If you plot T against P for equi-enthalpy states, then below a certain temperature (the maximum inversion temperature, Tinv max), which depends on the gas, the curve has a maximum. If the gas has an initial pressure less than the pressure at this maximum, then throttling to a lower pressure will produce cooling. Above Tinv max, the enthalpy curves have their maximum at zero pressure, and throttling can't produce cooling. For most gases Tinv max is well above room temperature, so throttling from a suitable starting pressure can produce cooling. But for hydrogen, Tinv max = 202, and for Helium, Tinv max is even lower. So at room temperature throttling won't produce cooling. Argon's a bit of a puzzle because, according to M W Z, Tinv max = 723.
Maybe it's possible to understand all this intuitively in terms of very simple ideas, but, as I said, I don't. I think my best chance of doing so would be to look at the mathematical treatment in detail and then look to understand each step in simple terms. But I haven't the time just now.