Modification of QA403 for DC measurement

1

Hi,

for the integration into a scientific experiment we are considering
the QA403 due to its excellent specs.

Since the experiments requires to measure DC signals the question
is if removing/bridging the AC coupling filters on the inputs would be
a suitable hack or are other measures to be considered additionally.

Thank you and kind regards!
Peter

2

What DC precision are you looking for? Beware that audio DACs/ADCs may have poor absolute precision (or undocumented even), and don’t usually use references like a general purpose precision device.

0.1dB error would be good for an audio device, but that’s an absolute error of 1.15%

The ADC (ES9822PRO) in the QA403 has a worst case left/right gain mismatch of +/-0.4dB (nearly 5%), for instance. And I doubt the gain is calibrated either.

Then you have to factor in things like gain drift with time / temperature (not specified), and that there maybe DC-blocking digital filters internal to the device…

I suspect its not the tool for the job if precision is needed, not just linearity.

3

Hi Mark,

thank you for the quick reply!

Absolute DC precision is actually not relevant. More specifically the goal
is to acquire noise spectra with lower starting frequency than the ~1 Hz.
If a hard DC blocking might cause problems reconfiguring the AC filter capacitor to reach a cut-off frequency of < 0.1 Hz might be an alternative.

I will check the data sheet of the ES9822PRO if DC-blocking digital filters
are implemented.

Thanks again
Peter

4

Hi @Peter_T, in addition to what MarkT says, there are some peculiarities with the ADC around 0V where spurious phantom tones can be introduced as you get closer to exactly 0V on your input offset. These are low level tones, but still there. If you don’t need the -110 THDN and could get by with -90 or -100, you might look at a generic DC-coupled capture card from National Instruments instead.

5

Thanks for the heads up regarding the phantom tones!

Since harmonic analysis next to the noise spectra is also
of interest to us the high THDN of the QA403 is actually a
big plus.

6

Thanks @Peter_T. I think you could just bypass the caps and try. If it were me, I’d not try to remove the surface mount caps. But I’d tack on to the BNC connectors on the bottom of the board and then fly a wire up to the analyzer side of the BNC. That is easy to solder to AND if you wanted to remove after eval it’d be easy.

Note the caps play a role in circuit protection. There are short term protections that come from some active electronics, but the longer-term protections are provided by the DC block of the cap. And so, with the caps bypassed, I’d limit the input voltage to +/-8Vrms or so when the attenuator is off.