Hi,


Newbie here.


I am a mechanical engineer with very limited knowledge of electronics. I am trying to educate myself by reading (Great Forum, BTW), and , like Shockmyselfsmart, have been experimenting with a 741 just to see what does what..


With V+ of 1.33v, I get a Vo of about 0.82v, with no Vin connected. Having read the previous posts in this thread I understand the idea of the offset null on the 741.


However, I have also been playing with a 6462 dual opamp, and with a V+ of 1.33v I get a Vo of up to 1v, again with nothing connected to Vin. The 6462 does not have offset null pins.


My question is this: Is a Vo normal for a 6462, or have I blown the amp, or am I just being a muppet? If a Vo is normal, how do I compensate for it?


Sorry if this is probably a very basic question, but please be gentle with this virgin!


Thanks,


Terry,

 

Welcome to the forums.

Please start your own thread instead of hijacking an existing one - http://forum.allaboutcircuits.com/showthread.php?t=47789

With the inputs open, any stray voltage will cause an unbalanced condition and therefore an output. There is no feedback, so the output will swing as far as possible and stay there.

The LMC6462 has an input impedance about 500 times greater than the 741. The inputs really need to be tied to ground or some reasonable input to give a sensible output.

 

I don't really understand what you did... could you post your circuit on the thread?

 

Hello Terry,

741 opamps are ubiquitous, but are now ancient relics as they were introduced some 40 years ago. At the time, they were the greatest thing since sliced bread.

They require 8v to 9v difference from +V to -V to get any kind of range. It's best to use a dual supply with them, otherwise you'll have to bias the inputs at 1/2 your +V.

Have a look at the attached simple schematic and simulation. I have set up a dual +10v/-10v supply to power two opamps, a 741 and a 6484, the latter a relative of your 6462. Note that these two particular opamps are unity-gain stable (gain=1), which is not the case with all opamps; some require gains considerably higher than 1 in order to be stable. They are wired as voltage followers (gain of 1); the inverting input connected to the output, and the non-inverting input connected to ground.

I have not shown "bypass capacitors" across the V+/V- supply inputs. At a bare minimum, 0.1uF capacitors must be connected from V+ to ground and V- to ground on each IC if configured with dual supply rails, or just from V+ to ground if powered with a single supply rail. Frequently, an additional capacitor of 1uF or higher is also required. Consult the relevant datasheet(s) to be certain.

The outputs from the two opamps are shown at the right. The 741 model has an input offset of 1mV; the 6484 roughly 750mV.

Note that if the opamp were connected with a gain higher than 1, the offset would be amplified along with the input signal.

The inputs to opamps or comparators must always be connected to something; even if their outputs are not used. Otherwise, the unconnected inputs will "float" at arbitrary levels, and will cause unexpected fluctuations/oscillations on the other channels' outputs.

I realize that this does not yet address your input offset adjustment question, but first you need to have a method to test what your input offset actually is.

 

Thanks for the quick replies.


Beenthere: Thanks for the advice regarding the necessity to connect the inputs to ground and an input.


Chudiandeyu: I have attached a diagram. Please note that I have applied 1.3v to Vin(+) and gounded Vin(-) since my previous post.


SgtWookie: Thanks for the extensive reply. Don't worry about “not addressing my original offset question”; I will forget the741 and concentrate on the 6462 from now on. I will study and hopefully digest your spreadsheet.


Sorry I didn't reply as quick as you guys; I had to go out, then I had another play with the breadboard, then had my dinner, then had another play and it is starting to make sense now. In my playing I did realise a couple of things which you guys learned in first grade!


Thanks again for your help. You have effectively answered my question. I'm sure I'll be back picking your brains again at some time.


Terry

 

Terry,
I'm afraid that the schematic you posted is not a good way to test an operational amplifier for its' offset voltage.

Since you only have a single 9v battery to work with, here's how you can use the other half of the opamp to create a "virtual ground"; since R1=R2, the "virtual ground" will make it appear to the opamp that you have dual 4.5v supplies instead of a single 9v supply:

Note that all of the ground symbols are electrically connected together, and considered as the 0v reference point.

Have a read through our E-book Chapter 8, which covers some basics of operational amplifiers: http://www.allaboutcircuits.com/vol_3/chpt_8/index.html

Keep in mind that opamps have limited output current, and some of the tutorial schematics use very low-value resistors. To get more realistic results, try to keep your resistors in the range of ~2k to ~500k or so. Much lower than 2k, and you'll start using a fair amount of power and the output of the opamp may not be able to "keep up". If you start getting into really large values of resistance, electrical noise can get to be a problem.

 

Thanks sgtWookie,

I will read up on the link you suggested. I already have a couple of questions, but I will read the link before asking.

Thanks

Terry

 

Why did Gitbash say he connected the Vin(-) to ground when the schematic doesn't show it connected to ground?

