How much input impedance is enough?

[Russ White]

I have been experimenting by building several different preamps for my chip amps. I have a few questions for the gurus here.

I have studied several designs and note that most non-inverting opamp based preamps and headphone amps use an input inpedance of 100K by using a resistor from input to ground(R-in) after the input capicitor if used (C-in), while a few other use 47K or even 22K values for R-in. My question is how does the input impedence effect the signal? Is a lower inpedance better or a higher one? I know that the value of this resistor effects the high pass corner frequency of the RC filter created by R-in and C-in so the value of C-in would have to change with R-in. Besides that is there some other effect?

Also, I generally use standard metal film 1% resistors for the resistor from input to ground, is there something better?

I have been using 1uf non-polar electrolytics for my input caps. Is this a poor choice? If so, what is ideal?

Does the input impedence presented by R-in effect the choice of the value of a volume pot? I have been using 10K pots with 100K R-in with good success but is that optimal? Are there some guidlines for choosing the pot value?

Thanks for helping a fledgling.

[Russ White]

http://www.whirlwindusa.com/tech03.html

Any ideas from the gallery still entertained.

[GregGC]

Most CD/DVD players would have Zout around 500 Ohms to 1k. So, Zin of 22k is high enough.

Greg

[Russ White]

Like around 1Mohm? Is there any down side to very high input impedance? Is there any gain to keeping it as low as possible? I see 22K on a lot of chipamps but as much as 1M on headphone amps. Why the difference?

[GregGC]

Originally posted by Russ White
Like around 1Mohm? Is there any down side to very high input impedance? Is there any gain to keeping it as low as possible? I see 22K on a lot of chipamps but as much as 1M on headphone amps. Why the difference?

The higher the input impedance the more chance for RF interference. In the other hand you don't gain anything by going higher than necessary. It makes sense if you have a Valve preamp with significantly higher Zout, though still going higher than 100k is overkill.

Greg

[theChris]

the obvious benefit to a high Zin is in the RC highpass input to most amplifiers. if R is very high then C can be very low. with a 100k input impedance, C can be made 1uF, a value that you can find film capacitors in fairly cheaply. if 10k is used, you may try using a capacitor that is 10uF, one which a film capacitor is impractical for.

the downsides is that the thermal noise of most resistor series will begin to increase near 100k. also the PCB must be made well enough so that the high imedance is maintianed. capacitive coupling may be an issue as well. an interesting thing about high input imedances is that you should not touch the resistor when the circuit is in use. it not dangerous to you (except in tube circuits where the lead you touch may be at 500V or more)''

as for the components:
input caps. for many designs, electrolytics will work, however they are generally not optimal. in power supply uses they have high ESR/ESL. they also have fairly poor tolerance and some issues with changes over their lifetime. Most designers try to use some type of film (film and foil or metalized film). for cheap 1uF caps, try goldmine-elec.com they have some small 1uF mylars 10 for $1.

resistors. it debatable. i use metalized resistors myself, but some people have good luck with either.

note that with potentiometers, Rin will be seen in parallel with part of the potentiometer. if Rin is much greater (say 10 times) the resistance of the potentiometer then it is pretty much out of the picture. however if the resistance is closer to the value of the potentiometer, you will affect the resposne of the potentiometer. see sound.westhost.com for an article regarding fake-law pots. note that this effect can be used in a positive manner.