This produces that ~20 bit noise free bits every 10 us (if it is anything like the 2376, that I've thoroughly tested)...
Once you actually use a 250x part, you're not really going to see that, honestly. Maybe 18 or 19 bits. Probably like 2376. But your Vref is modulating your input data, and at low ppm's that becomes more and more critical. As your signal increases dynamic range then the Vref noise becomes
even more critical. Then if you're running these 250x fast with an FPGA good luck keeping the that digital noise out of the faster measure (if you're chasing PPMs)...Uggh. It can be done, but you have to be careful.
Speaking of noise: Also notice that the AC specs (SNR, THD) of the '250x parts are pretty lackluster. That's more like a 16 or 18 bitter.
An '2400 / 2408 has 4ppm INL but it is normally corrected out in software, and has been since the 80's. Linear even gives you the code for it, it's not like you have to re-invent the wheel. You should get to ppm absolute or sub PPM relative measures on every part in a production run. Very repeatable and stable correction curve is very similar even across parts - and is stable over time (decades). We just use a 10 point test cal correction during manufacture and it's quick and easy. You can get to ppm (or 20 bits) relatively easily, and if you keep your system very very quiet you can get into the 21~22 bit range. No good way to correct a '2500 / '2508 long term, they seem to bounce around every 1kHrs or so. We've been looking at them for the past year or so and feeding suggestions back to LT, but they seem a bit confused these days. This seems to be a part that marketing wanted to have a datasheet for to compete with AD, but now they are the same company. The will probably drop either the LT or the AD 32-bit part at some point, no need for two parts competeing against each other.
Again: These '2500 serires
are not meant to be a 32 bit, sub-PPM absolute measure device, and LinearT is the first to tell you that. Use these when you need to measure a high-resolution (not necessarily with high absolute accuracy) ratiometric sensor. Use a '2400 / '2404 / '2408 with a 6655 or LTZ's for accurate DC measuring at a MUCH better profit margin.
Zhtoor: Metrology and "poor man's" anything are generally mutually exclusive. You really need to understand that in the ppm world you get exactly what you pay for - and as you get below 10ppm the cost goes up more than exponentially every ppm below that. By the time you're taking accurate measures under 8~5ppm absolute requires more than a single 3458a... And by the time you're at real, calibrated 2ppm (the realistic limit if you're not a jjarray-refered cal lab) you've got a several 3458a's & 732bs or equivalent, with some thousands of $$$ per year in cal services alone. IF you're doing it right. And at that point you're talking a solid, noise free 19 bits measure no matter how you slice it. To get to a real, calibrated 20 bit accurate absolute voltage measure (1 ppm accuracy) takes an extreme level of measurement ability, and really talking about a JJ-Array reference at that point - or a bank of 732b' refe'd to a JJ-array at regular intervals.
The other thing to understand is that very few applications require ppm-level anything unless it's VoltNuts entertainment or usually trying to calibrate / test something else. That means a very small market with no real profit center, and that means very expensive parts. We use ppm-level test jigs on a laser diode semiconductor process line where each device in production sells for over $500 - but I can't think of a single instance where I'd need that in daily life at home.