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.

TPS70933DBVR problem

Other Parts Discussed in Thread: TPS709, TPS7A4101

Hello,

I recently designed a circuit which is powered by small 3.3V fixed voltage regulator (TPS70933DBVR) and draws only 9mA. 10 units have been built as part of engineering evaluation.

In front of TPS70933DBVR, I have a diode CRS09(TE85L,Q), followed by two shunt capacitors (100nF and 10nF) and TPS70933DBVR IC.

On its output, there are 3x100nF capacitors, 1x10nF and 1x470uF capacitor followed by two small ICs.

According to the datasheet, TPS70933DBVR should be able to withstand input voltages up to 30V.

I run into a problem when part of my design requirement called for input voltage test between 4V and 28V.

I first tested the circuit at 12V & 4V without any surprise. I then increased the input voltage to 28V and TPS70933DBVR failed almost immediately (dropped output current to approximately .4V and started sinking over 400mA on my current limited analog PS).

I replaced the defective IC and then tried again, the circuit failed again after a few seconds and a few brief connect/disconnect power cycles.

I powered second, third, and forth unit and experienced an identical problem; the circuit powers up fine at first but if you briefly power cycle the unit, it fails immediately (starts drawing large input current).

I then removed the large 470uF capacitor on the output hoping that would solve the problem (as there may have been large current flowing from TPS70933DBVR output to input). The problem is still present.

Last experiment I’d performed was I gradually increased voltage from 12V (1V at the time) and power cycled the circuit and everything worked fine until I reached 20V at which point TPS70933DBVR failed again.

Can you please explain what may be causing this issue?

Is this IC supposed to operate up to 18V?

Should I dump this voltage regulator and pick a different supplier?

 

 

 

  • Hi Veli,

    TPS709 has an operating input voltage range up to 30 V. Are you able to provide any scope shots to help with debugging?

    Very Respectfully,
    Ryan
  • Hi Ryan,

    Thank you very much for such a fast reply.

    I am aware that the chip has been designed for up to 30V use; this is why i picked it for my new design.

    During my troubleshooting, i "blew up" 7 out of 10 units (TPS70933DBVR regulators) by powering these units with voltage levels 20-28V while 18V and below works just fine. 

    Unfortunately, I don't believe i can capture anything useful just before the regulator fails as this is basically a manual quick connect/disconnect from a power supply (GW Instek, GPS-3303).

    Instead, i may be able to send you one of my test boards in a few weeks time upon completion of my engineering evaluation.

  • Hi Veli,

    There are a couple quick things that we are able to check with a detailed schematic. For example, how are you utilizing the enable pin?

    We also need scope shots in order to help support. A scope shot of Vin, Vout, and Ven (if you are using a second supply to drive EN) during startup would go a long way in helping us debug.

    Very Respectfully,
    Ryan
  • Ryan,

    The enable pin on the chip is left unconnected per data sheet suggestion.

    At this time, I have just a few working units left but i did order 25 more pieces of TPS70933DBVR which are due to come in on Friday this week.

    At that point, I my be able to do some more experiments; I will let you know how it went and perhaps email you a simplified schematics and a few screenshots.

    Regards

  • Hi Ryan/Veli
    Me too face the same issue.
    I pulled down the EN to Ground using 10K and a 2V is applied with a voltage divider network when required.
    Same thing happened to me too. After few iterations, it failed. For Scope output, nothing is special, if working, it is perfect and if not nothing. I am not able to copy the waveforms during the issue as I cant expect when it can happen.

  • Hello all,

    According to the datasheet, ENABLE input is a low voltage input and this pin can be left floating if the chip is to remain enabled at all times. Input voltage specification in my design is 5V to 28V (the chip is designed to operate up to 30V) and current consumption of my load is 10mA.

    Nevertheless, if I exceed 18V (i.e. 20V) and repeatedly connect/disconnect power supply to the regulator, the chip (regulator) catastrophically fails (sometimes it fails open and draws no DC current from the source while sometimes fails short and draws huge current such as a few hundred mili-amps from DC source).

    My first guess was that I had too much capacitance on the output and somehow the capacitance was causing the chip to fail so i removed all large capacitance from the output (all that was left was about .2uF) and the regulator still catastrophically fails at 20V and above. I tried a few other ideas and at the end of my experiments blew up about 10 separate regulators.

    At that point I realized that it was too risky to move forward with this chip in my new design as this would eventually cause our medical product in the field to fail and lots of headaches to the company.

    I picked MIC5233-3.3YM5 TR instead. This new chip has absolutely no such problems (i.e. it operates just as expected at any input voltage between 5V and 28V) while sourcing 10mA in my design.
  • Hi Jamshid, Veli,

    When you say you are connecting and disconnecting your power supply, are you hot-plugging the supply? Hot-plugging is where you supply is turned on and you then connect your power supply to the LDO. This event can cause large voltage spikes at the input of the TPS709 that could exceed the abs. max of 32V, where the Micrel device has an abs. max of 38V which could be why it survives. If this is the case we do have a few other LDOs such as the TPS7A1633 and TPS7A4101 with higher input voltage ratings.

    Regards,
    David
  • Yes, I hot plug the device and it fails with as little as 20VDC on its input regardless of regulator output capacitance. If on the other hand you gradually increase input voltage from 0 to 28V, the device operates fine. Last scenario however is very unlikely as normally the power to a voltage regulator is always applied suddenly up to its rated voltage input.

    Given all this, it appears that this voltage regulator has been over-rated in the datasheet; correct operating input voltage should be stated as "up to 18V" and not 32V because the device works fine up to 18V.
  • Hi Veli,

    As I stated earlier hot-plugging can cause large voltage transients on the input above the 32V abs. max. The reason is that the wires from the power supply to the LDO are pretty inductive. When you hot-plug the charge stored in the wire gets quickly discharge into the capacitance on the board, causing a large voltage spike.
    The below app note explains hot-plugging and its effects on the actual input voltage:

    www.ti.com/.../slua446.pdf

    The scope shot below, from the app note, shows how large the spike can get (up to a 40V peak for their setup, channel 3 is the input voltage on the board);

    As such if its a requirement that you survive hot-plugging at 28V, I would advise a part with a significantly higher voltage rating such as the ones previously mentioned.

    Regards,

    David

  • Notice that the voltage on pin EN should be less than 6.5V.
  • Exactly David,

    Me too find the reason was same. My PCB track was a long one and because of Hot insertion, the inductive effect of the track is causing the Voltage to increase.

    I introduced a higher value capacitor 100uF along with a TVS diode at the incoming point in parallell with the supply. Now it is working perfectly.