Remove DC offset of pulsed signal

Status
Not open for further replies.

Bengt

Junior Member level 1
Joined
Aug 19, 2004
Messages
15
Helped
4
Reputation
8
Reaction score
4
Trophy points
1,283
Location
Sweden
Activity points
151
remove the dc level

I need to remove the DC offset of a signal that consist of small pulses as a result of sampling. The problem is that the circuit must be very accurate as the resulting signal shall be amplified several undered of times afterwards and that the circuit must be very small as there are very limitid space.

I have done it using feedback from a microcontroller and a DA but the resolution is a problem and an anlog solution might be to prefere. I fugured to use a peak detector and remove the top that way but I have accuracy problem.
 

circuit to remove dc level

What is the frequency of your pulses?
What is the DC voltage level in your signal

The solution can be as simple as using a High pass filter at input
You can also use the difference amplifier (simple configuration using the DC level as reference at the inverting terminal) if the DC level in the pulses is known and constant
If DC level is not known then you can use a low pass filter at inverting terminals such that it will eleminate the pulses from the input and only DC voltage will be avialable and you can use the same difference amplifier to amplify you signal

if you can let us know the frequency and DC voltage then it will be easy to provide a solution
 
remove dc from signal

The DC level is from 0V to 3V. The frequency depends on the sampling rate and can be from minimum 100 Hz to 10 kHz. The sampling rate is however tuned for a specific application and can be fixed somethere in this interval. The system will be designed for a maximum sampling rate of about 10 kHz but if slower signals are sampled a much slower sample rate is good as it reduce the data.

A high pass filter will not work as the signal in it self have no DC. I understand now that I was very unclear in my description. The signal is small sampling spikes there the actual signal is fluctruations of the level of theese spikes. I need to take som sort of mean value of theese fluctruations and bring this down to 0V to be able to amplifiy them prior to feeding the signal into an AD converter. If the signal change very rapidly, this will of course result in that the signal temporary gets out of range but I am aware of that. There is often methods for forcing the signal fats back into range in such surcumstances.
 

Status
Not open for further replies.

Similar threads

Cookies are required to use this site. You must accept them to continue using the site. Learn more…