Application of the Current-sensing in the compressor of air conditioner

Hi Zhihong,

hmm, I see no special reason why the TLC2272 and OPA2374 are preferred. Sometimes it's quite simple: Someone develops a circuit and uses these OPAmps. And afterwards the circuit is copied by the competitors. Or the manufacturer of another component used in the circuit publishes an appnote which contains these OPAmps.

Having said this, the designers of these OPAmps have definitely found a successful compromise of specifications.

Kai

Hello Wesley,

I am not too well-versed in this end equipment, but can offer some guidance for this application.  Here are my comments to your questions.

1.  For current-sensing, you'll typically want a part with low offset voltage and low bias current to give you the most accurate result possible.  Low noise is also probably desirable.  Additional considerations will likely include bandwidth and quiescent current.

2.  My guess is that these parts are popular because they have acceptable offset, bias current, and voltage noise all for a decent price.  By the way, another decent part for this application might be the TLV6742.  It has better voltage noise and offset specs at the cost of greater bias current.  I'm not sure who the competitor parts are for this.  I would loop in the field team or someone else in the product line if you'd like to dig into this some more.

3.  I would imagine the current-sensing is used on the motor which drives the compressor.  This is especially likely if it's an electric motor where the motor speed needs to be controlled.

I hope this helps.  Please feel free to email me if you would like more details or would like to be put into contact with other individuals in the product line.

Regards,

Daniel

As a final note, I would agree with Kai's point as well.  Sometimes the first design gets copied by others later on.  So there may be a better part available, but it's simply left in place as the combination of price and performance are good enough to not justify spending time on a redesign.

Regards,

Daniel

Hi Daniel,kai

Thanks for your specific reply.

1. I agree your opinion about the Vos and Ios.

2. But I don't understand very well about why the slew rate , bandwidth and Iq effect the current sensing of the compressor, could you please explain that generally?

Hello Wesley,

Bandwidth and IQ should not be important in the sense that they will primarily determine the accuracy of you system.  Rather, you may need a certain speed or may have certain power constraints which limit, you in picking these specifications.  I suppose this could be said for almost any application.  But, I wanted to point it out for completeness.

I hope this helps.  Please let me know if you have any further questions.

Regards,

Daniel

Hi Zhihong,

I'm adding more color to this discussion on current sensing in a compressor / motor system. Depending on how you want to run the motor (i.e. different motor algorithms), you can use either low-side current sensing or in-line current sensing.

• Low-side current sensing is the easiest and most common way of sensing current in motor control. In addition to the op-amp products mentioned above, you can also look at the specialized current sense amplifier INA180, which simplifies the design for you a little bit by integrating gain networks and at the same time provide better accuracy. Depending on how many shunt resistors you want to use, i.e. 1-shunt vs. 2-shunt vs. 3-shunt, you'll need 1 to 3 op-amps or current sense amplifiers accordingly - obviously I'm assuming you're using a 3-phase motor here.
• Now more and more people try to adopt FOC algorithm to control motors for better efficiency and acoustic performance, and to get the best out of FOC you want to sense current in-line, to get continuous, real-time current information. The challenge here is that you're dealing with a high voltage (220V AC gets rectified to 310V DC), and there's not a lot of options available out there. You can take a look at TI's latest TMCS1100, which allows you to sense current up to 600V, through magnetic current sensing instead of shunt technology.

In addition to current sensing for motor control algorithms, more and more people also want to monitor the total current the motor is consuming. The best way to do this is to measure the total current on the high side, which gets rid of the ground bouncing resulted from low-side current sensing, but you're faced with the same high voltage challenge here. TMCS1100 is an awesome way to solve this challenge.

I recommend you to look at the following documents if you'd like to understand a little more about current sensing in compressor and other motor systems:

http://www.ti.com/lit/an/sboa161b/sboa161b.pdf

http://www.ti.com/lit/an/sboa351/sboa351.pdf

Hi Zhihong,

I don't think that a compressor circuit places unusual demands on the used OPAmps. These are just standard circuits.

To have enough feedback, or better called "loop gain", to keep the OPAmp always in its linear range, the use of an OPAmp with high enough bandwidth and slew rate is preferred. For standard controlling applications as mentioned by Zhao an OPAmp should have a bandwidth of let's say 3...10MHz and a slew rate of 3...10V/µs. And if the OPAmp combines these specs with a low current consumption, so much the better.

Kai

Hi Miller,

Thanks for your reply, your sharing benefits me a lot.

Best Regards,

Wesley Huang

Email: wesley-huang@ti.com

Hi Kai,

Thanks for your reply, your sharing benefits me a lot.

Best Regards,

Wesley Huang

Email: wesley-huang@ti.com

Hi Zhao,

Thanks for your reply, your sharing benefits me a lot.

Best Regards,

Wesley Huang

Email: wesley-huang@ti.com