The schematic you posted shows 1k in series with the 10k pot to ground. Now you say you don't have a resistor from the +input to ground.
If you are changing the circuit, you need to post a schematic of what you are describing. We can't read your mind.

We can tell you the purpose of each resistor in an op amp circuit, but we have to know what the circuit looks like.

Yes, the 1K resistor shown is in SERIES with the input.

Based on the data sheet, I tried several resistors from the input to ground ALONG WITH the 1K resistor in series with the input.

Did you even look at the data sheet?

I didn't redraw the circuit because it's so simple anyone should be able to picture in their mind another resistor running from the noninverting input to ground.

As I stated above, this resistor had a direct effect on the offset with this particular device.

Nobody is asking you to "read my mind".
Comments like that I don't care for and contribute nothing to this thread!

 

Yes, I looked at the datasheet. I posted the bias current specs in post #5.
The reason I thought you were talking about a different circuit was this statement:

In my circuit, I don't have a resistor going from the noninverting input to ground.

I thought you would realize that 1kΩ in series with a 10kΩ pot would result in a total resistance of between 1k and 11k from the +input to ground, depending on the setting of the pot (assuming nothing else is connected to the top end of the pot). I guess I assumed you knew that series resistors add.

Op amps have bias current. Some are insignificant. This one has A LOT of bias current.
A simplified explanation of the effect of bias current follows:
The bias current, multiplied by the resistance from the input to ground, results in a voltage on the input of V= Ibias*Rtotal (Ohm's law). This voltage will be multiplied by the closed loop gain of the op amp (5.3, in your case). If you are getting more than about 350mv of output offset, something else is wrong.
Can you post a picture of the circuit as built? It may be oscillating due to layout or wiring issues.

 

When selecting op amps, more (higher frequency rating) is not necessarily better than less. The disadvantage of using a high frequency op amp, when you only need to amplify audio frequencies, is that it is much more susceptible to high frequency oscillations (as well as having more high frequency noise). It thus requires a more careful layout, very short leads on the decoupling capacitors (surface mount preferred), and possibly a ground plane for good decoupling. As Ron noted, that oscillation is a likely explanation for the strange offsets you are seeing.

A good low distortion, audio frequency op amp such as the MC34071 single opamp, the MC34072 dual, and the MC34074 quad would be better and easier to use.

 

Yes, I looked at the datasheet. I posted the bias current specs in post #5.
The reason I thought you were talking about a different circuit was this statement:I thought you would realize that 1kΩ in series with a 10kΩ pot would result in a total resistance of between 1k and 11k from the +input to ground, depending on the setting of the pot (assuming nothing else is connected to the top end of the pot). I guess I assumed you knew that series resistors add.

Op amps have bias current. Some are insignificant. This one has A LOT of bias current.
A simplified explanation of the effect of bias current follows:
The bias current, multiplied by the resistance from the input to ground, results in a voltage on the input of V= Ibias*Rtotal (Ohm's law). This voltage will be multiplied by the closed loop gain of the op amp (5.3, in your case). If you are getting more than about 350mv of output offset, something else is wrong.
Can you post a picture of the circuit as built? It may be oscillating due to layout or wiring issues.

Yes, I realize you just add series resistors to get the total.

I see now how bias current can become important. Thank you for the example.
Even .35V sounds like too much offset for injecting into a audio amplifier.

The circuit is assembled on a breadboard just for experimentation. I actually made up a couple of PCB's with a large ground plane on the bottom side. They should be here in the next few days.
I'll transfer everything to one of them and see how it goes.

 

When selecting op amps, more (higher frequency rating) is not necessarily better than less. The disadvantage of using a high frequency op amp, when you only need to amplify audio frequencies, is that it is much more susceptible to high frequency oscillations (as well as having more high frequency noise). It thus requires a more careful layout, very short leads on the decoupling capacitors (surface mount preferred), and possibly a ground plane for good decoupling. As Ron noted, that oscillation is a likely explanation for the strange offsets you are seeing.

A good low distortion, audio frequency op amp such as the MC34071 single opamp, the MC34072 dual, and the MC34074 quad would be better and easier to use.

I have a LOT of the popular single and dual op-amps from LT, TI, Burr Brown, etc., but I have never heard or seen any MCxxxxx series op-amps.

If you know where I can order a couple, I'm certainly open to trying them.

 

There is a bit of an anomaly here. In post #14 the OP stated that varying the pot position had no effect on the output offset.

 

When selecting op amps, more (higher frequency rating) is not necessarily better than less. The disadvantage of using a high frequency op amp, when you only need to amplify audio frequencies, is that it is much more susceptible to high frequency oscillations (as well as having more high frequency noise). It thus requires a more careful layout, very short leads on the decoupling capacitors (surface mount preferred), and possibly a ground plane for good decoupling. As Ron noted, that oscillation is a likely explanation for the strange offsets you are seeing.

A good low distortion, audio frequency op amp such as the MC34071 single opamp, the MC34072 dual, and the MC34074 quad would be better and easier to use.

