Op Amps suitable for Adjustable Current Source Circuit

Sainer,

I suspect your stability issues are more related to the feedback loop of the switching regulator as opposed to the op amp itself. In general, capacitive load or capacitance on the inverting feedback node causes stability issues. Also, the output impedance can interact with the load impedance. Chopper amplifiers often have complex output impedance. One optional amplifier that has a resistive output impedance is the OPA2277. This device has similar offset and drift to the OPA2188.

If you are interested in some other adjustable current sources, below is a list of several tested designs. I think these topologies are simple and the stability is tested.

http://www.ti.com/tool/tipd101

http://www.ti.com/tool/tipd102

http://www.ti.com/tool/tipd103

http://www.ti.com/tool/tipd107

Hi Kay,

Thanks for your reply.
OPA2277 and OPA2188 have Common Mode Range V(+)-2 and V(+)-1.5V respectively.
Our power supply to the Op Amp is 5V and we want to drive the output of DC/DC converter up to 4V,
so in case of OPA2188 because the Max input becomes 3.5V, we can not use it.
Can you suggest a rail-to-rail Op Amps with the same characteristics as OPA2277 and OPA2188?.
Or should I put a higher Supply to the Op Amp as the last choice?

And one more thing, we have tested OPA2330 in the past and found that it has a switching/changeover
of PMOS/NMOS at a certain high temperature. OPA2277 and OPA2188 don't have this kind of characteristic, right?.

Sainer

1. 1.      Based on your common mode requirement, I suggest that you look at 5V CMOS rail to rail devices.
2. 2.      The OPA2330 is a chopper stabilized rail to rail device. To achieve the rail to rail performance the input stage does use a PMOS/NMOS input pair. During one part of the common mode range, the PMOS stage is active and during another part of the common mode range, the NMOS stage is active. This is a common method used to achieve a wide input common mode range. There will be some small amount of “crossover” distortion during the transition from one input pair to the other. This distortion is minimized by the offset correction from the chopper circuit. This effect should not be related to temperature. The same method is used on the OPA188. The OPA277 is a bipolar device and does not use this method. That is why the OPA277 common mode range is limited.
3. 3.      I would actually expect the OPA2330 to be a good option for your application and would not expect to see any noticeable crossover distortion or temperature effects. If you can send measured results or more detail on the issue perhaps we could better understand the problem. The OPA2335 is another zero drift amplifier with rail to rail inputs that you might consider.
4. 4.      If you are interested in a rail to rail 5V amplifier that does not have a PMOS/NMOS input pair this is generally done using an internal charge pump. The internal charge pump allows a single input pair to get rail to rail performance by boosting the internal power supply. A good example of this type of device is the OPA320.

 Art

Hi Arthur,

If you give me your personal contact, I will send you our measurement results.
Regarding OPA2188 I got an info that it contains PMOS only on it, so there isn't any crossover must be afraid of.

I have 2 questions for you.

1. Reading your post, it is still not clear whether OPA2188 is Chopper Op Amp or not.
On datasheet it is written that it is an auto-zeroing Op Amp,
and according to following link it is different with Chopper Op Amp.

www.analog.com/.../zero_drift.html

If OPA2188 is a Chopper Op Amp, then it suppose to have a noise density peak at
a chopping frequency. Can you tell me what the frequency is and how to deal with it?

2. I did the transient simulation of Current Driver Circuit using PSpice,
but the simulation went divergent with never ending calculation and stop at 22%.
But, if I replace the Op Amp with another one such as LMP2012, it worked well.
Seemed like there was a problem with OPA2188's model when putting it together
with LMZ21700 model. Can you fix this problem?.

For your information, I prototyped the circuit and up to this point, measurement result went well
 without any unstabilities occured. I put 12V Supply to the Op Amp to get the Common Mode
Range to 5V though.

Regards

Sainer

Sorry for the late response.  The opa188 is a chopper amplifier and its frequency is 750kHz.  You are correct that you will see a noise at the chopping frequency.  The OPA188 is pretty good at minimizing this issue.  This effect is minimized when gain is used as the choping noise is attenuated by the bandwidth limitation inherent in the amplifier.  The figure below shows the noise density out to wide frequencies in several gain configurations.  Also, this effect can be minimized by external filtering.  Finally, using small value resistors in the feedback and input is a good idea as this chopping noise is also a current noise.

Regarding the convergence issue, I will ask some colleagues for help.  The only quick solution I can offer is to try Tina Spice as Tina has a very good math engine and will often converge when other simulators will not.

Art