TL331: TL331KDBVR TIKG vs T1K8

Hi Kevin,

Thanks for your post. Yes, TL331 family is undergoing PCN so it is you may be receiving a mix of the old and new die when ordering the TL331. This new die is the same die as the "TL331B" device. See the document here: Product Change Notification (PDF) . To tell which is the old die and which is the new die, you can look at the assembly site code "CCO" field on the box label. If it says "USA", it is the old die. If it says "CHN", then it is the new "B" die. The main differences between TL331 and TL331B can be found on the front page of the datasheet. The new die has much lower bias current which would affect the resistor network. It looks like the input common mode range is being violated. Can you show waveforms of the inputs and outputs and the supplies of each pass and fail case? 

hi Chi Nguyen1:

1. Confirmed ：die is  "TL331B" device

 2.  It looks like the input common mode range is being violated

  reply:According to the schematic diagram, the maximum voltage between V - and V+ is 3.6V, the power supply voltage is 15-3.3=11.7V, and the maximum input common mode voltage is 11.7-1.5=10.2V. It should not exceed the maximum allowable common mode voltage

 Is this how to explain the common mode voltage range?

Hi Kevin,

shall this be an oscillator? Are you sure that the schematic is drawn correctly and that the given component values are correct?

R6 and R6 produce a voltage of 3.55V at the +input. To allow the output of TL331 to toggle, the output voltage must be able to exceed 3.55V. So, "Z2" must not be a 3.3V zener diode. If the circuit still works, then only because "Z2" shows manufacturing tolerances.

Keep in mind, that 3.3V zener diodes are highly non-ideal (compared to 6.8V zener diodes, e.g.) and show only a very soft knee in the current versus voltage curve. There's no abrupt change at all and the voltge drop caused by a certain flow current can drastically vary from diode to diode and with temperature.

To make the voltage drop across "Z1" more immune against supply current changes of TL331 and of the circuit (R1 !), you could think about adding a dummy load from the anode of "Z1" to signal ground and by this to force a more constant current to flow through "Z1". Another improvement could be achieved by uncoupling "Z1" and "Z2" from each other. Right now they are coupled via R1. And you should think about zener voltage of "Z2" and/or the set threshold voltages of TL331.

If this really is an oscillator, then you should also provide hysteresis as shown in figure 17 of datasheet of LM2903-N, e.g..

So, beccause your circuit is non-ideal, even the least and subtly change will make the circuit stop working properly. This is not the fault of TL331.

Kai