LTC2983 Error reporting question, 4 wire connection, PT1000, Kelvin Rsense, Rsense sharing

Question

Hello, We are using LTC2983 to measure the temperature on four PT1000 probes connected by 4 wires to LTC, Rsense connection is Kelvin type (5 kOhm). Below you can find scheme: 
 Measurements work properly, but we noticed that disconnecting specific cable from the probe does not generate a measurement error in the LTC (Read Results, bit D24 is always 1). Disconnecting the following two cables from the probe causes the LTC to report a measurement error: but... We have the problem with disconencting the cables below (only when disconencting one per test, diconnecting two at the same time will produce an error), we don't see the LTC reporting a measurement error, but the temperature reading goes up by about 200 degrees Celsius. We wonder if we are doing something wrong, or is the LTC simply behaving like this when disconnecting a single wire from the two that I have described? We just want to make sure if we can still improve something in our electronics or in the program, or simply assume that LTC works like that. Thank you Best regards Zbigniew Czerski

KSMendoza · Accepted Answer

Hi zbyxon , The bit D24 just reports if the conversion cycle was finished, and the conversion result is ready to be read. Disconnecting either CH2 or CH3 won't report an error, as the ADC is still able to read the differential voltage between the connected pin and the capacitor placed beside the pin where the wire was disconnected. 
 However, disconnecting both channels 2 and 3 will cause an error, as this would read a value outside of the expected range of the PT1000. Disconnecting the wires between CH2 and CH10, or the one at CH4 will also produce an error as this would disrupt the path of the current. I hope this helps. Regards, Karlo

Electronic Design Services - PartnerZone

LTC2983 Error reporting question, 4 wire connection, PT1000, Kelvin Rsense, Rsense sharing

Top Replies