# change the DC bias of an AC signal ?

#### ralphnev

Joined Jul 8, 2018
3
i have a fairly fast (~5Mhz) signal that is ~5-8VDC
& reading this signal using a ADC that takes an input from 1-4VDC

looking for solutions to shift the bias with little loss of fidelity of the signal?
(for testing i had used a simple resistor divider but loose 1/2 the signal ...)

#### ScottWang

Joined Aug 23, 2012
6,882
If your reading from ADC is 1/2 of the input signal then you need to double the input values (multiple *2) or shift to left for one bit with assembly.

#### ebp

Joined Feb 8, 2018
2,332
[EDIT] reread your question and now I'm not clear on what ~5-8VDC means. For the following I took it to mean that your signal swings between approximately +5 V and +8 V relative to circuit common ("ground"), so what you need to do is preserve the 3 volts peak-to-peak amplitude and just sift it 4 volts negative.

You would typically do this with a differential amplifier with a gain of 1 and an offset of -4 volts, The signal goes into the non-inverting input and (functionally) +4 V goes into the inverting input. Given your ADC input span, this should be possible with an op amp with a single 5 volt supply since you don't need the amp to work very close to the supply rails. You would want to use good quality resistors. I would recommend 0.1% tolerance with 25 ppm/°C temperature coefficient if you need good stability and accuracy, but ordinary 1% types might meet your needs. You could likely use the voltage reference for the ADC for the offsetting voltage.

e.g. (I think this is right - have to run off to so something & haven't checked properly)
say 5V reference
- reference through 1k to inverting input
- 800 ohm feedback
- signal through 444 ohms to non-inverting input (NII)
- 556 ohms from NII to common (ground)

gain of -0.8 for +5 V reference, so -4 V offset
non-inverting gain is.556/(444+556) * 1.8 = 1
Resistor values are for arithmetic simplicity - ratios are what is important & there may not be "real" resistors in these values

You can find lots of info on single op amp differential amps on the web.

#### ralphnev

Joined Jul 8, 2018
3
[EDIT] reread your question and now I'm not clear on what ~5-8VDC means. For the following I took it to mean that your signal swings between approximately +5 V and +8 V relative to circuit common ("ground"), so what you need to do is preserve the 3 volts peak-to-peak amplitude and just sift it 4 volts negative.

You would typically do this with a differential amplifier with a gain of 1 and an offset of -4 volts, The signal goes into the non-inverting input and (functionally) +4 V goes into the inverting input. Given your ADC input span, this should be possible with an op amp with a single 5 volt supply since you don't need the amp to work very close to the supply rails. You would want to use good quality resistors. I would recommend 0.1% tolerance with 25 ppm/°C temperature coefficient if you need good stability and accuracy, but ordinary 1% types might meet your needs. You could likely use the voltage reference for the ADC for the offsetting voltage.

e.g. (I think this is right - have to run off to so something & haven't checked properly)
say 5V reference
- reference through 1k to inverting input
- 800 ohm feedback
- signal through 444 ohms to non-inverting input (NII)
- 556 ohms from NII to common (ground)

gain of -0.8 for +5 V reference, so -4 V offset
non-inverting gain is.556/(444+556) * 1.8 = 1
Resistor values are for arithmetic simplicity - ratios are what is important & there may not be "real" resistors in these values

You can find lots of info on single op amp differential amps on the web.
thank you / I'm mostly a digital guy / haven't worked with opamps in may years ...
but i do understand above - now component choices ....
is a video opamp appropriate ? ala AD8013 ?

Last edited by a moderator: