INA148: INA148+MUX507 protection

Hi Vladimir,

I have  a few questions:

(1) What is the maximum common-mode voltage you are required to measure in this application? 

The schematic shows an input voltage of +/-50mV to +/-12V differential signal flowing through 1..15km, but it does not indicate the maximum input common-mode voltage requirement.  The INA148 can operate/support common-mode input voltages up to +/-200V continuous without damage, however, the MUX507 can only support a maximum voltage limited by +Vss and -Vdd supplies.  In your schematic, +Vdd=+15V and -Vss=-15V, so the common-mode of the signal can not exceed this multiplexer +/-15V supply voltage.

(2) The INA148 is a difference amplifier, where the internal INA148 1MegaOhm input resistors are carefully matched to meet the gain, and common-mode rejection performance.  Please keep in mind, adding series resistance at the difference amplifier INA148 inputs will cause gain and common-mode rejection degradation. Therefore, although we could add protection resistors, there could be some degradation due to the mismatch of the resistances in the 1-15km of wire and any mismatch of the R3/R4 resistance.  For Example, 0.1% resistance mismatch will degrade CMRR to less than 60-dB. 

Keep in mind, if the common-mode voltage is expected to exceed +/-15V supplies, the MUX507 can not support this higher input common-mode voltage; and you will need an INA148 at each line (without the mux) input to reject the common-mode voltage prior the PGA281... 

Thank you and Regards,

Luis Chioye  

Dear Luis, thank you

Point is that, I do not need high measurement accuracy. Error 10-15% will suit me.

In fact, the signals, transmitted on this line, are pulsed signals with a duration of 10 ms. On the transmitting side, the signal has an amplitude of  +/-12 V.

But due to the line length it can weaken up to +/-50 mV at the receiving side. On the receiving side I need to detect whether there is a pulse or not. Since there are noises in the line, the analysis is performed using the ADC (after PGA281). If the amplitude of the pulse at the receiving side is more than 500 mV, then there are no problems at all - even the ADC can be omitted (I can use a comparator).  But if the amplitude is less than 500 mV, then against the background of noise it is already more difficult to analyze.

Honestly, I don’t know if the  common-mode voltage will be more than  +/-15V.  Actually shouldn't be.  But with lightning discharges, everything is possible. Gas arrester and suppressor do not always help. That's why I expect to "increase" the protection of MUX507 with resistors R3/R4.  Аnd I would like to know what size I can choose them: 1 kOhm,10 kOhm or 100 kOhm ?  It may not be very good, but putting INA148 at the input of each line is expensive. 

We used DG408DY multiplexers earlier.  In principle, everything worked, although sometimes there were failures. Are they any better, than MUX507? (in terms of surge protection).  Please, see attach. 

TIA

Vladimir Naumenkov

www.agat.by

DG408_.pdf

Hi Vladimir,

The R3/R5 protection resistor values are chosen so the input current into the MUX508 does not exceed the absolute maximum ratings during a fault condition.

The MUX508 internal ESD structure will turn on at voltages above +/-VDD/VEE supplies, and the analog input current should not exceed 30mA during an over voltage fault condition. The schematic shows that you have TVS diodes VD1, VD2 and VD3.  What is the clamping voltage of this TVS diodes during the worst case fault condition?  

Typically, you will select the resistor values to limit the input current to <30mA when the TVS diodes clamp the voltage during a fault condition.  If you know the maximum clamping voltage of the TVS diodes on the expected worst case fault, the resistors are chosen so the maximum current is ~10mA (less than the absolute maximum 30mA) when the internal ESD structure of the MUX508 is turned on.  The resistor value can be calculated using the worst case TVS diode clamplng voltage:

R3, R4 = (TVS_Clamping_voltage - VDD) / 10mA  

Thank you and Regards,

Luis 

Hi Vladimir,

For your reference, Section 12 of Precision Labs covers this over voltage Op-Amp input protection:

 https://training.ti.com/ti-precision-labs-op-amps  

The same concepts will apply when designing protection clamp circuits for the inputs of the multiplexer.

Thank you,

Kind Regards,

Luis