I would like to use the LTC6090 as a high voltage unity gain buffer into a voltage divider ranging circuit driving the front end to a 24-bit SAR ADC. I am wondering if I am going to have noise issues using this topology versus directly dividing the voltage and the buffering it prior my ADC front end?
Tony,
I like to get the signal out of the noise and then divide.
You have another problem: The ADA4637 is only stable with a noise gain greater than five.
Harry
In my actual design I am using the ADA4898-2 which wasn't available in LTSpice so I just dropped in the 4637 to get things running. Not sure what you mean about getting the signal out of the noise before dividing.
Hi Tonyraca and Harry,
To make it easier, I've imported the ADA4898-2 pspice model into LTspice. I've also simulated your single ended to differential converter circuit section and it seems fine for the speeds you seem to be looking at (200Hz sine wave).
At higher speeds, the time delay between the stages will cause the differential output (IN+ and IN-) to deviate from each other and distort the differential output. But, there is no instability in simulation. I've reduced the feedback cap on your 2nd stage because that aggravated the deviation I've mentioned above.
ada4898.zip
Regards,
Hooman
Let's say that the first stage op amp with resistors has a noise of xx uV. If your divided signal is less than
xx, you won't see it. If you gain it up first, then you can divide.