INA188: large error when input=0

Hi Justifice,

you don't show what is connected to the -Vs pin of INA188 and you don't say whether the thermocouple is connected to the input during the measurement. If the thermocouple is disconnected, you have to consider the input offset current of INA188 flowing across R3 and R44  and causing an input offset voltage error.

Also, keep in mind that a current of (1.65V-0.283V) / 40k = 34µ is flowing out of the REF pin of INA188. Can the reference voltage generator handle this reverse current?

I would decrease R13 and R47 to 1M. But in any case I would mount a 100nF filtering cap in parallel to R13.

Kai

Hi Justifice,

In addition to Kai's suggestions, here are my comments:

1. The offset voltages on the instrumentation amplifier is specified in RTI or Referred To the Input. If the TC is shorted, Vout = Gain*)V + Vref + Vos (total).

Since Vref = 0.283V, I will expect the output will be approx. 0.283V + Vos, see the total Vos offset calculation below, where Gain =  159.228 V/V. Your figure of 0.255V is close and I do not know the actual system configuration. 

BTW. there is a polarity on the TC leads and please make sure that it is oriented properly. 

INA188 TC 01272022.TSC

2. Did you mixed up the INA188 with our other instrumentation amplifier (IA)? The output transfer function does not match INA188's datasheet. It is more resemble to INA826 or similar other IA parts. Please check on the part. On page 7 of INA826 EVM application note under 3.2.1 Input Filtering section, it is shown how you can design your differential and common mode LPFs. 

Enclosed are 3 application notes for your references.

https://www.ti.com/lit/an/sloa123/sloa123.pdf?ts=1643299826619&ref_url=https%253A%252F%252Fwww.google.com%252F

https://www.ti.com/lit/an/sbaa354/sbaa354.pdf?ts=1643300021242&ref_url=https%253A%252F%252Fwww.google.com%252F

https://www.ti.com/lit/ug/sbou115c/sbou115c.pdf?ts=1643238109838&ref_url=https%253A%252F%252Fwww.google.com%252F

3. You may be able to increase your TC temperature range if you place Vcm higher than 1.65Vdc. Please consult it with our Analog Engineer Calculator. I do not have your Temp sensing requirements, so I am unable to comment on it. 

https://www.ti.com/tool/ANALOG-ENGINEER-CALC

If you have additional questions, please let us know. 

Best,

Raymond

Kai, thank you, -Vs is grounded.

We've measured no matter thermocouple is connected or disconnected, the output is almost the same, when disconnected the output is 0.259V and when connected, the output is 0.255V.

Hi Justifice,

Can you show us the schematic marked in red rectangular block? 

C11? at the input of Vref pin is unlikely. Vref pin is required to be a low impedance node, which is capable of source and sink current. With Vref = 0.283V, which is likely generated from an op amp, see the image below. If that is the case, C11 should not be placed at the output of the op amp. Please check on it.  

Another question is your INA818 exposed to excessive high heat during reflow or wave soldering? We are assuming that INA818 is functioning and not damaged for some uncontrolled process. 

Best,

Raymond

Raymond,

thank you.

Yes it's driven by an OPAMP. 

In fact we've built 6 pcs boards and all behaves the same.

• connect thermocouple, both sides are room temp, OUTPUT=0.255V, in theory it should be 0.283V
• unconnect thermocouple, J2 unconnected, OUTPUT=0.259V, in theory it should be 0.283V
• unconnect thermocouple, J2 pin1 and pin2  shorted, OUTPUT=0.252V, in theory it should be 0.283V。
• remove R13, R47，OUTPUT=0.283V.

Hi Justifice,

remove C11 !!!

A cap at the output of U4A can destabilize the circuit.

Kai