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.

In-rush Surge using TPS61170 in Soft Start Mode

Other Parts Discussed in Thread: TPS61170

I am using a TPS61170 boost converter in "Phantom Power" mode, exactly as shown in Fig. 22 of the data sheet. Vin (5V)is sourced by a USB host. The output current draw is very modest - less than 2mA @48V. Even though the circuit is in Soft Start-up mode (Ctrl hard-wired to Vin) I see a high in-rush current (>500mA) as soon as the output voltage starts ramping up. This exceeds the capacity of my powered USB hub and causes Vin to drop to <3.5V.

How can I limit the input current to<100mA?

To Summarize:

Vin = 4.5 - 5.5V

Vout = 48V

Iout <2mA

Cin = 4.7uF

Output Load ~ 33K, 4.7uF

  • Hi Donald,

    The typical start-up waveform is shown in  Figure11 in the datasheet.

    Could you attach your start-up waveform on the forum? which including CTRL, Vo, IL?  Is your input wire very long?

    BR,

    Helen

  • Hi Helen,

    Here are a few scope shots of my 48V regulator circuit.  The trace labeled VBUS is the 5V input, connected, via T4, to both VIN and CTRL of the 61170.  FB_48V is the FB input to the 61170.  L100_CURRENT is a current probe measurement taken at the output of the inductor.  48V is the output of the regulator circuit.

    As you can see, the 5V input drops from about 5.2V to about 4.5V as the 48V output ramps up.  Although not shown in this trace, we also see a surge of about 400mA on L100 and a drop of about 0.5V on VBUS every time the 48V output is muxed onto a load.

  • Hi Danold,

    I noticed the input voltage has a big drop or (oscillation) during the  start-up. This is usually caused by the long and thin input wire. Please try to use a much thicker and shorter wire, it will be helpful.

    BR,

    Helen

  • You may also want to consider some bulk electrolytic capacitance at the input for testing purposes.  Maybe 100uF.  The actual value is not critical.

  • This regulator is on a USB device.  The cable we used for this test is about 8 feet long and has a 24AWG power pair.  Per the USB specification, however, the cable can be longer and have wires as small as 28AWG, as long as the voltage drop over the length of the cable is <125mV.  How can we reduce the input  voltage droop while maintaining compatibility with the USB spec?  (Incidentally, the USB spec also limits the capacitance on the VBus pin to <10uF, so we can't just hang a big cap on the regulator input.) 

    It looks like the current surge happens right around the time the soft-start period ends.  Is there any way to prolong the soft start?


    Thanks for your ideas!


     -Donald

  • Hi Donald,

    From the spec, for the first 5 msec after the COMP voltage ramps, the current limit of the PWM switch is set to half of the normal current limit specification or below 700mA (typical). Form your experimental result, I noticed that the peak inrush current is about 600-700mA. So it is a normal phenomenon.

    From your schematic, I noticed that you already add an inrush  current limit circuit: T4+R138, so you can limit the inrush current by control the timing of the t4's turn-on time, before the output voltage reach a certain value, t4 keeps turn-off, all the current flow through r138. Make sure the power rating of R138 is OK during the start-up.

    So you need to spend some time on experiment to make a judgement on when the T4 should be turn-on and the R138 selection.

     

    BR,

    Helen

  • Hi Helen,

    You raise a very interesting idea... R138 was not intended to be used as an inrush-limiting device.  It is normally not populated and was included in our design for debugging purposes.  We could insert another FET switch / resistor combination downstream of T4, but what would be the effect of a reduced voltage on Vin and CTRL?  If Vin drops below 3V the regulator isn't guaranteed to work, but if Vin dropped, let's say, from 5V to to 4V as Vout ramped up, would that have the effect of reducing the input current, or would the regulator work harder to boost Vin to 48V, thereby drawing even more current from the 5V source?

    Regards,

     -Donald

  • Donald,

    You can connect VIN and CTRL in front of the insert switch, thus you needn't to consider the Vin drop problem. Because the control and power is seperated.

    BR,

    Helen