TINA/Spice/TLV2434: TLV2434 High temperature controlled

Hi Charles,

It looks like you are likely running into the offset voltage drift increasing your offset voltage and therefore your error at the higher temperatures. Our colleague Pete Semig wrote a great article on measuring high voltages at temperature, please give this a read:
www.eeweb.com/.../considerations-for-high-voltage-measurements

You can also try to find an amplifier with a lower offset and drift specification so that it is contributing less offset. By my calculation, a typical device has 300uV offset and 2uV/°C of drift, so your offset voltage could go from 300uV to 420uV, which is a big increase!

You labeled something on the schematic 600V - is that BUS- and BUS+? What is the voltage at the gate of the two transistors Q3 and Q1?

-Paul

Hi Charles,

is R351 shorting the source and drain of Q1?

Where exactly do you measure the error voltages of 2V and 12V?

Kai

Hi Ron，

I used the same circuit as the 4-channel channel, but only CHB has deviation coefficient,is the temperature coefficient of the chip problematic? Can it be improved?

The op-amp's temperature coeffienct can't be changed. If you need something with lower drift, you'll have to choose another part.

Were you able to try the following suggestions from Ron and Kai?
* Removing Z5 and Z7 and test again.
* Measure all the node voltage to find out where the error occurs most.

Can you answer these questions?
* Is R351 shorting the source and drain of Q1?
* Where exactly do you measure the error voltages of 2V and 12V?
* What is the voltage at the gate of the two transistors Q3 and Q1?
* You labeled something on the schematic 600V - is that BUS- and BUS+?

Hi Paul,

Can you answer these questions?
* Is R351 shorting the source and drain of Q1?

    R351 and Q1 are selective relations, only R351 is the fault PCB, not applied Q1.

* Where exactly do you measure the error voltages of 2V and 12V?

    The errors of 2V and 12V were read by the upper computer, and the errors were 12.345mV and 74.085mV after dividing by the partial pressure ratio.

* What is the voltage at the gate of the two transistors Q3 and Q1?

    Q3 and Q1 did not stick, and the voltage of R351 and R352 was 3.704V

* You labeled something on the schematic 600V - is that BUS- and BUS+?

    The labeled fault channel is BUS-

Hi Charles,

after your additional information I agree with Ron and do also think that the error is caused by the leakage current of Z5 and Z7. Datasheet of MMZ5232 says that the leakage current at 3V reverse voltage can be up to 5µA at 25°C. And your signal voltage across Z5 and Z7 is even 3.704V, which is much higher than 3V and will cause a much higher leakage current.

The simulation shows that an error of 12.345mV corresponds to a current of 1.5µA through the zener diode. This is well within the specification of MMZ5232. So, there seems to be no problem at all with the TLV2434 but with your zener diode.

The simulation shows that 9.09µA must flow through the zener diode to give an error of 74.085mV at 85°C. Such an increase of leakage current with raised temperature is quite normal for your zener diode.

Kai