 

Hi Audioguru

I take your point: I was pointing out that, based on Beenthere's first reply to me, in which he told me that I needed to connect Vin+ & Vin- to "either ground or a reasonable input signal", I had taken his advice, and my circuit had evolved a bit more. What I should have stated is that "I have connected Vin(-) to ground through R2". Sorry for the confusion.

Also, just to be sure we are all singing from the same hymn sheet, I was referring to my pdf attachment. (I don't know how to embed my diagram into my message body like sgtWookie has). SgtWookie's diagram is considerably different from mine because.
sgtWookie is, I believe, trying to illustrate to me how to measure the offset voltage of the 6462.

sgtWookie. My diagram represents the circuit I was playing with to validate the theoretical gain from ampA. I had disregarded the offset because (if I've understood correctly) it is far less of an issue with the 6462 than with the 741. OK, perhaps the offset does exist with the 6462, but for the purposes of my exercise (simply checking the gain just to prove to myself the fundamentals) I've chosen to ignore it, and maybe revisit it at a later stage.
Also I have totally ignored ampB, because I have assumed (correct me if I'm wrong) that ampA and ampB can be considered as two independent amps, and I am only considering ampA.
Just to clarify, what I was doing was using a 9v battery to supply V+ , but a separate 1.5v battery (slightly discharged, so only actually supplying about 1.3v) connected to Vin(+). Both batteries -ve connected to the common ground. The purpose of my exercise was to achieve a gain of 1. (which eventually I did!).
Regarding your advice about the best range for the resistors, prior to your advice I had tried 5 ohms, then 151 ohms, and then after reading a tutorial somewhere on the web in which the writer suggested between 1k and 100k, I used something like 16k (can't quite remember at the moment) and the results were much more meaningful, so I appreciate good advice like yours; Thanks for taking the time to explain the "real world" reasons why the values make a difference.

I still haven't gotten round to reading your suggested link, but I will when I get a chance. I'm just trying to fit this electronics stuff in with everything else!

Thanks once again,

Terry

 

Here's Terry's original circuit from his PDF, just so nobody else has to download it:

And yes, I was trying to illustrate both a way to create a "virtual ground" and an easy way to measure the input offset using unity gain. Note that not all opamps will be stable with unity gain; you have to refer to the datasheet to find that information.

SgtWookie. My diagram represents the circuit I was playing with to validate the theoretical gain from ampA. I had disregarded the offset because (if I've understood correctly) it is far less of an issue with the 6462 than with the 741. OK, perhaps the offset does exist with the 6462, but for the purposes of my exercise (simply checking the gain just to prove to myself the fundamentals) I've chosen to ignore it, and maybe revisit it at a later stage.

That's OK, but you do need to keep in mind that even though most modern opamps are trimmed for a low input offset, an offset usually still does exist - and it is multiplied along with whatever gain you have the opamp set to. This can "bite" you if you have the opamp programmed for a lot of gain via the feedback resistors. Knowing what the input offset is by comparing it with a known input (in this case, referenced to ground) will let you know what that fixed error is. Errors, like enemies, accumulate.

Also I have totally ignored ampB, because I have assumed (correct me if I'm wrong) that ampA and ampB can be considered as two independent amps, and I am only considering ampA.

Well they are, but "ignoring" the unused opamp channel can (and most often will) "bite" you, particularly if the inputs are high impedance (like JFET and CMOS opamps). The neglected channel will oscillate unpredictably, slamming the outputs into saturation and causing odd behavior in the channel you're using. For opamps that are unity-gain stable, you wire the output to the inverting input, and the non-inverting input to ground, or an arbitrary reference that's between V+ and V-.

Regarding your advice about the best range for the resistors, prior to your advice I had tried 5 ohms, then 151 ohms, and then after reading a tutorial somewhere on the web in which the writer suggested between 1k and 100k, I used something like 16k (can't quite remember at the moment) and the results were much more meaningful, so I appreciate good advice like yours; Thanks for taking the time to explain the "real world" reasons why the values make a difference.

You're welcome. If you look in the datasheets, you'll often see values for Rload specified as >=2k. When you're operating on batteries, keeping the current consumption down should be a priority, as batteries are expensive. You can kill a 9v battery pretty quickly by using low values of resistance.

Resistors are inherently noisy. Small capacitors can quiet the noise down considerably. Think of resistors more or less like the nozzle on a garden hose. When you have them partially closed and restricting the flow of water, they can make a good bit of noise. However, toss the nozzle in a bucket of water and things get really quiet again. The capacitor acts like a bucket of water.

I still haven't gotten round to reading your suggested link, but I will when I get a chance. I'm just trying to fit this electronics stuff in with everything else!

Busy is good. Reading that tutorial will help a good bit in understanding how to figure out gain and various opamp configurations.

Texas Instruments published a PDF entitled "Opamps for Everyone"; here's a link to it:
http://focus.ti.com/lit/an/slod006b/slod006b.pdf

Have a read through it; lots of good stuff in there.

 

Thanks again sgtWookie

Terry