Input Impedance (RG?)

[RedHouse]

While looking at this schem of a line amp, I noticed the text explaining the spec on this said it had a 10K input impedance, how is that?

I thought that in a case like this, the first resistor (before the input cap) set the input impedance.



I would have guessed this is a 100K input impedance (silly me).

[R.G.]

Well, it's not 10K, you're right there.

It's actually close to 50K.

Here's how. When you look at input impedance, you have to look at all of the things that can eat current from the input. In this case, that's the first 100K, then the second 100K, because that's in parallel with the first one for all frequencies which are passed by the 47 (uF?) cap, and then the 1K in series with the non-inverting input of the opamp.

You have to check the opamp datasheet on things like this, but most bipolar opamps (this is one) have input impedances over 1M. So the 1K in series with the 1M is 1001K, and that's in parallel with two 100K's. I would ignore the 1.001M and just call it 50K.

The first resistor before the input cap sets the entire DC input impedance, since the cap blocks any DC loading, but the AC input impedance is set by anything that can load the signal, and in this case it includes the second 100K.

[amz-fx]

You have to check the opamp datasheet on things like this, but most bipolar opamps (this is one) have input impedances over 1M. So the 1K in series with the 1M is 1001K, and that's in parallel with two 100K's. I would ignore the 1.001M and just call it 50K.

Last time I checked the 5532 (and it's been a while)  it was 500k, which is the drawback of this otherwise excellent opamp.  That is in parallel with the two 100k resistors and makes about 45k for this circuit, which is close to the input Z of the LPB1...  not terrific for a full-range sound if you are driving it directly with a guitar, but it might suit the taste of the builder as the LPB is quite popular.

Best regards, Jack

[Sir H C]

For noise sometimes it is better to have lower input resistance, and if it is made to drive 600 ohms out, makes sense that it is not too high z-in.

[R.G.]

Last time I checked the 5532 (and it's been a while)  it was 500k, which is the drawback of this otherwise excellent opamp.

Yep. Like I said, there's no substitute for checking the datasheet.

[johngreene]

Well, I did check the data sheet and the input impedance is typically 300k with a minimum of 30k!
www.njr.co.jp/pdf/ae/ae04058.pdf

So the input impedance could be as low as 18k.

It also mentions putting a 1k current limiting resistor in series with the non-inverting input to keep from popping the protection diode on power up.

--john

[amz-fx]

Well, I did check the data sheet and the input impedance is typically 300k with a minimum of 30k!
www.njr.co.jp/pdf/ae/ae04058.pdf

Ouch!  That's lower than I recalled!

The 4558 is about 1M as is typical, which RG related...  in some cases it might get a better freq  range than the 5532 at the expense of a little more noise.

regards, Jack

[R.G.]

Well, I did check the data sheet and the input impedance is typically 300k with a minimum of 30k!

It just gets better and better. 

Like I said, there's no substitute for checking the datasheet.

[George Giblet]

> Well, I did check the data sheet and the input impedance is typically 300k with a minimum of 30k!

I think you will find that is the differential input impedance (not common mode to ground).

When you use feedback (like 99% of circuits) the differential input impedance is essentially eliminated from the scene.

You can even make a rough estimate of the input impedance opamp assuming it'a a simple differential pair:

Rin_diff = 2 * Vt / Ibias

where Vt ~ 26mV

Ibias for the 5532 is 200nA to 800nA, 800nA gives about 65kohm, 200nA gives 260kohm.

[RedHouse]

Thanks RG and Jack, ...you guy's rock.

I'm glad you guy like to hang out here, and I'm ever grateful you like to share.

[Roobin]

So basically, for guitar signal, its jsut 3 resitors in parallel.

Can it really be that simple?

[RedHouse]

It's actually the line input on a primary channel of an M72 mixing console.

[R.G.]

So basically, for guitar signal, its jsut 3 resitors in parallel. Can it really be that simple?

It can - it can also be much more complicated. Opamp circuits in general are much simpler than discrete circuits.

But the essence is always the same - find all the places that the current can go, and take account of how much signal each of them steals.

[WGTP]

I think this is a related question.  On my dual op amp distortions, I run a non-inverting op amp (tube screamer like) into an inverting op amp with resistors of from 1k to 100k connecting them. 

Are the smaller ones, say 10K resulting in a lose of treble due to the low impedance of the inverting op amp input???  IIRC the output impedance of the first op amp is very low.

[R.G.]

Remember why you get treble loss from guitar pickups with loading - the output impedance of a pickup rises with frequency because of the inductive nature of the coils.

As the driving opamp has a quite low output impedance, resistors from 10K up certainly don't load down the opamp, so you should see little if any treble loss. If the signal level is smallish - line level, for instance - you'll likely see no signal degradation between 1K-10K. This is why buffering a guitar prevents treble loss - the buffer can supply the current to force the lower resistance loads to follow along.

For 1K-10K, you are in the region where some opamps have trouble driving big output voltages.  Some opamps can only supply full current into resistances greater than 2K or so. Driving 600 ohm loads is problematical for them. But the degradation is a loss of drive capability, signal level at high frequencies or distortion, not treble loss.

The 5532, by the way, is one of the few opamps that can drive 600 ohms easily, and that is one reason it remains in high regard in pro audio circles.