Ideal feedback resistor values

[CBS240]

Hi

I agree with sawreyrw, if the values are reasonable, there would be no difference in output Z. The global feedback resistors are just a voltage divider after all. But there are some capacitive components, whether inside a device as part of a transistor's AC model, or as an external component in the circuit that may pose limitations. The values of the resistors don't matter much to the closed loop gain as long as the ratio is the same. However a higher impeadance will be more suseptable to instability and frequency limitations. Also the circuit will be more sensitive to noise. The 'perfect' value would be determained partly by the input components used. But, IMO, to use the lowest possible input impeadance is the best solution, as determained by the application, i e source inpeadance or DC blocking cap/servo, ect.

 

[Leolabs]

Usually a few kohms will do,without loading the OPAMP and generating resistor noise due high ohm value.

 

[hags]

Conrad, please explain.

As for sources and RF caps, I like everything defined. It always drove me crazy to see people adjust a gain stage by adjusting the feedback. BW, slew, noise and who-knows-what-else changes. Thus, I don't like things that vary my offset and BW with gain or different input sources. I suppose it can't be helped at some level, but the more obvious variables should be minimized.

I'm not sure I follow you as it pertains to line amps and op amp line amps in particular.
I'm not using RF caps or any kind of low pass filter, is this necessary?
My input sources see the impedance of the volume control which is a constant value.
I just need a little insight into your reasoning.

 

[Conrad Hoffman]

My comment stems from what I've seen a lot of engineers do at places I've worked, not so much with audio stuff. I'm pretty sure we all agree that voltage amplifier stages should be fixed gain and bandwidth, and the conventional volume control used as a divider, should feed the input. Your source sees a constant load (the volume control), but the following stage sees a varying source impedance, depending on the setting of the volume control. That's where I apply the "at some level" escape clause. Everything can't be unchanging unless you design for one setting (choose it wisely?), and I don't think the volume control is a significant problem. I don't put rf caps after it, as I want them at the very input to the box. I do worry about some series resistance to the control, as I've seen some opamps go unstable and oscillate when the input is tied directly to ground, as it might be at zero volume. It's possible to build a constant impedance control for both input and output, but I can't imagine anyone thinking it's necessary or going to that amount of trouble.

 

[hags]

My comment stems from what I've seen a lot of engineers do at places I've worked, not so much with audio stuff. I'm pretty sure we all agree that voltage amplifier stages should be fixed gain and bandwidth, and the conventional volume control used as a divider, should feed the input. Your source sees a constant load (the volume control), but the following stage sees a varying source impedance, depending on the setting of the volume control. That's where I apply the "at some level" escape clause. Everything can't be unchanging unless you design for one setting (choose it wisely?), and I don't think the volume control is a significant problem. I don't put rf caps after it, as I want them at the very input to the box. I do worry about some series resistance to the control, as I've seen some opamps go unstable and oscillate when the input is tied directly to ground, as it might be at zero volume. It's possible to build a constant impedance control for both input and output, but I can't imagine anyone thinking it's necessary or going to that amount of trouble.

Thanks for the clarification. Is RF consideration important with high bandwidth op amps used in audio?
Is this something I should take precautions for?

 

[Conrad Hoffman]

IMO, the big worry with high bandwidth opamps is good bypassing and correct single point grounding. The bypassing should be located as close to the opamp as possible. That way you won't have RF generators inside your amp! Be sure to take care of the power supply rectifiers too. My belief is that external RF sources have to be stopped at three places- input jacks, output jacks, and the power cord. I'm not saying I put RF suppression on all of them every time (heck I'm not even saying everything I build even has a chassis!), but people have had problems with all three. It depends on where you live- how close are the RF threats? When CB radios were more popular, I'd get full volume "breaker breaker" audio if they transmitted on the road in front of the house. Once this stuff gets inside the box, it's hard to do much about it, so suppression is typically done right at the connectors and terminated to the chassis. By the time you get to the board, unless you're using board mounted connectors, it's too late. I also recommend trying a commercial EMI filter on the line cord to kill RF, but I've yet to fully resolve low level hum problems when everything has a 3-wire cord, so the best practice there is a bit uncertain.