This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
Hi,
I need to measure a single ended signal with an ADC (14 or 16 bits) at 100MSPS (ideally powered in 3.3V but not compulsory). My signal can vary between 0V and 10V and I need to measure low level (different to 0V) and high level.
At this speed, I find only ADC with differential inputs like ADS5482.
One solution seems to connect INM to VCM, and my signal to INP. For that, I must divide my signal and add an offset to be in the range VCM +/- xx V (0.75V for the ADS5482).
With this solution, the input signal seen by the ADC at his INP input will be an unbalanced signal (could be VCM - 0.75V for low level and VCM for high level).
What will be the behavior of the ADC with an unbalaced input?
Is there a better solution to measure such a signal (single ended with a measure of low and high level)?
Thanks
Matthieu Baque
Hi,
The ADS5482 requires a balanced fully differential input, as does pretty much all of our modern high speed ADCs with sample rates above 10Msps or so. For a single ended input signal there would need to be some kind of signal conditioning circuit before the ADC to convert the single ended signal to differential. For input signals with frequency content above a MHz or so the cleanest conditioning circuit is a transformer or balun coupling. But magnetics such as this will not pass low frequencies, so if the input signal is to contain frequency content down to DC then usually an amplifier circuit is used.
Regards,
Richard P.
Hi,
The VCM signal should just be biasing the output of the amp to the desired level so that the common mode is what the ADC wants it to be. The amp should be set up with a certain gain such that the max output signal is less than the full scale range of the ADC. But I would think that there should still be a fixed transfer function of the amp stage such that an input voltage of 'x' relative to ground is known to produce a differential output voltage of 'X', and an input voltage of 'y' relative to ground should produce a differential output voltage of 'Y'. (I am using lower case and upper case to distinguish) So I would think you should still be able to tell from the ADC samples what the original input voltage is if you know the transfer function throughout. At some point though the amplifier forum may begin to be a better place to discuss this aspect of the signal conditioning, though.
Regards,
Richard P.
Hi,
I can have a transfer function like you explain, but I think I need to have an unblanced input on the ADC.
Let say, if my signal have a low level 0.5V and a high level 1.5V (referenced to ground), because I need to center it around the VCM of the ADC, I will have a differential voltage of 1V centered on VCM. Now, if my signal is 1.5V low level and 2.5V high level, result will be the same, am I wrong?
If ADC accept unbalanced signal, for the first signal, I will have for example 1V differential centered on VCM - (or +) x (some mV) and for the second signal a 1V differential centered on VCM - (or +) y (some mV).
So I really need to know if ADC accept unbalanced signals. If not, I have to use another solution more complicated for my system.
Thanks
Matthieu Baque
Hi,
If the differential input voltage to the ADC is such that the INP is at 0.75V below VCM and the INM is 0.75V above VCM, then the input differential voltage would be -1.5V and this would correspond to the lower limit of full scale and the output code would be all zero. If the INP is at 0.75V above VCM and the INM is 0.75V below VCM, then the input differential voltage would be +1.5V and this would correspond to the upper limit of full scale and the output code would be all ones. So the range of input would be +/- 1.5V around VCM or 3V peak to peak. You would need to design your input conditioning circuit such that your minimum input signal would be transformed to the -1.5V differential input to the ADC. And your maximum input signal would be transformed to the +1.5V input to the ADC. Then you would know what your input signal was when it was sampled and turned into the output code.
Regards,
Richard P.