Opamp gain setting resistors values

[deduikertjes]

Dear All,

Im'sure it must be somewhere on this forum, but even with extensive searching I couldn't find it.

Well here is my problem:

I've build Rod Elliot's project 88 to fit in a Quad 33 pre-amp. (http://sound.westhost.com/project88.htm ). It is a very straightforward design in 2 stages design using OPA2134 and series feedback.

As the balance pot in the Q33 is only 1K (and has 100 Ohm's in series on each side (as well as a cap of 100uF in series on each side)), the configuration in that project cannot be used by me.

So I thought of putting the balance pot in the position of one of the gain setting resistors (the one going to ground). To keep the gain equal I have to divide the other gain setting resistor (between pin 1 and 2) by 10 giving me a value of 1K.

So depending on the position of the balance pot I wil have the two gain setting resitors to be about 1K giving me about 6dB gain in the second stage.

The question: Is it possible to use gain setting resistors with such a low value (1K as opposed to 10K as is more often used). It should give a lower thermal noise, but as nothing is usually gained without a trade off there must be a catch somewhere. Is there?

Any help greatly appreciated, MArco

(attached the the picture of the second gain stage of the project 88).

[darkfenriz]

1k load to ground will result in 1mA output current for every 1V output voltage, so no problem. Check datasheet.

[AndrewT]

Hi,
r7=1k0 pot?

or

r7L=half 1k0pot and r7R=other half of 1k0 pot?

In the second case the effective r7 is only 500r and at centre rotation your gain will be 1000/500 +1 =3 (+10db instead of +6db)

I think it is more usual to use a gain setting pot in an inverting opamp layout but the disadvantage is that to get useful input impedance you will need a higher pot value.

Darkfenriz has given you the current limitation warning by defining the output current relative to input voltage. Have a look at the datasheet and find the distortion and drive capability graphs for low load values and then decide if you can live with this 1k0 (or 500R) pot value.

[myhrrhleine]

Hi,

There is a reason for higher resistance values.
While the datasheets do show the opamps can drive lower impedences, the distortion also goes up.

For minimum distortion you want to keep impedence higher.
Many dadasheets show comparisons of distortion at 600ohms and 2kohms. Quite a difference.

[deduikertjes]

Dear Andrew,

The pot is actually a 1k pot, so r7L=half of 1k pot and r7r the other half.

But ... there is already 100 Ohms in series with it on each side and I was planning to put another 300 Ohms or so in series which wil the give me the desired gain and gain variation (about 3 db to each side if i'm not mistaken). Besides that there is a 100uF cap in series on each side of the pot.

Have a look at the datasheet and find the distortion and drive capability graphs for low load values

Well, easier said than done (at least for me). I'm not at all sure (or actually totally cueless) in determining the load. That will have something to do with the gain setting resistors, the output impedance and the attached poweramp

greetz, MArco

[richie00boy]

The noise issue of a 10k pot is so trivial that there are far more important things to address in the construction of the system. Just build the design as published, it's proven and effective.

[AndrewT]

Hi,
you predictions of gain variation due to balance control of 1k0 plus 400r to each side seem to be good.

I find +4db to -2db at balance extremes.

The output load on the opamp will be 1k0//cable capacitance//input impedance of next stage. This will be lowish particularly at high frequency.

The opamp will not perform at it's best into this kind of loading. Adding a series resistor of 100r to 200r will help slightly but personally I would look for a new balance pot that will fit, to replace your existing one. Try 5k0 to 22k

[deduikertjes]

Hi All,

It took me some time to work on it again, but after all I decided on this:

5K for the feedback resistor and 2K2 and 100 Ohms in series with the 1K balance pot. This gives me just enough balance control and a good gain.

Listened to it for a short while. Didn't notice a degrade in sound compared to no balance and a 10 k feedback resistor. But then, maybe I have to listen better.

Thank you all...

greetz, MArco