8.5 digit dvm

Wiki

"Open Source 8.5 Digit Voltmeter from CERN: Build and Test" -

It is accurate to 23 bits in 10 Hz bandwidth, which is 7 decimal digits. On an 8.5 digit DVM, the last digit is completely random :-)

Worse than 'accurate', that's the "effective resolution".

So maybe 10x worse than a world-class COTS instrument. Not bad for something thrown together.

Of course it can't be any better than the AD7177-2 chip that they are using. AD claims "24.6 noise free bits at 5 SPS" which is 7.4 digits, and at a leisurely 5 SPS, so no more than 2.5Hz bandwidth.

You also have to worry about linearity too if you care about such things. It isn't too hard to have 24bits all monotonically increasing but unless it is compensated out they can sag a bit in the mid range.