OPA188 & OPA140 & OPA237 VCM Range not as in Datasheet

Hello Said,

The three op amps that you mention are all rated with a common-mode input voltage that extends to the negative supply rail, or even a bit lower. That isn't the issue that prevents the output from achieving a true zero volt level.

Even though the OPA140 and OPA188 are listed as rail-to-rail output, the output doesn't quite swing all the way to the rails. The datasheet Electrical Characteristics tables list how close their output will swing to the rail. It runs anywhere from tens of millivolts, to hundreds of millivolts. Therefore, if you apply a common-mode voltage of 0 V to the non-inverting input in a single supply application, the output won't quite make it to 0 V. It is surprising that you found the OPA237 swings closest to ground of the three. It has the least swing to the rail specified, (V-) + 1 V.

A remedy to get their output to true 0 V is to provide a low negative voltage for V-. Usually that is inconvenient because the person using the device wants to only use a single, positive supply. One way to get around using an actual negative supply is to incorporate a negative bias generator, and power the op amp V- pin with it. In particular, the LM7705 low-noise, negative bias generator is used for this purpose. You can find its information here:

There really aren't a lot of options when it comes to the output swing issue. 

Regards, Thomas

PA - Linear Applications Engineering

Thank you Thomas for the valuable answer and infos, I didn't know about the negative bias generator.

I know about the output swing and have seen the graphs for all of them and really was confused to see the opa237 can deliver even negative output (-8mV) if you measure it with oscilloscope(while with multimeter (Average value) it´s near to 0V don´t exceed +2mV in worse case) .
extracted from datasheet description of the opa237: " When operated from a single supply, the input common-mode range extends below ground and the output can swing to within 10mV of ground". Another thing that I have seen the input common mode range specified from -0.2V independent from the negative rail (not like V- - 0.2V)

any difference ? A special feature for this op amp in concrete?

The opa237 was used in our applications for more than 8 years and I'm planning to do a small redesign. It would be helpful if we can have a zero drift- very low noise op amp with the same output swing "feature" without using a bias generator.

I will qualify the neg bias generator for my application in the meantime.

Thank you again, I was pleased to use the very impressing TI forum

Hi Said,

If there is only a positive supply source in the circuit the only way a negative voltage can be developed if a capacitance/inductance in the circuit is being charged, and then discharged, and the direction of the resulting current flow supports a negative voltage. We have mostly observed small negative voltages relative to ground being developed when fast voltage transients are present on waveform edges.

I am not sure why you are observing the behaviors you have noted with the OPA237. If you do use the LM7705, you should be able to go all the way to 0 V at the output without going into output saturation. The OPA140 and OPA188 would be good, modern op amps that have high performances.

Regards, Thomas

PA - Linear Applications Engineering

Hi again,

I understand perfectly your point, there is no added capacitance/inductance in the circuit. The op amps are qualified as pick and place in a very simple circuit and I have noted the difference. I have tested many different higher gain 101 (100K and 1K) 51 and 26 to see the effect of bias current(without compensation), I have seen that in the case of the opa140 and opa188 no big change still the 7mV - 8mV. Except for the opa237 you can see a proportional change it raises 5mV-6mV for the gain 101 and 3mV-3.5mV for the 51 and so on .

one more question the LM7705 can drive 2 op amp (opa188) at the same time ?

Regards ,

Said Khalil

Hi again,

I understand perfectly your point, there is no added capacitance/inductance in the circuit. The op amps are qualified as pick and place in a very simple circuit and I have noted the difference. I have tested many different higher gain 101 (100K and 1K) 51 and 26 to see the effect of bias current(without compensation), I have seen that in the case of the opa140 and opa188 no big change still the 7mV - 8mV. Except for the opa237 you can see a proportional change it raises 5mV-6mV for the gain 101 and 3mV-3.5mV for the 51 and so on .

one more question the LM7705 can drive 2 op amp (opa188) at the same time ? Also I'm thinking how I could get a true GND always, If I compensate the -232mV from LM7705 should be enough for all the temperature range and load driving capability ? or my compensation should be adaptive ?

Regards ,

Said Khalil

Hi Said,

The capacitance and inductance I mentioned would be parasitic in nature such as ones associated with PC board traces, etc.

It appears from the LM7705 datasheet that the device is capable of sinking as much as 35 mA, but 20 mA is probably a practical limit. The OPA140, or OPA188 negative supply current, plus their output currents would be sunk through their V- supply pins, and those currents would flow into the LM7705 output to Vss. Therefore, the total V- current would have to be accounted for when considering the LM7705 current capability. I would think that one LM7705 would provide enough current capability for two of the op amps used in practical applications.

Do note that the two amplifiers are available as dual devices, OPA2140 and OPA2188.

Regards, Thomas

PA - Linear Applications Engineering

Hi Thomas,

I´m again back here with a question for the OPA237 and OPA188. There is any application note or articles related to robustness between those 2 OP amps.

Or between Bipolar and CMOS in general !!

How you could define robustness if you compare both OP amps with different technology (CMOS vs Bipolar) ?

Thank you in Advance

Hello Said,

There isn't a report specifically comparing the OPA237 and OPA188 robustness. The closest measure of product robustness that we and other semiconductor companies use is Failures in Time (FIT) and Mean Time Between Failures (MTBF) information. These are statistically derived reults and numbers based on High Temperature Operating Life (HTOL) tests performed on the products as part of their initial qualification. Both the OPA237 and OPA188 received this testing and their MTBF and FIT numbers were identical because of similar sample sizes and that no failures occurred for either one during the life test. So one might conclude that they have similar robustness.

If you are interested in general comparisons between CMOS and bipolar devices I found a couple of interesting articles on line. They may provide you with some additional insight. The URLs are provided after the DPPM/FIT/MTBF information I provided below.

Regards, Thomas

PA - Linear Applications Engineering

Hi Again,

Thank you for your support.

Can you estimate the life cycle of the opa237 ? it is getting obsolete soon ? and the OPA188?

This will help a lo us to decide

Hi Said,

I checked with our PA Op amp Marketing Engineering team and they were able to provide me with some insight on the OPA237 and OPA188 longevity plans.

Despite the OPA237 being introduced 20 years ago, it is a widely applied op amp. The semiconductor process that it uses is completely viable and supported by TI. The OPA188 was just released in 2013 so it is a relatively new device on a modern process. It was the first high-voltage, auto-zero op amp introduced by TI. It will be a available for many years to come. There are plans for obsoleting either device in the near-term.

In general, most of our TI and the op amps, and those originally produced by Burr-Brown, have very long lifetimes. Three to four decades of lifetime is not unusual. 

Regards, Thomas

PA - Linear Applications Engineering