PGA4311 600ohm input source impedance

[DragonMaster]

OK, I found how to make an inverting buffer here:

http://hyperphysics.phy-astr.gsu.edu...ampvar.html#c2

I suppose that the 39kOhms in parallel can act as R1, so I can leave out R1 and put a 19kOhm Rf, right?

Also, to make the buffer's output 600 ohm, I just put a 600ohm resistor in series with the output, right?

[anatech]

Hi Antoine,

I suppose that the 39kOhms in parallel can act as R1, so I can leave out R1 and put a 19kOhm Rf, right?

For this to be an inverting buffer, R1 = Rf. Otherwise you will have gain. R1 also determines the input impedance as the input terminals are forced to be the same instantaneous voltage. That is why the inverting input is called a virtual ground when the non-inverting input is referenced to ground or some common AC reference point (As in a single supply version).

-Chris

[DragonMaster]

OK.

Oops, I meant a 19.5kOhm resistor, not 19kOhm.

Now about power rails, will powering the inverting buffers with ±5V cause problems since the S&H circuit uses ±6V?

I would personally feed the inv.buffer + PGA circuit with ±5V using a LM2941/2991 combo fed from the not-so-well filtered ±6V lines already available. The PICmicro and digital section of the PGA could use the digital voltage 5V line. I'm trying to go cheap as I might have to build this circuit for a few persons which might not want to spend lots of money on it.

[anatech]

Hi Antoine,
Do you want your input impedance to be 19.5 Kohm? This was my point. I wasn't nit picking on your values.

-Chris

[DragonMaster]

Well, there currently are two voltage followers with a 39kOhm output impedance in parallel. To match this, I use an Rf with half the value?

Oh, I see what you mean, I think... I'm trying to make the buffer's output 600 ohm to reduce noise and THD at the PGA level, but the buffer seeing an high impedance output actually causes noise too.

[anatech]

Hi Antoine,
Why not reduce those mixing resistors and run the entire thing at a lower impedance? Your numbers are correct. The only thing to watch for is loading down your inverting buffer with the feedback network.

There will be give and take, so try for 10K total and use 20 K resistors for your mixing section. Normally for current sharing I would go lower still. You could then use a 10K dip resistor where the tracking will be very good.

-Chris

[DragonMaster]

Well, there's that feedback problem. I have some experimenting to do...

Anyways... I think it's time to buy a breadboard!

[DragonMaster]

Hey, old topics...

I had an idea a few weeks ago that would solve the PGA input impedance vs. feedback problem : I could simply use two PGA4311s instead, and mix the signal pairs after them rather than adding an other stage, would that work and be a better solution than working on feedback resistors?

[veteran]

The best results I had with PGA2310/11/20 was using input preamplifier with 3-5V/V gain. This chip like low input impedance source. Don't link tho of them - just use input buffer/preamplifier stage

[DragonMaster]

Would the sample and hold circuit work well for the task?