Attenuation and signal conditioning circuit

This schematic is a complete disaster.

How about 74HC4052 ?

I am afraid everything should be done different way.

How much of noise and distortion exactly? How little of DC?

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

I expected this kind of response. Can you write what exactly is a bad idea? That will be more didactic for me :)

Are you able to show me direction? I have no idea.

About 90dBc And DC as small as possible.

This is going to be business. My email is at the website.

I.e. THD = 0.003%. Wow. This is not going to be very simple.

Let's put it this way: how much are you willing to spend to have low DC?

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Dnia Sun, 18 Nov 2007 16:32:50 -0600, Vladimir Vassilevsky napisa³(a):

THD and noise is more important than low DC

Excellent suggestion. A nice low-capacitance MOSFET switch.

Noir, you looked at the irf7455, and thought, what a nice switch, only 0.007 ohms of ON resistance. But, that nice low ON resistance comes at the expense of a very large die, a large part that has very high capacitance. In this case Coss = 870pF at 25V, or 3200pF at 1V as we see from fig 5 in the datasheet plots, and even more at 0 volts! So this nice MOSFET can never be OFF at higher frequencies, but ON and lowering the gain, shorting out signals. A bad choice!

Dnia Sun, 18 Nov 2007 18:13:27 -0800 (PST), Winfield napisa³(a):

It's all true, thanks for response.

Have you considered using a MDAC as a programable attenuator? For the differential amp, you need to match drive impedance at each leg. [I don't do discrete designs, so I can't spit out part numbers without doing research.]

Regarding the other posts, this is a place to learn, not solicit business. If someone posts a RFQ, that's a different story.

Dnia Sun, 18 Nov 2007 23:15:39 -0800 (PST), snipped-for-privacy@sushi.com napisa³(a):

If I good understand: using MDAC to controling opamp gain by switching internal resistors?

And how about noise and distortion in this case?

Thats good because I want to learn and that what I'm doing is not a comercial project. I want to do this myself I only looking for information, opinion

I only have first hand knowledge of charge based circuits on chip, i.e. not discrete solutions. But I recall that you could take a MDAC and apply the signal that you want to attenuate as the reference, then change the code to perform attenuation. Overkill, but simple, which is good for homebrew stuff. You should be able to find a circuit online with a google search. There are also "digital potentiometers" as chips.

Getting back to the MDAC, meditate on it a bit. The M in MDAC stands for multiplying. The part takes the reference voltage and multiplies it by a digital code, producing an output. Say the reference is a volt, the code is half scale, so you get half a volt out. Now replace the input to the MDAC with the audio signal. Then the audio gets cut by half. You need to research the "quadrants" that the MDAC can handle, i.e. you want the reference to be able to go negative, since the reference is really your audio signal. Your digital code should be postive (one quadrant), since you only want to attenuate and not invert.

Poke around the analog.com website and I'm sure this will be on an app note. It's a very old scheme. I really surprised the board level engineers on the list haven't suggest this.