ADS124S08: how to interface 3 NTC with ADS124S08?

Hi Bill Xu,

Can you please help respond to my question in the Notes section of the Thread Tracking Toolkit? We will review your questions while we wait for your response

-Bryan

Hi, Bryan

I visited my customer last Thursday which design high voltage power supply for x-ray/CT system. during the visiting, customer ask us how to monitor the IGBT/SIC-MOSFET temperature to improve system reliability. then I study the IGBT/SICMOSFET and find there had integrated 1 or 3 NTCs in the module.

my first idea is to use ADS1220 to monitor IGBT/SICMOSFET temperature, however, for a 3-phase SIC-MOSFET/IGBT module , there has 3 NTCs, these means that I need 3 PCs ADS1220 to monitor modules temperature. this will cause cost increasing and large PCB area. So I want to try ADS124S08 for cost down and compact design.

I feel this is a general issues in electrical car since SIC-MOSFET was widely used in electrical car to drive motor. but I could not find any application notes to monitor SIC-MOSFET temperature in ti.com. thanks for your supporting.

regards,

Bill

Hi, Bryan,

use internal reference to build an resistor divide is workable. but the function between resistance of thermistor with bias resistor is non-linear. in actual application, customer may use MOSFETs without integrate NTS resistors, then potential long wire need compensation. the third issues is an general ADC can easy to test divider voltage, this will make our solutions has lots of competitors and no much benefits. so, I feel that my original idea should more attractive for customers compare competitive solutions.

do you think that the orginal solutions can work or not? is there any risks that I'm not considered? thanks.

regards,

Bill  

Hi Bill Xu,

The RREF resistance needs to be larger than the NTC resistance for your proposal to work. Do you know what the NTC resistance range is that needs to be measured? This will help you determine what the maximum current can be. It is possible that the ADS124S08 does not offer a small enough current to bias the resistor divider

Let me know what the resistance range is for the NTC, then we can figure out if your solution will work

-Bryan

Hi, Byran

the resistance range is 300ohm~10kohm. thanks.

regards,

Bill

Hi Bill Xu,

This should be possible. If you assume that RREF is maybe 12kohm, that's 12kohm + 10kohm (max) = 22kohm. Assuming the ADC AVDD = 5V, and the compliance voltage is AVDD - 04V = 4.6V, that's a maximum current of 4.6V / 22kohm = 210uA. The ADS124S08 can output 100uA, so that should be doable.

Of course the output signal will be very small when the NTC = 300ohms compared to when the NTC = 10kohm using IDAC = 100uA.

Do you really think the lead resistance on the NTCs will be significant enough to require lead compensation using 2x IDACs?

-Bryan

hi, Bryan

when NTC resistance equal 300ohm, we can adjust internal PGA's gain to improve the accuracy. Am I right? in actual application, customer designer potential varies from case by case. this means that in some application, wire resistance maybe neglectible. but some application could not neglect. So, I feel it is good to keep lead compensation. BTW, do you think that my original scheme workable and no errors? thanks.

regards,

Bill 

Hi, Bryan

yes, I try to use a 3 wire RTD for lead compensation. does the original circuit runable? could you please propose a circuit between ADS124S08 and three 3-wire NTCs? big thanks.

regards,

Bill

hi, Bryan

one thing: why need remove the resistor and filter capacitor? I'm not get your idea. could u please a little explain? I feel it should be workable if not remove the resistor and filter capacitor. thanks.

regards,

Bill