Hey Harry,
Thanks for the reply. I waited a bit until I got my parts and dev board in so I could get some hands on testing and maybe a better understanding etc...
I've changed up my plan a little in hopes of eliminating an unneeded op-amp. I've attached a pdf of what I'm trying to do now. Same idea, but I'm trying to control the gain in the sallen key filter. I've got inputs to an adc that I'm trying to run through a low pass filter. The low pass filter I think I'd like to go with is the Sallen key. The filter seems to work ok on the bread board, but I'm still having some concerns about the gain. I read your recommendation tonight. Very good and I will be reading it a couple more times this week. For now though I wanted to ask you about my current attack plan and what I'm seeing.
Here is a schematic describing what I'm doing:
Image may be NSFW.
Clik here to view.
I have the AD4661 (the picture has a 4662 which is a typo) on a PCB and because of my lack of planning I had to break out the Vin+, Vin- and Vo pins to a bread board where I have the passives. Anyways, I'm shooting for around a 20kHz fc and this is close. I'm seeing some issues with the gain though and am not sure if it is my signal generator or the configuration. when I amplify a small signal ~200-500mV I can do 2x where i set the AD5272 to 10k and Ra is a TH 10k resistor. If I start to increase the 5272 with a uC the output starts to break down and get noise.
One thing I'm not sure about is that the signal generator only does Vpp so I have to add an offset. I'm wondering if this is causing it to reach Vcc early and clip. Either way it gets very noisy when it should be outputting a slightly increased magnitude of the input signal.
I'm mainly looking for some feedback here because I know most of this post is very vague. Not being a well versed analog guy yet I could really use any pointers. In reading the App note you pointed me to, the only thing I can think of is to add a cap to the feedback path where the AD5272 is to help clean up the feedback path a bit.
Long story short, does this look like a viable path forward? I'm looking to low pass filter 8 input channels to an ADC. I'd like to have a corner frequency of 20kHz because my max sampling speed in 8 channel mode is 50ksps. This would give me a little over 2x.
Thanks for the help,
Rob
