**Part Number:**THS4524

**Other Parts Discussed in Thread:**ADS1278, , THS4552, THS4551, REF5025

I'm very new to designing with these components so I may be misreading some of these values or lacking some nuggets of basic info. Can someone please help me by showing me what I am missing? (Links to good references are always welcome)

I'm planning to use differential inputs. I see on pg 7, table 7.1 that the max differential voltage (Vid) is 1V. To me that implies that this is only as much as my Vin+ and Vin- can be apart (+/- 0.5v around whatever their common mode voltage is if they are symmetrically out of phase). Am I reading that value correctly?

Both in this data sheet (fig 85) and in the ADS1278 data sheet (figs 88,89), the THS452x series is demonstrated as a unity gain input buffer to the ADC. Yet, it was described (on the THS sheet) as tested to drive a signal that is -0.5dB below the ADC's full scale. On the ADS1278 sheet, its full scale (FSR, first table line, page 3 and footnote 1) we see "= 2Vref" and "+/-Vref". Above that table it lists "VREFP=2.5V VREFN=0V".

One way (for me) to read that is that Vref is just 2.5V (which would also be Vcom which is tied to Vocm on the opamp) and that AINP could be +5 while AINN would be 0 (a rail-to-rail signal at the ADC) which gives me a Full Scale Range reading. Did I tie all of that together correctly? Which part did I mess up?

So how can an opamp set to a unity gain with a maximum input differential of 1V push that ADC to within 0.5dB of its full scale which is at least 2.5V (possibly 5v)? If it's all in the math, I just don't see it. (Technically I believe the gain, as drawn, would be 1.05V/V but that's still fairly close to unity).

If indeed the full scale digital value the ADS1278 would emit is 0x7FFFFF for a +Vref differential analog input (and I'm still not sure what the max Vref could be), what number would it give me for a value that is 0.5dB below that?

And one more thing. The circuit in figure 79 is described as having a 0.27V/V gain. Shouldn't that be (270+40.2)/1000 = 0.31 ? If we aren't supposed to count the R in the RC low-pass filter as part of the feedback leg, why not?

Again, I apologize if these are overly simple questions. I'm coming at all of this without a degree in EE so I'm trying to learn by example. Thank you for your patience.

--

Shawn