Hi Kaushik,

When you say that you get variation in the measurements, is that a good thing or a bad thing? I'm not certain if you mean that when the fluctuations are present the instrument should be reading the same value each time but it starts fluctuating such that the measurements become unrepeatable and therefore untrustworthy, or if you mean that when the fluctuations go away that even though the measured value should be changing it starts giving a single value that doesn't change regardless of input.

In either case, given that this is a medical instrument, you really should limit the advise you accept to that from people who are qualified and certified to work on such instruments. Do not take advise from me, or anyone else on this forum for that matter, until and unless they first demonstrate those prerequisite qualifications.

As for the general case of analog and digital grounds, which may or may not be applicable to your case, the reason they are generally kept physically isolated is because of the noise induced into one signal by the return (ground) eddy currents of the other signal. Yes, in general problems can arise from either analog noise inducing into digital signals or vice versa.

Also, in the general case, it is usually desirable to short both grounds together, but to do so in a location where there is no risk of crossover return currents. In other words, both ground nodes will still be physically isolated but electrically united, thus ensuring a common 0-volt level in both.

In your specific case, I have no idea and you should only accept the advise of someone who knows more about your specific instrument than either you or I.

Joe