Hello ADI support. I am with 2 customers using ADE9000 evaluation board and they are targeting the ½ rms voltage measurement feature. They work on different markets but the need is similar.
Regarding this, I would like to check 2 main topics:
For the ½ rms cycle RMS Table 18 of the technical reference manual indicates that the settling time is 60 ms, One of the customer has performed some tests with a relay test setup and noticed that the convergence actually occurs at this time, however the signal behaves in an oscillatory way. They noticed that such an oscillation does not come from the test setup, but from the ADE itself.
Their doubts are as follows:
- They obtained an oscillatory behavior during this transient of 60ms, after that the signal is very stable. Is this the expected behavior for the component?
- Is there any way to get the part-time measurement without going through additional filtering, which adds 60ms? (for a 50Hz signal then the settling time would be 10ms)
Attached is a step test of a signal of 110V-> 120V (@ 50Hz), with duration in each 200ms. It is observed that the signal converges in 60ms, but in an oscillatory way.
The pictures refer to the RMS_SRC_SEL = 0 and =1. We would expect faster settling times between those but it was not observed.
Important note: The readings are being synchronized with the 0 crossing.
1. How do you swithc from 110V to 120V? if you use a switch, the line period value need 2 or 3 line cycle to settle down and 1/2 RMS needs extra time. Here this is translated in an oscillating behaviour but it can be improved with some pre-filtering of the signal and/or the internal sinc4+lpf
2. Sorry but you can not have part-time measurement. You can remove the intergator and the hpf but not the additional component reported in the channel datapath (figure 56, 57, and 58 in the DS) since signal will be too noise and the measurement inaccurate
3. Indeed, high-pass filter has settling time associated with it but its effect on the RMS1/2 computation are more relevant on the final error (%)
Thank you Valeria.