INA198: Unnecessary voltage measurements on INA198

Hi Jame,

what is the load?

Kai

Hello Kai,

Fig1. This is Load in Block Diagram.

And "Simplified Schematic" is described in the datasheet.

Did you see the INA198 datasheet or post upper?

Load is a device consuming current.

Best Regards,
Jame,Shin

Jame,

I will look at this on the bench tomorrow. 

The reason that Kai asked about the load in your application.  I think he was asking in case there was something from the load that could supply some of the voltage when the high supply +5V is not supplied.

Have you disconnected the load as well to test if the INA198 was supplying the voltage?  What is the range of the current pulled from the load?  Could there be a load to ground.  If there is then the Voltage should drop close to GND.

Hello Javier,

Thank for your reply

Javier Said> Have you disconnected the load as well to test if the INA198 was supplying the voltage?

              A> Yes, it is.  And control the power supply to Load in front of INA198. (Picture attached.)

Javier Said> What is the range of the current pulled from the load?

            A> For maximum current of 1.5A and maximum voltage into ADC VREFP of 3.3V,   ( min 0A,  typ 0.7A , max 1.5A)

                 with a multiplication factor of 100 for the INA198 the maximum voltage across the sense resistor is 3.3V/100 or 0.33V.

                With a maximum current of 1.5A the maximum resistor is Rs = V/I  = 0.33/1.5= 0.22 Ohm.  ≒ 0.2 Ohm

Javier Said> Could there be a load to ground.  If there is then the Voltage should drop close to GND

              A> Load is unnecessary voltage.

                    What should I do?

Best Regards,
Jame,Shin

Hi Jame,

Javier is right. I wanted to know whether the input bias currents of INA198 have a chance to flow to ground. See figure 9 of datasheet.

Without a low ohmic path to ground (through a low ohmic load, for instance !), the only path to ground might be the multimeter which you have connected to the inputs to measure the "unnecessary" voltage.

Please connect for a test a 1k resistor from the Vin- pin to ground and measure again the voltage between the Vin- pin and ground.

Kai

Hello Kai & Javier ,

Thank you for reply .

Q1> Is R (1K ohm) solved for input bias current problem?

Q2> Is the unnecessary 1.52V voltage 0V? ((How few Volt)

Q3> If input 5V is supplied, does INA198 work normally?

Best Regards,
Jame,Shin

Hi Jame,

I recommended the 1k resistor for a test. What voltage do you measure with this 1k resistor mounted? Still 1.52V?

I would remove the common mode choke. Without adequate filtering cap or at least TVS on its left side the choke can produce huge voltage spikes (inductive kick back !) whenever you open the switch on the left. This can kill the INA198 and the MCU!

To your third question: Yes, the INA198 (like any other amplifier) always needs a path to signal ground at its inputs. This can be a supply voltage, a regular load or a resistor as dummy load.

Kai

Hello

In the INA198 datasheet, specify the R1(5Kohm) , RL(500Kohm) resistance value.

And specify Input Current into Any Pin (max) 5mA.

VR1 = I R , 0.005 * 15Kohm = 75V

V Rsolution = ( Rsolution / (Rsolution + R1)  ) * 75V

                  =   1K / (1K+15K) = 4.6875V

Actual VRsolution(1K) = 20mV is measured.

Q1> If so, is IRsolution(1K) = 2uA current value ?

Q2> Please explain the difference between 4.6875V and 1.77V ( Javier said  ).

Q3> If common mode choke is removed, is the power output to the outside bad for Pi (Power Integrity)?

        As a side note, the application of the Tagetget Board works after the 5V switching is fixedly switched.

Thank you for explaining your TVS needs to avoid damaging peripheral devices due to excessive voltage switching noise for power switching.

Best Regards,
Jame,Shin

Hi Jame,

I don't understand your calculation. Where does the 1k and 15k in your calculation come from?

Also, please carefully read the footnote (2):

Kai