So when is appropriate to call and not fold?

Make the decision that maximizes your equity.

Sounds like you are a beginner with a question like that, so I will gear my response in that manner. Apologies if that is not the case.

I will talk only about cash games since these are simpler. The same principle applies to match play, but the value of a point changes with changing match score, so that adds an extra layer of complexity. In money play, each point has the same value, so it makes things easier.

The basic principle is that of risk vs reward. If your opponent doubles you have two choices - the safe choice of losing only the current value of the cube or the risky choice of playing on and maybe losing twice as much. The reward for taking on that risk is that you might turn it around and win the game. Like any risk/reward calculation the relevant question is how likely is it that I might win? That is not usually an easy question to answer in backgammon. That comes with experience, practice and study. But what we can do is quantify exactly what probability of winning you need to have to take a cube.

It turns out that in any risk/reward situation, the reward probability must be at least risk/(risk+reward) in order to justify taking the risk. I will focus for clarity on an initial double from 1 to 2; the math is exactly the same for further doubles (at least in money play, but not match play). If you drop you lose 1 point. That is our reference for evaluating risk and reward. If you take and lose you lose 2 points, so your risk is 1. If you take and win, you win 2 points. Winning 2 is 3 points better than losing 1, so the reward is 3. This implies you must win at least 1/(3+1) = 25% of the time in order to take a cube.

That is a simplifothough. There are two factors that modify this. The first is gammons. Your risk and reward should both be modified to account for gammon wins and losses. Usually, it is more likely that you will lose gammon than win if your opponent is foubling, so often we must adjust the takepoint upward to account for the increase in our risk. Again, practice, sgudy and experience are needed to more accurately assess gammon risk.

The other factor is cube ownership. Owning the cube makes it more liikely that you will win. Your opponent cannot use the cube after you take to cash the game. You will be able to play everything out to the end and win a few games here and there you would not have if your opponent had cube access. Conversely you might get to a position where your opponent still has winning chances, but you can force him to give up by doubling yourself. The net effect is that you can sometimes lower your takepoint a bit.