Protecting biases from feedback

[LoveLaika]

I work with some silicon photomultipliers a lot, and they need to be properly biased at the right voltage in order to work correctly. In my cases, this involves adding a voltage at the cathode/anode. I typically pass the bias voltage through a 2nd-order Low Pass Filter (RC), and that's good enough usually. The trick is when you put two or more SiPMs in parallel.

When working with two or more SiPMs, rather than connect the common cathode/anode together, I tend to pass the voltage through their own separate LPF for each separate diode (though there are instances where I can't, when the common bias points are connected together already). I found that if the capacitor values of the LPF are not large enough, the current from one SiPM will affect the others (this was evident when common bias points were already connected together). Thus, increasing the cap values reduces the 'crosstalk'/'feedback' amongst the SiPMs.

However, I wonder if there isn't a better way, to have each bias voltage separate on each SiPM in order to prevent 'crosstalk'/'feedback'. One way I thought of was using an adjustable linear regulator for each sensor. The thing is, the bias voltage that I want will be the same as the input voltage going into each regulator. Theoretically, if I make it so that the regulator has the largest gain via resistor ratios, then the output should be the same voltage as the input. Setting it up like that, would using a linear regulator like that provide some kind of protection?