0

I was reading about the Input bias currents for an op-amp. And I had a question regarding the ones internally compensated for.

So in general if I was to use an Op amp for any particular application and I have an op amp with the internal compensation for the bias currents, then do I ever have to worry about the bias currents or connecting resistors to compensate for the design ?

Also if the answer is yes to the above question, then why don't people just use the internally compensated ones ?

Dallas Carter
  • 367
  • 1
  • 7
  • 16
  • Please do tell us which op-amp has "internal compensation" for bias currents. – Andy aka Mar 18 '15 at 12:26
  • I am afraid, you are mixing external bias compensation and internal frequency compensation? – LvW Mar 18 '15 at 12:32
  • Otherwise - look here: http://www.analog.com/media/en/training-seminars/tutorials/MT-038.pdf – LvW Mar 18 '15 at 13:09
  • Are you talking about a chopper op amp? I think the best thing you can do is give us a part number. – Scott Seidman Mar 18 '15 at 13:51
  • Looks like I misunderstood it then. I was talking about figure 2 in http://www.analog.com/media/en/training-seminars/tutorials/MT-038.pdf . Since it talks about internal current source, I thought it was something inside the op amp – Dallas Carter Mar 18 '15 at 13:53
  • Yes - correct. I rather would say that it is a kind of internal biasing instead of "compensation". – LvW Mar 18 '15 at 15:16

2 Answers2

0

The compensation you're talking about will reduce the bias current, but not eliminate it. The improvement will be reflected in the specifications for the data sheet of any op amp. Whether you need to worry about it or not will depend on your application and specific circuit design.

Scott Seidman
  • 29,274
  • 4
  • 44
  • 109
  • Thanks. But since the bias current usually mentioned in the circuit is that max value, then why would different applications/circuit designs will not have the bias current reduced from the compensation? – Dallas Carter Mar 18 '15 at 14:50
  • @DallasCarter The number shown in the data sheet as max already reflects any bias compensation. What you as a designer can ignore (or not worry about) is VERY application specific. – Scott Seidman Mar 18 '15 at 14:54
  • Please read the rest of the paper. It states that the drawbacks of the compensation technique include increased current noise and uncontrolled bias polarity, and gives an example of a situation where bias polarity is important. Likewise, very high-impedance, high-gain amplifiers (like some transconductance circuits, for example) cannot accept the high current noise produced. As Scott Seidman says, it depends on the application. – WhatRoughBeast Mar 18 '15 at 15:47
0

Some op-amps have input current so low that there is no need for compensation (it's just nA or even fA leakage).

Some have uncompensated bias current (could be uA, practically enough to dimly light an LED, for some types).

Some have input bias current compensation, reducing the input bias current by a compensation scheme. They're generally not as low input bias current as op-amps with a MOSFET input but they can have other advantages (voltage noise may be lower than the MOSFET types, but input current noise is higher than the bias current would suggest).

Each have advantages and disadvantages, which is why there are literally hundreds of different op-amps.

Bias current is only one of a plethora of sources of errors in an op-amp circuit, and often not a very important one. Often just matching the DC impedance each input sees will do as good a job of bias current compensation as some internal network that hides the internal (noisy) bias current.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842