Yeah, I spent most of my career as a circuit designer of analog and digital video TV broadcast equipment (timebase correctors, wideband analog and digital video routing switchers). I REALLY know about video amps, and there perverse intention to oscillate.

 

My current guess is that the pot is not grounded as shown in the original schematic. Otherwise as it adjusted it should vary the output offset. Maybe the original measurement was done with the NI input open. This might account for the uncertain value noted by the OP between observations.

 

I have a LOT of the popular single and dual op-amps from LT, TI, Burr Brown, etc., but I have never heard or seen any MCxxxxx series op-amps.

If you know where I can order a couple, I'm certainly open to trying them.

Another good choice is the TL071 single, TL072 dual and TL074 quad.

You should be able to buy the op amps at Digi-Key or other online electronic suppliers.

The main criteria are low noise, low distortion, and good output swing (slew rate) at 20kHz for audio amps.

 

There is a bit of an anomaly here. In post #14 the OP stated that varying the pot position had no effect on the output offset.

With no input resistor going to ground, varying the pot position had no effect on the offset. It was a steady 1.6 volts regardless of position.

It's grounded correctly or I wouldn't be able to change the volume using other op-amps.

 

Yeah, I spent most of my career as a circuit designer of analog and digital video TV broadcast equipment (timebase correctors, wideband analog and digital video routing switchers). I REALLY know about video amps, and there perverse intention to oscillate.

Sounds like you know your stuff.

So what is the purpose of a resistor connecting a op-amp's input to ground?
I see some audio circuits using them and some only have a series resistor like in the circuit I drew.
Here's one with a 100KΩ resistor at the input:
http://diyaudioprojects.com/Chip/CMoy-Grado-RA1-Headphone-Amp/

 

Sounds like you know your stuff.

So what is the purpose of a resistor connecting a op-amp's input to ground?
I see some audio circuits using them and some only have a series resistor like in the circuit I drew.
Here's one with a 100KΩ resistor at the input:
http://diyaudioprojects.com/Chip/CMoy-Grado-RA1-Headphone-Amp/

All op amps need a DC path to ground (or to a voltage, which in turn connects to ground). They need this to establish a source of bias current. In the schematic you referenced, the input is AC-coupled. DC current cannot pass through a capacitor. The 100k is the ground return for the input bias current.

 

With no input resistor going to ground, varying the pot position had no effect on the offset. It was a steady 1.6 volts regardless of position.

It's grounded correctly or I wouldn't be able to change the volume using other op-amps.

It's interesting that you have observed three different readings at various times. Initially 1.3V, then 1.7V and subsequently 1.6V.

My reference to there being an anomaly was in relation to your comments about connecting the non-inverting input to ground via various resistor values in series with the 1k. If I understand you correctly, you were able to reduce the DC offset at the output using this method. As you pointed out earlier, the non-inverting input was originally tied to ground via the 1k resistor in series with the 10k Volume pot - actually the pot wiper terminal. This was indicated in the original schematic you posted. Hence with a full variation in the pot position one would have a 1k-11kΩ variation in the grounding resistance.

If the input bias current was actually flowing in that path comprising the potentiometer & 1k series resistor, then one would expect a variation in output voltage DC offset [as the pot position is changed] of the order 63mV to 413mV with a bias current of 6.6uA. In fact, you apparently observed no such variation but did observe a variation when you tied the non-inverting input to ground through an alternate resistance path.

Which led me to conclude there was no means for the bias current to flow via the potentiometer. Otherwise you would have seen a DC offset change .....

Unless you had made the original observation(s) with the pot disconnected from the cct.

This is the nub of the anomaly.

 

I believe the anomalies can likely be explained by the amplifier oscillating.

Does the offset change when you bring you hand near the circuit?

 

I believe the anomalies can likely be explained by the amplifier oscillating.

Which I suggested as a possibility in post #16.

 

It's interesting that you have observed three different readings at various times. Initially 1.3V, then 1.7V and subsequently 1.6V.

My reference to there being an anomaly was in relation to your comments about connecting the non-inverting input to ground via various resistor values in series with the 1k. If I understand you correctly, you were able to reduce the DC offset at the output using this method. As you pointed out earlier, the non-inverting input was originally tied to ground via the 1k resistor in series with the 10k Volume pot - actually the pot wiper terminal. This was indicated in the original schematic you posted. Hence with a full variation in the pot position one would have a 1k-11kΩ variation in the grounding resistance.

If the input bias current was actually flowing in that path comprising the potentiometer & 1k series resistor, then one would expect a variation in output voltage DC offset [as the pot position is changed] of the order 63mV to 413mV with a bias current of 6.6uA. In fact, you apparently observed no such variation but did observe a variation when you tied the non-inverting input to ground through an alternate resistance path.

Which led me to conclude there was no means for the bias current to flow via the potentiometer. Otherwise you would have seen a DC offset change .....

Unless you had made the original observation(s) with the pot disconnected from the cct.

This is the nub of the anomaly.

