# Compare two DC voltage signals

Discussion in 'General Electronics Chat' started by Anestis88, Nov 3, 2014.

1. ### Anestis88 Thread Starter New Member

Aug 9, 2011
19
2
Hi
I would like to ask what is the best approach if I have to compare two dc signals with a lot of fluctuations on them. I have already checked the frequency spectrum and I see no high frequency noise. However the signal fluctuates around +- 5% of the mean reference.
So, is it enough if I compare the mean values of these signals or any other digital filtering technique is better ?

2. ### Alec_t AAC Fanatic!

Sep 17, 2013
5,968
1,135
It depends what the aim of the comparison is.

3. ### Anestis88 Thread Starter New Member

Aug 9, 2011
19
2
I want to compare which of the signal is higher

4. ### Alec_t AAC Fanatic!

Sep 17, 2013
5,968
1,135
How much higher? Averaged over what time period? Or instantaneous? Updated how often? What will the comparison result be used for?

5. ### Anestis88 Thread Starter New Member

Aug 9, 2011
19
2
The signals are around 2.5V and I want to find a difference in mV accuracy. For example 3-4mV difference . My sampling is 2.5Gs/s with 12bit resolution. The period can be 10-20 usec. And inside this period I want to compare absolute values to find the difference of these signals.

File size:
73.6 KB
Views:
23
6. ### wayneh Expert

Sep 9, 2010
12,374
3,226
So if I understand, you want to capture a burst of two signals, each carrying a lot of noise, and produce an output that is proportional to the difference in the mean voltages of the two signals?

I think it would make sense to integrate each signal to produce a pair of "moving average" voltages, and then amplify the ∆V of the smoothed series.

It would also be valid to compute a mean and standard deviation for a captured data set. This approach would also give information about the noise and the significance of any ∆V between the means.