www.classdaudio.tk

4.096v Ultralow Noise, High Accuracy Voltage Reference

Another sample from Analog devices was this nice reference, with 4.096v output. You really need to check datasheet for numbers, because they are really really good, just to name a few:
- Maximum temperature coefficient (TCVOUT): 2 ppm/°C
- Output noise (0.1 Hz to 10 Hz): Less than 1 μV p-p at VOUT of 2.048 V typical
- Initial output voltage error: ±0.02% (maximum)
just to name a few.

I wired it, on SOIC to DIP adapter board with 1uF output capacitor, nothing on input, just HP6205B set to about 10v.
Output was measured with HP34401A and logged to PC. Sample interval was 10s and I let the test run for little over 48h.
Nothing special, long wires, no shielding or anything, just left it on my workbench.

So one data point = 10s, there are about 18000 data points.
Took that data into excel and I get this:

Started loging just after I turned it on for the first time, graph shows output voltage vs samples.

Here is similar graph, but shows, how voltage drifted away from average value of above curve in ppm (parts per milion).

One thing you can see just after 9000th sample is when I turned the LED (switching) lights above bench on. That made impact on output (not great lights).

Excel data of ADR4540

ADR4540BRZ visits:

How to make it loud - real loud. (230 watts rms into 8 ohms without clipping from 4 ICs. How to match LM3886s for parallel operation, then bridge the matched parallel banks. If you are not familiar with bridging amplifiers (BTL configuration), read the Application Note on bridging listed left in article Application Note on bridging audio amplifiers.