Preamp Input Selector & Buffers

[Devil_H@ck]

Hi,

I am brainstorming about what parts to use in a preamp, but I don't know very much about all this, so I guess I'll best ask some things here.

What should I use for input selection? Price is not a problem, since I will only need a few chips.
1. MPC506/ADG506/ADG406: All 16 input muxes, I am really clueless as to which one to pick, anybody has an idea what's the best of those 3 chips?
2. SSM2404: 4 SPST quad audio switch
Datasheet: SSM2404 - ADG406 - ADG506 - MPC506
The SSM2404 is especially made for audio, but it seems so are the '506/'406 ICs.

Next, a question about buffers. I will be using a PGA2310 as a volume control. Were should I place buffers? I plan on putting a DRV134 followed by 2 BUF634's for each channel to allow for XLR output. However I'm not sure about what to put between the input selection and the PGA2310. I want to have a tape loop in there to. I was thinking something like:

Input - BUF634 (tape loop)
       `- OPA627 - PGA2310

Would this be ok? Also, what kind of topology should I use for the buffer/opamp after the input selection, non-inverting or inverting?

I know it's a lot, but if you feel like responding, please do, it would help me a lot .

[Nuuk]

You will be dead within about 80 years!

The point is why make things so difficult for yourself? A simple rotaty switch for source selection, a stepped attenuator and a simple opamp buffer will do the job and allow you much more time to enjoy the music; and what's left of your life!

[Hybrid fourdoor]

As far as CMOS switches. Look at the R(on) [Resistance On]. Some of the generic switches have R(on)'s in the 100's Ohms or 200 ohms's where as some of the dedicated audio CMOS switches from like Dallas have as low as 1 ohm. Now I can't vouch personally to what 100 ohms added to 47K of input impedence will do, but the companies seem to make a big deal about it.


Theres always something like this, also a good read


Also remember that although CMOS switches are cheap (especially sample ones), you can never go above the Ve of the chip (5V), and you can only have so much current. However you get reaction times in the nanoseconds. Where as relays are more expensive, but are pretty much tanks. All in all I think the CMOS chips are a good alternative to relays and switches.

[adx]

Originally posted by Hybrid fourdoor
...Now I can't vouch personally to what 100 ohms added to 47K of input impedence will do, but the companies seem to make a big deal about it.

The big problemo relates to the flatness of the resistance (as signal voltage varies). A switch which varies 10 ohms with signal voltage while driving a 10k load with introduce distortion in the region of 0.1% (=10/10000). I don't know how audible this is, as it will reduce as signal level decreases.

[AuroraB]

The SSM series switches are specifically designed for audio, with extremely linear Ron vs. signal amplitude. Rated distortion is somewhere around 0.000x %. They are also rated for operation at +/- 15V. I don't know about the others.
One way to solve the Ron non-linearities is to use the switch in series with the resistors in the inverting input of an opamp. If you need a buffer amp anyway, this is a good solution.
I have used this in source selectors for sound mixers with good results.

[Devil_H@ck]

Originally posted by Nuuk
You will be dead within about 80 years!

The point is why make things so difficult for yourself? A simple rotaty switch for source selection, a stepped attenuator and a simple opamp buffer will do the job and allow you much more time to enjoy the music; and what's left of your life!

I want to do it this way . Also decend stepped attenuator will cost me a LOT more than a PGA2310. Let's just say I like challenges .

Originally posted by AuroraB
The SSM series switches are specifically designed for audio....

Hm, but so are the others. They are all rated at +-15V. Lower Ron is better, since it will cause less voltage swing when in series with resistors to the inverting input of an opamp, right?

Then I only have to worry about the buffers. The output from the switch to the inverting input of an opamp is the best choice?
Is my "schematic" OK?

Input - OPA627 - BUF634 (tape loop)
     `- OPA627 - PGA2310

[Nuuk]

Let's just say I like challenges

No arguments with that. Good luck in achieving your goal!

[Devil_H@ck]

Anyone?

Then I only have to worry about the buffers. The output from the switch to the inverting input of an opamp is the best choice?
Is my "schematic" OK?

Input - OPA627 - BUF634 (tape loop)
     `- OPA627 - PGA2310

[head_spaz]

QUOTE]I plan on putting a DRV134 followed by 2 BUF634's for each channel to allow for XLR output.[/QUOTE]

WHY would you wanna do this?
The DRV134 makes a fairly decent output buffer as it is, and it's made specifically FOR providing a balanced output to XLR connections.
You shouldn't need any output buffers in this situation !!!

[Devil_H@ck]

Originally posted by head_spaz
WHY would you wanna do this?
The DRV134 makes a fairly decent output buffer as it is, and it's made specifically FOR providing a balanced output to XLR connections.
You shouldn't need any output buffers in this situation !!!

Hm, OK. I though it was better to have another buffer behind it, but indeed now that I read the datasheet I noticed it isn't necessary at all.
One thing I'm not sure of though. In the DRV134 datasheet, it says that the input impedance it 10k ohm, does this mean that I don't need to put another resistor in front?
It also says that max input current is +-1mA. Is there any way to limit input current so that I don't overload the chip?
Oh, and should I worry about the 2x gain? Or is this normal for balanced applications?

Those are probably very basic questions, but I don't know to much of all this.