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.

INA199: INA199x2

Part Number: INA199
Other Parts Discussed in Thread: TPD4S012, INA210, INA181

I am using the INA199x2 in a phone charging application. I don't have any external components except for a .045 ohm current sensing resistor. The part has a V+ of 5V and I am using a high side arrangement where the phone is the load. The reference is tied to ground. The output goes to an ADC. That's basically it. What I am seeing is that occasionally the output stays at 2V-3V which basically indicates that the chip has been destroyed. I have surge protectors on the inputs to prevent ESD problems. The voltages presented to the Vin+ and Vin- are approximately 5.1V - 5.3V. The maximum current that goes through the sense resistor is 1.5A but typically less than 1A. I don't think I am violating any of the specifications so I am confused as to why the chip is being destroyed. I realize that at 1.5A, the output would be maxed out (greater than V+) but I don't think that should be a problem. Any one have any thoughts on this?

Thanks,

CT

  • Hey Courtland,

    I am sorry you are seeing this. I have some questions.

    1. Once output drops to 2V to 3V, does the device always output the voltage after system is reset? Basically is the damage permanent and the output of the damaged device always remains at 2V-3V regardless of input?

    2. Are you measuring the input of INA199 (across 45mOhm shunt or directly measuring the current) when the output drops to 2V-3v?

    3. How are you powering the device? Do you have a decoupling capacitor connecting V+ and GND pins of INA199?

    4. Can you elaborate or share schematic of the surge protectors? Are these on both inputs of the INA199?

    5. How fast are you running the ADC? Do you have a "charge bucket" RC filter between INA199 output and ADC input?

    Sincerely,
    Peter Iliya
    Current Sensing Applications
  • Peter, Thanks for the timely response. See responses to your questions below; The circuit being used is also shown

    1) The damage is permanent. J3A is the supply for the phone. The phone plugs into J4A. If there is no phone connected the PORT1 output should be 0V. However, I am seeing at least 1V whenever the the supply for the phone is connected. If the supply isn't connected, the output is 0V. It appears that the Vin- input is presenting some load causing current to flow thus causing the effect.

    2) The voltage drop is there across R17 regardless of whether the phone is connected.

    3) The device is powered with +5V.  C51 is the decoupling capacitor

    4) Schematic attached

    5) The ADC is running at approximately 3 MHz. No RC filter used.

  • Hey Courtland,

    Thanks for the response.

    You say there is a voltage drop across R17 regardless if phone is connected. What is the value of this? Is it 10mV, which causes INA199 output to have 1V? Are you saying this is an indication of failure? I ask about measuring voltage across R17, because if output is in between 2V to 3V, but VR17 is not Vout/gain, then we can determine actual failure of the INA199.

    How often are you seeing the failures? Do you have any indication on what action results in the failure? Does it happen after system running in normal operation or when phone/load is inserted or removed?

    Sincerely,
    Peter Iliya
    Current Sensing
  • Peter, I measure .0031V across the resistor. That number should be .0001 or less when there is no load. The output is setting at 0.7V. This value increases if I connect a load and then disconnect it. Over time, while power is removed it drops but never less that 0.7V. The Vo reading isn't correct because with .0033V I should see .3V out.

    I haven't done enough testing to determine what the failure rate is. I just delivered two boards to my customer so he will be testing quite a bit and provide feedback.

    I haven't been able to determine exactly when the failure occurs. All I can say at this time is that when the phone is disconnected, the Vo level remains at some value greater than 0.

    Based on the datasheet, the phone supply voltage can be greater the the supply voltage of the INA199. Is this correct because the phone supply voltage sits at about 5.2V and the INA199 V+ is 5V. Also, I wonder if the current requirement of the phone being charged can be too high causing Vo to try to go beyond 5V. Would that be a problem for the chip?

    Thanks,

    CT
  • Courtland,

    So you see VOUT = 0.7V and VIN = 3.1mV with no load. Once you connect phone, does VOUT become a correct value? Or can it become 2V to 3V and thus this does not represent actual current? These seem like separate problems. If you can measure VIN (VR17) when load is connected and INA199 is not driving the correct VOUT voltage that would be helpful, but I understand this would require a differential probe.

    Anyway, the phone's voltage at 5.2V should not affect the I NA199 powered at 5V or any voltage supply. The INA199 can accept a common-mode voltage (VCM) of -0.3V to 26V and the VCM seen by INA199 will be dictated by your charging device at J3A.

    I do have a couple debug suggestions. Try adding a 10kOhm load resistor from INA199 OUT to GND. This could help pull VOUT below 0.7V. Additionally, try adding an RC filter on the output or charge bucket. The values of the R and C are usually recommended in the ADC datasheet. It could be that your ADC sampling to quickly compared to bandwith of INA199.

    Another thing to note is that the ESD and transient protection diodes in your surge protector really are only helpful for INA199 if some curret-limiting resistors are included at inputs of INA199 (IN-, IN+). Resistance will limit current into IN+ and IN- pin to under 5mA. I'm not sure if your part is being damaged in this way or damaged at all. Unfortunately incorporating these current-limiting resistors will make the positioning of the diodes/TPD45012 useless for INA199. Please see this design on transient robustness for current shunt monitors for more information.
    www.ti.com/.../TIDA-00302

    If you become certain that ESD event are responsible for damaging the INA199A2 and do not want to add transient protection circuitry, then you could try the INA199B and INA199C versions which have more robust internal ESD protection circuitry.

