LM124A: Output is non-linear at near zero range

Hi Shinu,

I would think about adding a small negative supply voltage by the help of LM7705.

Kai

Shinu,

To get a low output voltage, the sink current needs to be less than 12uA. The current from V7 through R410 and R409. 

Also consider tuning the function to be 26V to 30V being 1V to 5V 

Is 5.1V is most accurate voltage to use as a reference? If not what voltage is the most accurate? 

If 13.7 is accurate then this could be done with two op amp channels.

Is the voltage across R416 (1 ohm) significant or could it be ignored?

Hi Ron,

Thanks for your reply. 

• Customer cannot use a negative supply in our circuit due to some limitations.
• The input range is 26V to 40V for which the expected output is ~0V to 5V.
• Customer have made some changes in circuit (PFA) after which the voltage reference at output stage is changed to 6.4V (temperature compensated zener is planned).
• For this circuit : Vbias (or +Vcc) = 13.7V and -Vcc= 0V (grounded) single supply opamp.

Modified the circuit with the following changes-

1. In input stage added an offset voltage reference of 2.5V.
2. In third stage limited the sinking currents to less than 34uA, by changing the  4 resistors R11, R12, R13 and R14. [All resistors are in the range of 45k to 120k]
3. Last stage ref changed to 6.4V as explained above.

By doing the above changes customer got the simulation result and hardware result in ambient temperature as expected without any non-linearity near zero. PFA- The circuit and the simulation result.

However can you please give your input regarding the behavior of this circuit or LM124 for temperature range of -20^oC to +60^oC.

Regards, Shinu. 

Hi Shinu,

the negative supply voltage generated by the LM7705 is only -0.232V.

Kai

Shinu,

Change R13 to 301k; change R12 to 229k ; to reduce sink current.  6.4V / (301k+229k) = 12.1uA  

I also suggest a capacitor in parallel with R13 to reduce pause lag in the feedback. 10pF is plenty.

This should provide enough margin. Margin is not certain because there is no full temperature parameter in the electrical table of the data sheet.

I can suggest a replacement op amp that does have full temp electrical table specifications.

Hi Ron,

Thanks for the feedback. Will update schematic. 

In one of the design LM124 is used with a single ended supply of 14V. Out of four ,one Amplifier is used and the rest are terminated.
Termination method adopted is inverting buffer configuration. But some error happened during jumper implementation.

In U1C pin9 inverting input got connected to GND instead of left open and in 'U1B' non inverting input is left open instead of GND.
Output of U1A measured 12.138V for the input of 0.756V and Vcc current calculated to be 13.5mA.

Customer would like to know how much the current would have been shared between U1A & U1C and the dissipation allowed in each stage.

After correcting the pcb can they use the same device ? Can the device fail or degrade due to this?

Regards, Shinu. 

Shinu,

Assuming the 13.5mA calalcation was correct. Power is 14V * 13.5mA  = 189mW. That is not a lot of temperature rise. I would expect the device to be good. Depending on the application and tolerance for risk, you may not want to ship this device to third parties. 

There is a chance that 13.5mA was not correct; that value seems incorrect to me. I would expect lower or higher value. Unless normal current is 13.5mA

Thanks Ron,

When one of the amplifier outputs shorted to ground via the input through feed back, how the current sharing happens to rest of the amplifier

pins of the IC?

Thanks.

Shinu,

Only pin 4 is shared. There is common VCC+ metal bus on the silicon die

Thanks Ron.