Power rail noise calculation

I am attempting to calculate the noise on the power rail that my ADC is receiving so I can calculate my effective number of bits. But the different units on the datasheets of the regulators are throwing me off. I don't understand the decibel unit in noise suppression. I clearly understand and visualize volt peak to peak and frequency.

So the ADC rail pass thought several regulators and filters.

Power plug (5v/3v3) 🡒 16 Hz RC Low Pass 🡒 16 kHz Low Pass 🡒 LP5097 🡒 ISOW7841 🡒 ADC

That is basically the chain for the power to get to the ADC. I do not know the units of the final answer to look into the data sheet of my ADC to see if it's acceptable.

The ISO7841 is the main cause of the noise, it specifies a typical output ripple of 100mV pk-pk, and its output frequency shifts with load, looks like in the ballpark of 30-100KHz,

ok, so your ADC following that has about a PSRR of about 80 decibels, so 0.1V * 0.0001 (-80 decibels ratio) = 0.00001V on the output, or 10uV pk-pk noise.

So your low pass's are in the wrong places, filter down the noise after the isolator, as that is where most of it is being introduced.

Assuming your using the internal voltage reference, and I don't go into the references own noise, your best case would be 17 bits

• You should not, providing you use the recommended decoupling capacitor size or larger, and that you have enough voltage headroom for the LDO to accomplish its role (the ripple + the ldo dropout) so you need atleast 350mV more than what you want to output – Reroute May 24 '20 at 9:57
• The ISO7841 is creating that noise, if you fitted your filter after it, then the attenuation would depend on what the corner frequency of that filter is, Isolated power converters are naturally noisy – Reroute May 24 '20 at 10:14
• filters reduce the amplitude by 20 decibels per decade, so as a rule of thumb, a 16KHz filter would in the worst case only be reducing it by a fraction of that, while a 16Hz filter it going to be somewhere around 60 decibels of attenuation – Reroute May 24 '20 at 10:25
• they sum up, that 80db is how much it reduces any supply noise, your filter also reduces it by X amount. sp 0.1V * 0.001 (your filter) * 0.0001 (PSRR) = 10nV pk-pk, your ADC max resolution would be roughly 80nV per code, so that would mean its reduced to a suitable level, site is sengpielaudio.com/calculator-gainloss.htm – Reroute May 24 '20 at 12:13
• The isolator attenuates input noise by roughly 54db, at which point, I suspect the noise produced on the output is still the dominant noise source, but you can run that through the calculation if you want, (line regulation), technically load regulation is also in play, but I still feel the added output noise is dominant. – Reroute May 24 '20 at 12:31

in the ADC there will be analog comparators, and opamps and probably DACs.

At really high frequencies, such as the 1 nanosecond RISETIME transients of MCU clocks coupling 1cm away into your ADC VDD, VREF, VIN(+) AND VIN(-), THE PSRR of circuits within the ADC becomes ???? zero??

Thus 10 quanta on trash coupled onto Vref ?may? cause a 10 quanta measurement error; probably not, because of the massive averaging within the ADC digital engine.

So let us be cautious, and predict te worst-case cleanliness of that VDD.

Assume we are willing to budget 1quanta of trash to the super_clean_VDD_rail. What should that be?

Divide 3.3 volts by 2^24 (or maybe 2^23) or 3.3 / 16,000,000 === 3.3/16 microvolts pp. Or about 0.2 microVolts.

What can we say here?

==================================

Once you have (at some expense, in cost and PCB area) created an ULTRA_CLEAN VDD, you need to treat that CleanVDD as a precious resource, to be protected from aggressor trash injectors.

• avoid ground currents flowing across or along the PCB between VDD filtering and the ADC power pin; 2 milliAmps of (varying) ground current, passing thru 2 squares of standard_thickness 1 ounce/square foot foil that is 0.000500 ohms per square, causes a VARYING 1 microVolt trash injection into VDD;

• avoid nearby e_field transients ---- example needed (planes, gnd or vdd, are great attractors of flux lines, causing 1/distance^3 [cubed] attenuation

• avoid magnetic fields of all sorts ---- expect to need some steel sheets, folded to wrap around your 24-bit system ---- example needed (copper foil has little benefit --- just a dB? --- for edges slower than 100 nanoSeconds)

• power supply interference --- you have already become aware of this, and the other answers are very good guidance

• cabling

• right, i should be able to clean as close to 0.2 microvolt. But i do not know the trash noise in the rail. Should i only look at the ISOW7841 output noise? (90-100mV pk-pk) ? and ignore the noise of the resT? – Jake quin May 24 '20 at 9:58