to hold the load off during inrush.

    These source/load line intersections are possible anywhere you have a nonlinear source (in this case power foldback) or load (e.g. dc/dc converter).

    In this specific case, a converter load would often follow the hotswap.  The most typical method of avoiding the latchup is to implement "PG" from the hotswap to the dc/dc and/or implement an undervoltag lockout on the dc/dc.

    In general, inrush controlled startup with the load "running" is a difficult battle.  Most applications aviod it if at all possible.

    In the case of the gate dv/dt with a capacitor/resistive load, the expected load current is a pedestal (I = CdVout/dt) + a ramp (Vout/R).  The gate driver current in conjunction with the gate pin loading sets the reference dv/dt, and the MOSFET source follows it to create the dVout/dt (source follower).  The trick is usually to keep teh peak current less than current/power-limit curve.