Just eyeballing the results, my conclusion is: in relative terms, yes, the filament does dry faster when injected with dry air. However, in absolute terms, “faster” is still a long time. i.e. not days of drying reduced to just an hour or two, but more like many hours reduced to fewer, but still many hours (so called Hawaiian math: 1,2,3, many).
So, at least for me, the better way to approach it is: dry to asymptote, even using the humid make-up air during the drying process. For most filaments, the result will be “dry enough” for printing. However, for problematic filaments, use that as the new starting point and only then dry further to an even dryer asymptote using pumped dry air. This minimizes the amount of depleted desiccant that will later need to be reanimated. i.e. it’s the lazy, least-effort solution. Either way, it’s a lengthy process, and it seems there’s no real getting around that.
Either way, the biggest time saver would be to figure out a way to pinpoint the right threshold for terminating the drying at a “good enough” filament moisture level rather than spending all the extra time needed to dry all the way to asymptote.
Drying all the way to asymptote will always be a “conservative” standard, but it comes at a price in the time needed. But it is not absolute, in that it itself depends on the ambient humidity. I think it will turn out to be drying to a particular RH number inside the drying chamber, and for most filaments the magic number seems to be less than 20% at ambient temperature (less than 40% for PLA), which then needs to be translated into the equivalent number for the RH number inside the drying chamber. i.e. drying to a particular dewpoint.
Say 15%RH at 20C is the target. Then the dewpoint target is -7.4C.
I’m going to switch to -7.4C dewpoint as my “inside the dryer” target for most filaments and see how it goes. The best way to save time is to terminate the drying as soon as it’s “good enough.”
In the case of the AMS HT, it seems that the temperature reads high by about 3C and the RH reads low by maybe 5%RH. So, converting that target to the AMS would mean retargeting to 10%RH at 23C. So, for the AMS HT, that would mean drying to a dewpoint of -10.3C as the “good enough” number target.
But that would mean drying PETG to the point where the AMS HT reports 0.98%RH after setting the dryer temp to 68C. Would it ever actually read that low? Perhaps there’s some kind of nonlinearity in the AMS HT’s TH sensor that needs to be accounted for.