    Best regards,
    Peter Iliya
    Current Sensing Applications

  • Peter, if I connect the load, Vo does not become the correct value. In fact it is added to the 0.7V. It appears that 0.7V is the new level with no load connected. So if a load was connected and Vo was supposed to 1v, it now would now be 1.7V. I have measured Vin(VR17) to be .040V when I connect the phone to be charged. Normally I would expect Vo to be 4V but I read 4.7V.

    I don't think adding a 10Kohm resistor on the output will tell us anything as the problem appears to be that the Vin- input impedance has dropped to a point where it appears that the circuit now has a phone(load) connected thereby causing current to be drawn. I don't think a load on the output will change this. However, i can try it.

    The ESD protection circuitry is what is recommended for USB interfaces. No resistors needed.

    I just performed an impedance measurement on pins 4 & 5 with respect to ground and found the impedance to be 200 ohms on the bad chip and over 1M ohm on the good chips. I think this verifies that the chip is bad. Now I have to figure out what caused the failure.
  • Peter, do you think it is possible that I could be destroying the chip with too high of a current being drawn by the phone thereby causing the Vo to try and output a voltage much higher the V+(5V) even though the output will by clamped to V+?
  • Courtland,

     

    Thanks for measuring the input impedance. 200 Ohm input impedance definitely indicates damage to the device.

     

    If the phone at some point draws excessive current causing INA199 Vo to saturate at power supply level (5V), then this would be caused by excessive input differential voltage, yes. The highest this input differential voltage (Vin) can go without breaking the part is + or - 26V (see Absolute Maximum Ratings, Table 7.1 in datasheet). Thus, in order for this voltage to be generated, the current sunk by the phone would have to be 26V/45mOhm = 577.7A. Thus I do not think this is the root cause.

     

    The damaged input impedance (Zin) indicates a root cause of ESD or EOS (electrical overstress). While this does not seem like it should be the case since you have an ESD protection IC (TPD4S012), I still think ESD is possible after looking through datasheet. One critical aspect to note about INA199 is that it’s absolute minimum input common mode voltage (VCM) is -0.3V (see Table 7.1). If VCM goes below -0.3V then the internal ESD diode to ground will forward bias, conduct excessive current, and permanently damage. This results in a low-impedance path to ground since ESD cell will become small resistance.

     

    While the TPD4S012 (and the ESD cells of INA199) protect against transients of + or – thousands of volts (see Table 7.2 of INA199 datasheet), the protection is not the same in both directions. If you look at the TPD4S012, you will see that inside are unidirectional TVS diodes for ID, D+, and D- pins.


    You will also see that the breakdown voltage (VBR) for the Zener diodes are 6 to 9V and the forward voltage (VF) for the diodes are 0.6V to 0.95V. So if there is a positive voltage transient at INA199 inputs, the Zener diode will go into reverse breakdown to dump energy and the diode on same line as Zener will forward bias. The other diode to ground will remain in off.

     

    If there is a negative voltage transient, then ID pin’s diode to ground will forward bias and dump the energy, but it will only do so after negative voltage has become greater than diode VF (0.6V to 0.95V). Thus, there will be a time INA199 ESD diodes being stressed and this will be when -0.3V > VCM > -0.95V. If you look at Figures 8 in TPD4S012 datasheet: “IEC Clamping Waveform, -8kV Contact, D+, 25 ns/div”, you see that even after diode has clamped the large ESD signal there is 225ns of the voltage hovering just below 0V.


     

    I am not saying that 225 ns is enough time to damage the ESD cell, but if this event occurs repeatedly it will wear down INA199 ESD cell causing it to become internal short. Just because the TPD4S012 will survive these ESD events does not mean the INA199 will, especially if the absolute minimum VCM is -0.3V. Protecting devices from EOS events of less than -0.3V can be tricky. You can read more about way to protect against this and also about limiting current into input pins at our following FAQ pages:

    https://e2e.ti.com/support/amplifiers/current-shunt-monitors/w/faq/3610.dealing-with-device-protection-when-you-cant-exceed-0-3v

    https://e2e.ti.com/support/amplifiers/current-shunt-monitors/w/faq/3599.exceeding-the-absolute-maximum-voltage-ratings-specifications-by-limiting-input-current

     

    Since you have only built a couple boards your issue might not be due to ESD at all. I would recommend continuing to refine the circuit until you have it working. Once you have something that works, then try stressing the ESD cells of the system by plugging in and disconnecting phone many times or powering phone on and off while it is connected to charger or get an IEC-61000 generator. You can also observe the VCM with oscilloscope to see what kind of load transients exist. Additionally, if you can’t ever get the circuit to work, try changing one variable at a time until it does: removing ADC, replacing phone with resistive load, move to prototype setup, trying similar current sense amplifiers like INA210, INA181, or the more ESD robust INA199B and C versions. You could also add series resistors to inputs of INA199 in working system to limit current during EOS event and then stress system to see if it will INA199 will break.

     

    I hope this helps.

     

    My references are:

    http://www.ti.com/lit/an/slyt492/slyt492.pdf

    http://www.ti.com/lit/an/slla305/slla305.pdf

    http://www.ti.com/lit/ug/tidu473/tidu473.pdf

    http://www.ti.com/lit/ds/symlink/tpd4s012.pdf

     

    Sincerely,

    Peter Iliya

    Current Sensing Applications