ADA4000 comparator

Hi,

ADA4000 is used as a shmitt trigger (non inverting) Transistor is a 2N4416 and there is a serie resistor (value 100kOhm) between op-amp/comparator and transistor.

Could you provide me with reasons people don't use op-amp as comparators ?

thx

Hello Jajajib,

OpAmps and comparators are designed for very different modes of operation at a fundamental level. An operational amplifier is intended to be used in systems which are linear and negative feedback is employed, keeping the operating point of all the internal parts in well behaved operation. Comparators, on the other hand, are designed to be tolerant of running open loop, or with some positive feedback, such as your schmitt trigger design.

In the design of a comparator, the devices internal to the part are usually designed to be predictable when they become saturated, similar to some families of logic. You can force the input stage of an OpAmp into saturation by subjecting the inputs to a very large difference. In the book, "Troubleshooting Analog Circuits," by the late Bob Pease, he mentions that an old OpAmp with a slew rate of 0.5V/us with its inputs overdriven by just 5mV will break the output's linear behavior and then slew at a rate of 0.01V/us. The device was simply not designed for that function. The complexities which make up this change in performance would lead to a discussion about device behaviors (such as transit time in the active BJT, versus transit time in saturation of the BJT which dictates the response time of the device), which is outside of the scope of this forum.

The result is that an OpAmp may be used as a comparator, but it will not be as fast or as predictable as an actual comparator. As such, by applying an OpAmp instead of a dedicated comparator, you will find that obtaining a very high slew rate (dv/dt) will be a very difficult task.

Hi,

I would like to extend my issue, not only on using ADA4000 as a comparator but also as an op-amp.

I am not sure that my electrical circuit is not the root cause of my problem (I cannot provide with the system circuit) but this ADA4000 op-amp seems to have some troubles (KO) and I would like to check if this op-amp has no internal sensitivity.

What would be the behaviour (or failure mode) of the ADA4000 op-amp with :

- a disymetric power supply or no synchronous power supply when power on

- ESD

- bad or far decoupling capacitor on power supply

- latch up phenomenae

- any cause which could lead to a failure on the op-amp

thanks in advance

Jajajib,

  Please attach a schematic of what you are trying to do.

Harry

Hi,

I am sorry but I have already mentionned that "I cannot provide with the system circuit" and it is not the purpose of my question. System behaviour of "what I am trying to do" work fine with an other op-amp.

I try to find the root cause at system level which could lead to op-amp damage but I would like to know more about this op-amp by valuable datas from AD, not general consideration about using op-amp and system analysis.

So, and maybe you cannot provide with an answer too for other reasons, are there precautions by using this op-amp (not mentionned in datasheet) due to problem on power supply (disymetric, over voltage), ESD, etc ?

jaja

Hello Jaja,

We do not necessarily need the full system schematic, but merely the section with the amplifier. Without the schematic, we have no means to understand how the device is being applied and with it, the possible failure modes. One can neither diagnose a malfunctioning black box, nor declare with certainty the root cause if the box cannot be opened or investigated.

Poor supply decoupling will generally not be a problem as that is a function of PSRR.

Asymmetric supplies are not a problem, so long as their values do not exceed the absolute max ratings of the part. ESD is a possibility, but again no information is provided about the failure mode to believe it to be ESD with any certainty.

Cheers,

-David

Thx for your help David,

I understand the black box problem. I will think about it.

In addition, and in order to find the root cause of failure, I have tested each op-amp by wiring them as an inverting circuit (Gain = -2). I inject a DC voltage source between 0V up to 20V on the inverting input of op-amp (not recommanded regarding the datasheet when op-amp is power supplied by +15V/-15V) and non inverting input is wired to ground.

2 failure modes are raised.

- output is always equal to -Vsat (i.e close to -13,7V or -14V) whatever input voltage is between 0V and 20V.

- output follow the inverting input, as to know the voltage source, and op-amp clamps output voltage to 15,7V or 16V when voltage source is between 15V and 20V.

I try to find a common failure mode, any idea on what I can do on my black boxes ?

Additional clue : it is possible that sometimes I have injected a DC voltage source on inverting and non-inverting inputs before powering op-amp on +/-Vcc. Not recommanded too isn't it ?

Hi,

Its it possible that there are a few sources of failure here. You did mention that the output actually rails to the maximum value it can on the low-side. This would tell me that it isn't a fault on the output, such as an ESD diode fusing to the low side supply due to an over-stress condition.

I have some questions:

If you take your input and ground it, what does your output read?

What are the values of resistors you are using for your gain of -2?

For a gain of -2, with 0-20V in, you would output 0 to -40V. As your supply on the negative side is limited to -15V, I would expect anything beyond +7V to rail the device. Once you go about 7V in, I would then expect the inverting terminal to follow (at an offset) the input: At 8V in, with Vout = -14, your inverting terminal would read about 0.66V. At 9V, the inverting node would go to about 1.333V, and continue to increase proportionately at 2/3V per V of Vin. When you continue to exceed the compliance of the device, you may be observing phase inversion of the part.

Really, without a schematic it is very difficult to debug the circuit. What I then would recommend is to look at our op-amp text book for more information: http://www.analog.com/library/analogDialogue/archives/39-05/op_amp_applications_handbook.html

Cheers,

-David

Hi David,

I just read your answer, I will try to check on devices what happens taking into account your question.

I also have additional clues for this issue. ADA4000 datasheet precises that without voltage supplies (power voltages), input current on inverting and non inverting should be limited to 10mA.

During nominal mode, input current should be limited to acceptable values on inputs but during power on/off on inverting input, an overvoltage occurs which could lead to a current over 10mA.

I have attached input voltages seen by each inverting input on all ADA4000 used ( 4 different op-amp used with 4 different functions) when they are not power supplied (no power voltage).

Do you know if these input voltages could lead to damage of ADA4000 devices when there is no voltage supply and why ?