I apologize, the 1.3 volt reading was a mistake. Without any resistor going from the input to ground(as in my original circuit drawing) the offset was always between 1.6-1.7 volts...VERY high and varying the pot didn't change it at all.

So after looking at the headphone circuit above and the data sheet for the AD828, I tried several value resistors(100K to 1K) from the input to ground.
I noticed that as I tried lower value resistors the offset also came down.
I stopped at 1K and the offset measures 35mV.
In the data sheet they're using values below 100 ohms.

Putting a 100 ohm or even a 1K ohm resistor in parallel with volume pot doesn't make much sense to me as I always see people using much higher values.
But when you use higher value resistors in that position, the offset jumps back up again.

 

I apologize, the 1.3 volt reading was a mistake. Without any resistor going from the input to ground(as in my original circuit drawing) the offset was always between 1.6-1.7 volts...VERY high and varying the pot didn't change it at all.

So after looking at the headphone circuit above and the data sheet for the AD828, I tried several value resistors(100K to 1K) from the input to ground.
I noticed that as I tried lower value resistors the offset also came down.
I stopped at 1K and the offset measures 35mV.
In the data sheet they're using values below 100 ohms.

Putting a 100 ohm or even a 1K ohm resistor in parallel with volume pot doesn't make much sense to me as I always see people using much higher values.
But when you use higher value resistors in that position, the offset jumps back up again.

You are correct. putting 1k or 100Ω in parallel with a 10k pot doesn't make much sense.
The AD828 is a video amp. It wasn't designed with 10k pots in mind. Video is wide bandwidth, and needs low resistance values to minimize the effects of parasitic and stray capacitance to preserve that bandwith.
Most video signals are transmitted and received on 75Ω transmission lines, so low impedances are not a big issue.
IMHO, you are wasting your time and ours, dikkin' around with a video amp in an audio application. Use a proper audio op amp, and most of the offset will go away.

 

I must say this has been a fascinating thread with all its twists & turns in attempting to find an answer for the OP.

Hopefully foxfire3 has grasped the concept of why three reasonably experienced forum members were confused by the information presented. We had assumed quite reasonably that the original schematic posted was a true representation of the the situation.

I commend Ron for his perseverance in spite of the serve he got from foxfire3 about not being sufficiently intuitive to realize the volume control wasn't connected to the circuit and that (if he had taken the time to read the data sheet) he should have known that a bias grounding resistor was needed to achieve correct circuit operation.

I wonder if had occurred to foxfire3 that the rest of us had reasonably assumed the bias current was flowing via the volume control's resistance to ground. An important matter which he seemingly failed to recognize.

Pity there wasn't a CRO at hand - then we could have known whether the circuit was oscillating as proposed.

Next time try an LM741 - (that's meant to be joke by the way)

Thanks for the journey.

 

I must say this has been a fascinating thread with all its twists & turns in attempting to find an answer for the OP.

Hopefully foxfire3 has grasped the concept of why three reasonably experienced forum members were confused by the information presented. We had assumed quite reasonably that the original schematic posted was a true representation of the the situation.

I commend Ron for his perseverance in spite of the serve he got from foxfire3 about not being sufficiently intuitive to realize the volume control wasn't connected to the circuit and that (if he had taken the time to read the data sheet) he should have known that a bias grounding resistor was needed to achieve correct circuit operation.

I wonder if had occurred to foxfire3 that the rest of us had reasonably assumed the bias current was flowing via the volume control's resistance to ground. An important matter which he seemingly failed to recognize.

Pity there wasn't a CRO at hand - then we could have known whether the circuit was oscillating as proposed.

Next time try an LM741 - (that's meant to be joke by the way)

Thanks for the journey.

All the twists and turns??
The original schematic was EXACTLY what I was working with. After measuring the huge dc offset at the outputs, I tried several resistors from the inputs to ground just to see if there was any change.
Well, there was a change as I have explained several times now.

The volume control was always connected, so I'm not sure what you were referring to there.

Hey, I'm no op-amp expert as you can tell. I was asking you guys why the huge offset?

With that much offset, I'm simply not going to just sit around and see if you guys can figure it out for me.
I'm going to bust out my meter and iron and start changing things just to see what happens.

So I add one simple resistor and some people seem to be "totally lost" since I've "changed" the circuit.

Anyway, I think Ron has nailed it with video op-amps having to work with only 75 or so ohm impedances vs. audio amps working with high input impedances and relatively low output impedances.

Thanks for your help as well TNK.

 

OK thanks for the clarification. The potentiometer was always in circuit.

I am still at a loss then to explain why you could test various resistor values from the non-inverting input to ground (or at least in parallel with the pot wiper to ground ?) and obtain a change in output offset whilst adjusting the volume control setting had no effect at all. The latter variation in potentiometer setting is in principle equivalent to placing the various resistors across the potentiometer. If that's not the case then you seem to have discovered an interesting phenomenon that appears to be at variance with expected circuit behavior.

Unless your circuit was actually as shown in the attachment with the potentiometer orientated the other way around ....

I shan't bother you any further but leave you with that conundrum to consider.

Good luck with your project!