Timeline for Are these low frequency FFT components due to insufficient op-amp slew rates?
Current License: CC BY-SA 4.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 7 at 13:54 | comment | added | Maxim | Let us continue this discussion in chat. | |
| Jan 6 at 16:52 | comment | added | bobflux | OK with this chip you won't have aliasing problems, so the 1st order LP input filter should be fine | |
| Jan 6 at 16:47 | comment | added | Maxim | I use ADC AD4134. Before the Input Buffer there is a 1st order LP in my real measurements (not in the ones I posted, as, there, I inject the artificial signal directly into the Input Buffer). I will update my question with the digitized ADC output soon. I just need to fire up Logic Analyzer. | |
| Jan 6 at 16:08 | comment | added | bobflux | Can you give info about what ADC chip you're using? Also if your analog signal chain includes analog filtering or not. | |
| Jan 6 at 16:06 | comment | added | bobflux | Can you post a plot of the digitized signal from the ADC? Most likely the peaks will be smoothed by the filter, as the high frequency components that make the waveforms sharp will be removed. But as the others said, the peaks have frequency components well inside your filter's passband, so these will not be removed. | |
| Jan 6 at 15:52 | comment | added | Maxim | To add on this: I am not interested in the energy of the peaks. Would it then be reasonable (as a kind of filter) to have a lower slew rate for the input buffer (I am only interested in signals up to 10 kHz -> slew rate of approx. 4 V / 25 us = 0.16 V/us). The Corona discharge of the peaks is above this threshold. | |
| Jan 6 at 15:34 | comment | added | Maxim | Thank you! That was educational. The full truth is that in reality I also see FFT components between 100 Hz and 200 Hz (or more) but usually peaking at approx. 150 Hz. The Corona signal above was constructed/reverse-engineered to see a peak in 150 Hz (as I do in reality) but tbh I do not know how the input looks in reality. However, I cannot wrap my head around, why the ADC does not filter out these high frequency peaks (I added its filter response above). I also added new measurements above. | |
| Jan 6 at 14:51 | history | edited | bobflux | CC BY-SA 4.0 |
edited body
|
| Jan 6 at 12:33 | history | answered | bobflux | CC BY-SA 4.0 |