• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 1121 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.

INA282 or LMP8640 and short circuit detection

Francisco Javier Sanchez Rivas
Prodigy 130 points
Other Parts Discussed in Thread: INA282, LMP8640, INA200, INA138, INA139

Hello all,

I would like to measure unidirectionals currents ranging from 0A to 20A and with DC voltages from 10V to 30V. Analog voltage output is also required. In my case the load will be a car horn, and the application is for life testing it. It could be the case that after some cycles the load would break and short circuit itself. My questions are:

-  If the load get short circuited, wouldn't all the voltage drop in the shunt resistor? Would the differential input withstand this voltage?

The measuring system can cut the power to the load, so monitoring the current I could act in consequence but I don't know if the IC could suffer any damage. Additional thermal-magnetic overcurrent circuit breakers is mounted to the load, but it needs some time to break the circuit.

Also, I would like to know how to calculate the error similar to the one shown in page 4 of this document: Current shunt monitor. I don't know if INA282 would be better in my application than LMP8640. I plan on using Vsense=50mV, but I could consider higher or lower Vsense if I get better performance or less total error.

Thank you.

  • 0
    TI__Mastermind 25825 points

    Hello Francisco,

    When the load is shorted you must ensure that the INA282 is not exposed to conditions that violate the absolute maximum ratings table (page 2 of the data sheet).

    Concerning the accuracy question...the link appears to be broken. Nonetheless, the INA282's initial input offset voltage, CMRR, and PSRR are all better than the LMP8640. For more information on how to calculate the accuracy of a solution, please take a look at Part III of the following article series. There is also an accuracy calculation on pages 21 and 22 of the INA282 data sheet.

    I do have one potential concern about using the INA282 in this application. Please understand that the INA282 has a switched capacitor input structure. While this yields excellent accuracy, it does not lend itself well to step changes at the input (it will take time for the output to settle). Please refer to the figures on pages 6 and 7 of the data sheet.  Then again, perhaps such a delay is not a concern in your application.

    Also, you may be interested in a current shunt monitor with integrated comparator. Please take a look at the INA200-2 family of devices.

    Finally, you state that the application is for a car horn. Will you require Q1 devices? If so, I can move your post to our Automotive forum.

  • 0 in reply to Pete%20Semig
    Prodigy 130 points

    Hello,

    thank you for the answer Pete, I was already reading these current sense tutorials, really nice work for current sense newcomers like me ;)

    I know I must ensure absolute maximum ratings in the datasheet, what I don't know is how can I manage high CM voltage, like 30V and still effectively detect short circuit in the load without damaging the IC.

    Regarding the step response of the INA282, I overlooked that... I want to power on and off the horn continuously and accurately measure the current, even in the transient. Also the horn will oscilate by his own inner working generating a 500Hz pulsating current, that is what I want to log, the current wave. Do you know if at that frequency I could use INA282? Any other suggestion for current shunt monitor?

    Even if it is for a car horn, I will use the IC in lab equipment, so no automotive class is required.

    Thank you.

  • 0 in reply to Francisco Javier Sanchez Rivas
    TI__Mastermind 25825 points

    Hello Francisco,

    Please take a look at Figure 31 in the INA200-2 data sheet. The comparator is used to disconnect the supply from the load. The frequency should not be a problem, either. Please keep in mind that the current shunt monitor in the INA200-2 is not as 'accurate' as the INA282. However, I think it would be better for this application (BW and on board comparator). One word of caution, though. Please ensure that you're operating the device in "Normal Case 1" conditions. In other words, the sense voltage must be at least 20mV (and Vcm=>Vs). You mentioned a sense voltage of 50mV in a previous post, so I don't think this will be an issue.

  • 0 in reply to Pete%20Semig
    Prodigy 130 points

    Hi again,

    first of all sorry for so many questions, but I'm just starting to learn about current sensing, your help is really appreciate. Looking at Figure 31 in the INA200 data sheet, if the load gets short circuited, supposing shunt option 2 and 30VDC supply, wouldn't all the voltage drop across the shunt just before the comparator could drive the MOSFET? Would it damage the IC?

    About INA282 vs INA200-2, why do you mention the BandWidth? Also, I want to measure a dc pulsating current that can range from 0 to 20Amps, so the Vsense will drop below 20mV. When I said 50mV I was making reference to the maximun Vsense, so for 20A I will have 50mV drop in the shunt. Also, it's not a problem to have separate comparator, as I will have to add also opto-coupling for interfacing Vout to my ADC.

    Just for additional information, here is a capture of a very similar current wave to what I want to measure, but with inverted sign. In this case it's 460Hz and 6A. Even if you continuously power the horn, the electro-mechanical principle of working will open and close contacts inside the horn, so the current wave if produced and the sound.

  • 0 in reply to Francisco Javier Sanchez Rivas
    TI__Mastermind 25825 points

    Hello Francisco,

    When the load shorts, the current will not change instantaneously. So, you should be able to detect the short by properly setting the current limit. You set the threshold voltage using R1 and R2 as depicted in Figure 31 of the INA200-2 data sheet.

    Concerning Figure 31, when the load shorts and is disconnected from the supply, there should be no current flowing through the shunt resistor and therefore the device should not be damaged. I prefer shunt location 1 for the common-mode will remain at the supply voltage whether or not the load is shorted.

    Concerning my mention of bandwidth...while the INA282 has a BW of 10kHz it performs best when measuring DC currents. Please take a look at Figures 16 and 17 in the INA282 data sheet.

    Our current shunt monitors actually measure voltage. In order to generate a voltage, you need a minimum load current. In other words, you can't accurately measure 0A. Instead, you'll really be looking at the offset voltage of the device. Therefore you need a minimum load current. The article series has an example of how to size a shunt resistor.

    For your application, our offering is limited beyond the INA200-2 (unless you decide to move forward with the INA282). The is due to the common-mode voltage of 30V. The only other devices that you may be interested in evaluating are the INA138/68 and INA139/69. They do not require a 20mV minimum shunt voltage and have a lower initial input offset voltage than the INA200-2. They are current-output devices that require a precision external resistor to set the gain accurately. They do not have an integrated comparator, but that does not sound like it will be an issue.  

  • 0 in reply to Pete%20Semig
    Prodigy 130 points

    Thank you Pete,

    I think I will give a try to INA282 and check the results. I cannot use INA200-2 because of the Vsense>20mV limitation to the accuracy.

    Thank you for all!

  • 0
    TI__Guru 60036 points

    Hello Francisco,

    I contacted to the designer of the 8640. Internally there is a 7V ESD diode across the inputs. If the full 12V was applied across the inputs from a low impedance source, this diode would be blown, damaging the device. Even just a few milliseconds may be enough energy to blow the diode.

    So I would NOT recommend the LMP8640 if you suspect the inputs may see the full load supply voltage across them.

    The LMP860x and LMP848x series should not have this problem.

    Regards,

  • 0 in reply to Paul Grohe
    Prodigy 130 points

    Thank you very much for the information Paul. I think I will go ahead with INA282, as it's accuracy is very good.