So I put a roll of BBL TPU in the HT to dry overnight and while it was hot (75C) it was still stuck at around 20% humidity after over 12 hours. By comparison I put some well-wet TPU in my AMS 2 Pro at home and within 4-5 hours it had the humidity down to 8%.
Does this seem normal at all?
Well, its the moisture in the air, not the material, but unless that spool was absoloutely soaking wet, no, it dont seem the norm for me, but I am in a relatively dry room.
It wasn’t, the room stays about 50% humidity or so with the AC running. At home I had the spool open to air for weeks with the windows open and crazy humidity outside so that one was really wet.
I pulled the one at work out yesterday to do some prints and by this morning was already getting bad stringing so I just threw it in the 2 Pro set to 65C.
I don’t even both looking at the humidity in the AMS’ any more. I don’t think its accurate at all. I just go by time its been in there for.
Anything blocking the air vents ? or is the seal bad ?
Yes - totally normal depending on the capabilities of the units and operating conditions.
When you read a humidity on a drying AMS or dryer that is running, it is showing you the balance between water given off by the filament and water being removed from the dryer or AMS.
Wetter filament or higher temperatures will help drive the humidity up. A more efficient/capable dehumidifier or dryer filament will help drive humidity down. There’s a lot wrapped up in “more efficient/capable” even down to the shape of the container, how much dead air there is, etc.
And obviously, if you take a container of air and heat it, the humidity inside will go down. But temperature has a big impact on how fast water comes off filament so hotter drying temperatures increases the humidity in a dryer (at first) in spite of the influence of temperature trying to reduce it. It’s just that there’s a lot more water given off at higher temps.
Also, you can test a dryer or AMS with nothing in it to see just how low it can go because there is no additional load on the dryer than water in the air in the enclosure itself.
Not only how low, but also how fast if you plot humidity vs time.
Those give you an idea of how strong a dehydrator is. Better methods/units will remove more water faster and hit lower humidities.
That’s really how units like these should be compared - empty. When you put filament in them, then things get wonky. You may not dry filament as much as another guy for various reasons and if so, your humidities will be higher since there is more water to remove.
So comparing with filament is “bad”. It confuses the issues and hides actual performance. The only variables reduces to starting/ambient humidity if you benchmark without filament.
Obviously the point is to dry or maintain filament and not just the air. For the purposes of testing though, testing without filament is the way.
Yep. It matches what I am seeing.
The AMS2 is obviously 2nd Gen (with lots of incremental improvements) while the HT is the first of its kind.
I have the feeling that there’s loads more room for improvement (sealing efficiency, connector location and quality, sensor location, airflow,…) for the HT 
& 
Is it possible, that the filament just wasn’t finished drying when you took it out of the AMS HT, so it still released water and kept humidity high. When you then further dried it in the AMS2, you removed the still remaining water. So another round in the HT could potentially have brought you success too.
I think an even better to test a dryer would be to introduce something soaked with a defined amount of water and see how much can be removed in a given time by weighing it? or see, how long it takes to get it to a predefined weight?
What you’ve described is absolutely plausible.
At the most basic, the AMS 2 Pro has a greater air volume, any humidity you drive out of the filament is going to be distributed across at least 5-6x more air volume than the HT.
There are also many potential sources of error in your comparison. Unless the two spools were 1) The same manufacturer 2) exposed to the same humidity for the same amount of time and 3) the same amount on each spool - then you have confounding variables that make this kind of comparison very difficult.
The easiest thing to use in that case would just be silica gel beads. To get the water content in the ball park with filament you could use one of the small bags (10g or so per simulated spool) which would put you somewhere near filament water content.
The defined amount of water can be gauged by humidity which is how the silica gel industry does it. You can buy beads pre-conditioned to a specific humidity which gives you a defined water content.
Under a load like that would be a good test.